I shall be discussing William as a writer of Latin, and as a virtuoso in that language. When he expounds his aims at the start of the Gesta Regum (bk 1 prol. 4) he talks of his wish to ‘give a Roman polish to the rough annals of our native speech’; and in that same prologue he comments on the styles of Bede, .thelweard and Eadmer. Every page he himself wrote shows his wish to give his own turn to the material he had so laboriously amassed. It is partly a matter of style. I have written a little on this, but there is much more to be done, not least because William's manner varies from book to book and even from passage to passage. But it is also, it seems to me, a matter of language, and especially of vocabulary. That too I have touched on in the past. More recently, I have embarked on a systematic investigation of William's words, of which these are the first fruits.
Work of this kind on a medieval text would hardly have been possible until quite recently. In a classical Latin author it is commonplace; there have for many years been concordances of the major authors, not to speak of the Thesaurus Linguae Latinae, scheduled for completion, after my time, in 2075. William's vocabulary could be worked on properly only after Martin Brett made it his job, more than twenty years ago, to commission a private concordance of William's historical and hagiographical works. The now yellowing and battered print-out has been of inestimable service. It unfortunately came before the full digital age, and the concordance cannot be searched. But as time has gone on I have accumulated searchable texts of most of the major works, with the important exception of Gesta Regum, for which I still rely wholly on Brett's concordance.
One further preliminary: to have any validity the work I have been doing requires proper comparative material. But the sort of authors with whom one would want to make a comparison, especially in the twelfth century, themselves lack concordances. Nothing can be more wonderful than the Brepols Cross Database, and I have made constant use of it. But it does not yet cover the English writers whose practice would be of such interest.