How representative are the teaspoonfuls of prose analysed in chapter 2, 8,000 words drawn from the great rivers of English published between 1700 and 1725, or between 1775 and 1800? Those who write grammars for schools are probably more sensitive to the orderliness of their prose than most people, and more conscious of correctness and formality. If gentrification was a linguistic as well as a stylistic phenomenon, if it affected most of the prose of the time, it would show up in proletarian as well as in academic texts, in novels and in stage plays, in philosophers and poets. Changes that look like gentrification may be detected by relatively cursory examination of major authors from the two ends of the century: Sheridan's dialogue is cleaner and more elegant than Steele's; Jane Austen's sentences are more periodic and her paragraphs more coherent than Defoe's; Godwin's critical essays are fussier and more prolix and more precise than John Dennis's. But what about the most careful and sophisticated writers of the first two decades of the eighteenth century, such as the third Earl of Shaftesbury or Alexander Pope? What about less-educated writers of the last two decades of the century?

The texts I have chosen to test the “gentrification and literacy” model are intended to be both representative and diverse; they speak for their age, but with more than one voice.