Published online by Cambridge University Press: 15 February 2011
In Modern English, function words such as this, that, and the are pronounced with a voiced onset, while content words have the original voiceless onset. A statistical analysis of the distribution of <ð> and <þ> in Vices and Virtues, a text preserved in London, British Library, Stowe MS 34 (circa 1200), reveals a distribution, whose most plausible interpretation is that these two letters were used to encode stress-conditioned differences in voicing. This sets the voicing of function words, normally dated to the fourteenth century, back to circa 1200 or earlier.*