Accounts of institutions and private individuals between the Norman Conquest and about 1500 were routinely written in a non-random mixture of Medieval Latin, Anglo-Norman, and Middle English. If the base language was Medieval Latin, then only nouns, stems of verbs, and certain semantic fields such as weights and measures could appear in English or French, with all the grammatical material in Latin and English and Anglo-Norman nouns, verbs, and adjectives Latinised by adding a suffix, or an abbreviation sign representing a suffix. If the base language was Anglo-Norman, then only the same restricted semantic fields and nouns and stems of verbs could appear in English. This situation changed over time, but was essentially stable for almost five hundred years. The chapter asks why, if English words could easily be assimilated into a Latin or French matrix by means of suffixes or abbreviations representing suffixes, were all English words not assimilated? Why did letter graphies such as <wr->, <-ck>, <-ght> persist in mixed-language business writing? One effect is to make the text-type of business writing very unlike any other genre—half a glance is all it takes to recognise a mixed-language business document and that may have been an advantage.