Sink or Swim?: a homonymic dilemma in Medieval French

Some years ago my colleague the late Frederick Whitehead wrote a paper dealing with the homonymic clash which has generally been supposed to lie behind the disappearance of the Old French verb noer 'to swim'. The loss of this word, it was claimed, caused the related restriction in meaning of nager from both 'to sau' and 'to swim' in the Middle Ages to the single modern sense of 'to swim': this restriction in its turn was said to have called into play the modern naviguer to take on the now vacant sense of to sau'. In his paper, read to the X Congres international de Linguistique et Philologie romanes at Strasbourg in 1962 and published in the Actes ... in 1965, Whitehead was at pains to show that such attempts to explain developments in historical semantics by recourse to the theory of homonymic clash were not satisfactory, because they concentrated upon the phonetic factor to the exclusion of all eise. This explanation by phonetics which Whitehead called into question was propounded most authoritatively by von Wartburg in his article Betrachtungen über die Gliederung des Wortschatzes und die Gestaltung des Wörterbuchs [ZrPh 57, 1937, 296-312] and summed up in succinct terms in the FEW: