The University Library of Oslo has a collection of letters from Henry Sweet to the Norwegian phonetician Johan Storm written over a period of twenty years from 1876. We here present extracts from the letters, selected with a view to showing Sweet's views on the terms narrow and wide during this period.

A sixteenth-century sound change in Scots has resulted in the appearance in that language of a system of vowel-length variations strikingly different from the common Germanic system still visible in other West Germanic speech forms. The present situation is described by the set of rules known as Aitken's Law (Aitken 1962 and 1977; see also Lass 1974): all vowels and diphthongs are long in stressed open syllables, before voiced fricatives and /r/, and before morpheme boundaries; and short elsewhere; with the two exceptions of /ι/ and /ʌ/ which are invariably short.

It is common knowledge that when one language borrows from another, the borrower often would incorporate the loan words into its phonological system by substituting ‘alien’ sounds by those from its own stock, breaking up consonant clusters to conform to its syllable structure and so on. In general it is not impossible to predict fairly accurately what a loan word would sound like in a language if one is familiar with the phonological systems of the lender and borrower. However, syllable structure and segments are just part of the picture. Other considerations like stress and tone would also be important if one is dealing with a stress language or a tone lauguage. The aim of this paper is to examine English loan forms in Cantonese in order to discover what happens when words from a stress language like English are borrowed into a tone language like Cantonese.

In all the formal descriptions of Haitian Creole published so far the phonemics of the vowels includes only front unrounded and back rounded varieties:

Analysts confronted with the presence in the inventory of the front rounded series attribute to them the status of exceptions or loans.

I have to confess to being to a degree pained at having it seems touched a nerve of my good friend Julian Pring's in my comments on linking /r/. He suggests that my article on the subject in its zeal to counter long-current false notions of the facts of usage has not escaped some degree of tendentiousness. I should be less than candid if I failed to grant that he has something of a point to make but equally if I let my high regard for him make me withhold the counter-comment that his criticisms strike me as by no means more successful than I was in avoiding tendentiousness in that article. I am at least glad to see that it is only on the sociological status of the so-called ‘intrusive’ /r/ that we are in disagreement.