The ‘facts’ and methods of historical linguistics

Reconstructing earlier linguistic states inevitably takes us on a slippery road, where ‘facts’ are harvested from accidentally surviving written records. We glean our information from orthography, from contemporary linguistic commentary, from typological comparisons, from loanwords and other language-contact evidence. These sources complement each other; added to them is the evidence extracted from the properties of verse: alliteration, rhyme, syllable count, the positioning of words in the line. This contribution will illustrate and evaluate the diagnostic powers and limitations of a subset of these sources in producing testable hypotheses about the history of English. Ideally, the techniques of gathering and interpreting historical stress data in English will be applicable to any language with extensive orthographic and metrical records.

The area of interest in this chapter is prosodic change. ‘Prosodic’ is a broad linguistic term, encompassing the organization of segments into syllables, the properties of syllables, the assignment of stress, tone, and intonation, and the interaction between phonology and morphology. Of all these, the scope of coverage here will be narrowed down to the reconstruction of stress.