Moreover, the use of so specific a label tends to lead to unsustainable assumptions about the variety's unity and homogeneity. It is far safer to think in terms of a range of speakers from various dialectal backgrounds, by no means all Norman; since their speech is of course not recorded, the diversity of it (both regional and social) is equally unattested.
The written records from the Conquest onwards are, it is true, relatively homogeneous, but this is in part the result of the writing process itself. The act of writing tends inevitably to lead to some measure of standardization. Nonetheless, these written records display certain striking features.
In the first place, they are early: medieval French literature first appears in medieval England, and some of the first non-literary documents in French (charters, etc.) are in Anglo-Norman. The most likely explanation for this is that there was a long-standing insular tradition of vernacular writing of religious, literary and historical texts, which the newly-arrived Normans adopted. Over time, the use of Anglo-Norman for these purposes expanded further into the fields of law, administration, commerce, and science, in all of which a rich documentary legacy survives.
But in parallel with the development of Anglo-Norman as a 'language of record' (Michael Clanchy's term), at first accompanying, later ousting Latin, before the advent of English which in its turn displaced Anglo-Norman, the language became less and less of a true vernacular, and increasingly an acquired, second language. In many cases, of course, it was only imperfectly acquired, and it is these texts which have fuelled the idea that all later Anglo-Norman is little more than a degenerate jargon.
How far this is from the truth may easily be seen from the wide range of documents well into the fifteenth century in which Anglo-Norman is used for complex administrative matters and indeed affairs of state, at home and abroad. At an international level, many Anglo-Norman diplomatic documents are virtually indistinguishable from the products of the Paris Chancery - a fact which (together with the substantial evidence of the use of Anglo-Norman in Gascony) rather undermines the notion, still current, that the insular variety of French was cut off from its continental roots after the loss of Normandy in 1204.
Yet as well as continuing as a written language of record for all sorts of purposes right through the Middle Ages (and in the case of Law French, beyond), in a determinedly multilingual context, it is clear that Anglo-Norman must also have penetrated sufficiently into all social classes to ensure numerous borrowings into various English dialects. On the one hand the bulk of the Anglo-Norman influence on the lexis of English can probably be attributed to the trilingual scribes in charge of records of all sorts from the late thirteenth century onwards; on the other, there is a layer of vocabulary (of lower status) not so readily explained by this process.
Whatever the route of transmission of individual words, though, one thing is clear: the Norman Conquest has had a lasting and dramatic impact on the development of English. Anglo-Norman is at once an important dialect of medieval French which cannot be dismissed as peripheral in any sense of the word, and a major constituent element in English.