This post by John Crace on the Guardian Arts blog is unfortunate. It contains a claim by a former lecturer of mine concerning Milton's contributions to the English language:
According to Gavin Alexander, lecturer in English at Cambridge University and fellow of Milton's alma mater, Christ's College, who has trawled the Oxford English Dictionary (OED) for evidence, Milton is responsible for introducing some 630 words to the English language, making him the country's greatest neologist, ahead of Ben Jonson with 558, John Donne with 342 and Shakespeare with 229.
There are two things wrong with this. The first one Gavin Alexander concedes quite readily, to some extent: 'The OED does tend to privilege famous writers with first usage,' he writes. This is true, though it's too often been used to create the mistaken impression that famous authors have been given first dibs because they're famous. Not so: the OED's policy has always been to include the earliest available example for a sense that may be found. It's easy to forget, in these days of EEBO, Google Books, etc., that during the initial round of editing (the first edition was published in fascicles from 1884 to 1928) there were quite severe restrictions on what was available in consultable editions or concordances, especially for the medieval and early modern periods. Though it's now routinely possible to antedate these first examples from Milton, Shakespeare, et. al., it wasn't then. It's a case of material constraint rather than editorial bias.
But the other mistake is more serious, and has begun to be pointed out in various quarters. The numbers suggest that Gavin Alexander has searched for 'Milton' and 'Shakespeare' in the 'first cited author' module available in the OED Online advanced search. What he's neglected to realize, though, is that until the current round of revision (from 2000 onwards) the OED abbreviated 'Shakespeare' as 'Shakes': type this in, and you will get a further 1663 results for Shakespeare as first-cited author. Milton does not even come close to second place, either: searching 'Chaucer' returns 2006 results, 'Cursor' (for Cursor Mundi) has 1609, 'Lydg*' (for John Lydgate) 783, etc.
A definitely cautionary tale, this one, about the dangers of hasty corpus linguistics. (Not a phrase I thought I'd use today.)
# Alex Steer (14/03/2008)