¹ See H. Kurath, The loss of long consonants and the rise of voiced fricatives in Middle English, Lg. 32.435–45 (1956).

² There is only one point in this paper that calls for separate comment. The statement (233–4) that being, buying, going, booing, etc. 'exhibit a syllabic structure exactly parallel to that of such words as bidding, bedding ... budding, pudding' is an extremely doubtful assertion, and cannot be used as evidence either for or against a unitary or a binary phonemicization of the diphthongal phones in be, buy, go, boo, etc. Hence it cannot be cited in support of the assumption that the 'free vowels' of be, bay, go, boo, etc. are phonemically complex, and that therefore ‘a strong-stressed short vowel cannot end a syllable’. The authors admit that in English unstressed syllables do end in a vowel (e.g. sofa) ; why not also stressed syllables? Why set the principle of pattern congruity aside in this instance?

³ On objective vowel length in two varieties of American English, see Roe-Merrill S. Heffner and Winfred P. Lehmann, American speech 12.128–34 (1937). On the devastating implications of these objective data for the system of six ‘short’ vowels asserted by Bloch and Trager, see Einar Haugen and W. F. Twaddell, Facts and phonemics, Lg. 18.230 ff. (1942).

⁴ American journal of philology 66.209 (1945).

⁵ I do not overlook the fact that initial clusters of fricatives occur in English in a few learned words taken from Greek, as sphinx, sthenic, phthisis. But I do not admit that this oddity is a part of the English system, and therefore do not regard it as a suitable basis for assuming that English has initial clusters of semivowels.

⁶ In 1941 and 1942 Trager found only six vowels (simple nuclei) in his idiolect of English, exemplified in bit, bet, bat, put, cut, cot. A decade later he found three more (Outline 14) : two rounded back vowels, /o/ and /ǝ/, and a high central vowel, /ᶤ/.

Trager reports that he has /o/ in whole ‘most of the time’ and in home ‘about one-fourth of the time’; that he has /ǝ/ in boss (a hump) and in gloss (translation); that he uses /ᶤ/ in the expression the willies, the preposition until, till, nearly always in twenty, ‘75% of the time’ in the adverb just, and ‘part of the time’ in this (22–3). H. L. Smith is said to have /o/ in the phrase gonna (going to), /ǝ/ ‘most of the time’ in the last syllable of alcohol, /ᶤ/ ‘more than half the time’ in children and ‘40% of the time’ in the adverb just (24).

If this is the best evidence the authors of the Outline can offer, it is shaky indeed—whether it is intended to be exhaustive or indicative of the evidence which they would adduce in support of positing three additional ‘simple nuclei’. The examples cited in support of /ᶤ/ actually have prosodic or positional allophones of the vowel in bit; Trager's /o/ in whole, home is either a rare relic or an isolated acquisition from another dialect, Smith's /o/ in gonna a prosodic allophone of the vowel in bone; on the meager evidence offered, the status of /ǝ/ is obscure, but it would seem that in any ‘simple‘-vowel system a rounded low-back or mid-back vowel must be included.

The plea in support of adding /o/ and /ᶤ/ to the vowel system for the sake of achieving an analysis of ‘the total pattern of all the dialects’ (9) is irrelevant. There is no such thing as a ‘total pattern’ that can be imposed upon all dialects of American English (not to mention the dialects of England, Scotland, and Ireland). The linguist must analyze the system of each dialect separately before he can know what systematic features are shared by all dialects, or by groups of dialects. He must distinguish between the systematic features and sporadic unsystematized features of each dialect, since every dialect has elements that are not built into the system. To regard unsystematized features as part of a ‘system’ and to impose an ‘over-all pattern’ are spurious notions that must be rejected.