This chapter introduces the first concrete application of the premise that poetic meter, in all of its diverse forms, is a valuable source of historical information on the language it deploys. Having outlined the structure of the various metrical types which form the basis of the source material, I will now use verse evidence to assess and challenge some established positions concerning the evolution of the velars and the palatals in the history of English. The hypothesis that will be tested is that /tʃ/ and /ʃ/ were not phonemes in English until after c. 1000. The phonemic split of early OE [Ɣ] into /j/ and /g/ and the merger of the voiced palatal fricative [ʝ] with the pre-existing /j/ occurred around the middle of the tenth century.

The Old English consonant system

The starting point for the discussion of the segmental histories in this chapter is the reconstructed consonantal system of Old English. The standard philological reference books are unanimous about how the graphemic–phonemic correspondences in Old English should be represented. Except for the terminology based on phonemic distinctions used in the second half of the last century, it is largely nineteenth-century scholarship that must be credited with setting up the inventory of Old English consonants still used today. Moreover, with some notable exceptions, such as the simplification of long consonants and consonant clusters, and the individual changes affecting /x/ and /r/, and possibly /l/, the English consonants are assumed to have had a rather bland and uneventful history.