Many years ago I bought and enjoyed Molly Lefebure‘s book Samuel Taylor Coleridge: A Bondage of Opium, and ever since I’ve wondered vaguely about her surname — it was obviously equivalent to French Lefèvre, but how did that -u- get in there? Now it occurred to me that with the resources of the internet I could probably find out, and sure enough, I found that Saussure (see this LH post) had addressed the issue in his Cours de linguistique générale (Course in General Linguistics):
Ainsi pour le nom de famille Lefèvre (du latin faber), il y avait deux graphies, l’une populaire et simple, Lefèvre, l’autre savante et étymologique, Lefèbvre. Grâce à la confusion de v et u dans l’ancienne écriture, Lefèbvre a été lu Lefébure avec un b qui n’a jamais existé réellement dans le mot, et un u provenant d’une équivoque. Or maintenant cette forme est réellement prononcée.
So for the family name Lefèvre (from Latin faber) there were two spellings, one popular and straightforward, Lefèvre, the other learned and etymological, Lefèbvre. Thanks to the confusion of v and u in the former writing system, Lefèbvre was read as Lefébure, with a b which never really existed in the word, as well as a u which came from an ambiguity. And now this form is actually pronounced.
He goes on to complain that in Paris, you can already hear sept femmes [seven women] with the t pronounced, which apparently appalls him (an odd attitude for a linguist); frankly, I had no idea it was ever silent. Thus we see once again the pointlessness of peevery!
Incidentally, the English name Lefebure is pronounced with the stress on the first syllable (LEFF-ə-byoor), whereas Lefebvre is lə-FEE-vər (according to the BBC Pronouncing Dictionary of British Names).