1 part I (Vocabulary) and Part II (Grammar and Syntax) of this series appeared in the JCLA, 1 (Oct. , 1954), 13-8 and (Mar., 1955). 14-9. The information comes from three sources, which accounts for the disparity in the total number of responses from item to item: two questionnaires (one circulated at Queen's University in 1949-1950, the other at R.M.C. in 1954-5), and, where relevant, the records of the Linguistic Atlas for eastern and western Ontario.

2 For a detailed discussion of these difficulties see Cassidy, Frederic G., “A Method for Collecting Dialect,” Pub. Am. Dialect Soc., 20 (Nov., 1953), 5–96 Google Scholar, which includes a fully developed questionnaire.

3 The symbols used here are substantially those of the International Phonetic Association. Since only broad distinctions are here dealt with, phonemic notation is generally used; where subphonemic variations are discussed or where the phonemic status of the item under discussion is obscure, broad phonetic notation is substituted. The following partial key, based on my Ontario speech, might be helpful: (vowels) / i / as in beat, / I / bit, /e/ bait, /ε/ bet,./æ/ bat, /a/ balm (/bam/, midway between the /æ/of lather and the /a/of father), /α/ bot (also bought in my speech), /b/ bog, /ɔ/ law (the last three vowels are probably not phonemically distinctive in my dialect), /o/ boat, /U/ book, /u/ boot, /iu/ beaut, /ə/butt, Bert, sofa, duffer (/bət/, /bərt/, /'sofə/, /'dəfər/),/aI/ bide, /aU/ brow, /ɔI/ Boyd. The consonant sounds not represented by symbols having the conventional alphabet values are symbolized as follows (all examples having the relevant sound in final position): /θ/ bath, /ǒ/bathe, /∫/ dish, / ʒ / mirage, /t^∫/ ditch, ridge, /ŋ/ rang, / ’ / main stress symbol placed before stressed syllable as in above /ə'bəv/, /’/ secondary stress symbol placed before stressed syllable as in /'sεkrə, tεri/. The symbol /j/, representing a phoneme which does not occur in final position in English, may be illustrated by you /ju/ beyond /bI'jαnd/.

4 In referring to a pronunciation as “usual” in Ontario, Britain, or America, I do not deny its currency elsewhere, I simply mean that to the best of my knowledge the form is the most commonly used by educated people in Ontario, in Britain (SRS) and in America (Northern speech area).

5 The form /vez/ is occasionally heard on the American side of the border; it is, according to the Linguistic Atlas records, unusually common in metropolitan New York.

6 See McDavid, Raven I. Jr., “Postvocalic r in South Carolina: A Social Analysis,” American Speech, XXIII (1948), 195–203 Google Scholar.

7 See Penzl, Herbert, “New England Terms for Poached Eggs,” American Speech, DC (1934), 90-5CrossRefGoogle Scholar.

8 See Avis, Walter S., “Crocus Bag: A Problem in Areal Linguistics,” American Speech, XXX (Feb., 1955), 5–16 CrossRefGoogle Scholar.

9 The final / r / often heard from British and other r-less speakers in Indian /'Indiər/, idea /aI'diər/, alpha /'ælfər/, beta /'betər/, etc. is the result of a somewhat different process. Such speakers pronounce the isolated words father, mother as /'fαǒə/, final historical r having been “lost” in their dialect. Thus the final vowel of these forms, /ə/, is identical with that in /aI'diə/ and /'Indiə/. When these speakers say father and mother, that is, where the word father precedes a word beginning with a vowel, the “lost” final r is restored to function as a hiatus-breaker, /'fαǒər an 'məǒa/. This so-called “linking r”, which in the speech of many people occurs regularly under these circumstances, is transferred by some speakers to such expressions as the idea of it /ǒi aI'diər əv It/ and India Office/'Indiər,ofIs/, where /r/ has no historical justification. This type of intrusive r also finds its way into the speech of persons who speak dialects having final /r/, for when idea or India, being often heard in the linking sequence, are spoken by these people, they are on occasion pronounced with a final r, just as in father /'fαǒər/ and mother

10 This Canadian pronunciation is frequently reflected in student spellings and occasionally appears in print: “… [he] wore glasses, kharki-colored trousers and jackets….” The North Shore Leader, Newcastle, N. B., April 8, 1955, p. 2, col. 4. I am indebted to Mr. C- J. Lovell for this citation.

11 Since the pronunciation of the second syllable of produce was not enquired into, I can offer no information as to how prevalent the American pronunciation /'produs/ is in Ontario. My personal observations indicate that this pronunciation, like /'ævənu/ for avenue, is not very often heard from Ontario speakers.

12 An English Pronouncing Dictionary (London, 1919), p. 303 Google Scholar.

13 I have heard at least one American who regularly uses /aI/ in these prefixes say /kən,stæntaI'nopəl/ for Constantinople, an interesting hyper-urbanism indeed.

l4 See Robertson, Stuart (revised by Frederic G. Cassidy), The Development of Modern English (New York: Prentice-Hall, 1954), pp. 396-7 and fn. 45Google Scholar. Chapter 12 contains a large number of up-to-date general statements concerning American regional usage based to a large extent on Linguistic Atlas records. It is a matter of interest that dictionaries do not record rooves as an alternative plural form of roof, the stated or implied plural being roofs. Questioned as to their usage, 76 of 139 informants claimed to use rooves, 63 roofs. For the plural of hoof 89 offered hooves, and 50 hoofs. The latter is the form offered by the dictionaries, though some list the former as a “rare” variant. Whether or not the extensive use of roove s and hooves reflects a grammatical usage peculiarly Canadian has not been determined; personally, I would be inclined to doubt it, in spite of the dictionaries.

15 In discussing the divergence between the British “broad α” and the American “flat α”, Mencken, H. L., in The American Language (New York, 1937), p. 338 Google Scholar, claims that in Canada the American has displaced the British pronunciation. This conclusion is highly questionable, for there is no reason to believe that the British “broad α” was ever generally used in Canadian popular speech.

16 see McDavid, Raven I. Jr., “ h before Semivowels in the United States,” Language, 28 (1952), 41–62 CrossRefGoogle Scholar.

17 ‘See Bagby Atwood, E., “ Grease and Greasy: A Study of Geographical Variation,” Univ. of Texas Studies in English, 29(1950), 251-3Google Scholar.

18 There are of course a large number of items which are similarly pronounced by most speakers of American and Canadian English but which are differently pronounced by most speakers of Standard British English. In the following examples the usual North American pronunciation is given first: privacy /'praIvəsi/,/'prIvəsi/; solder/ ‘sαdər/, /'soldə/; questionnaire /,kwεst∫an'εr/, /,kεst∫an'εə‘ garage /gə'rαʒ/ or /gə'rαdʒ/, /'gærαʒ/ or /'gærIdʒ/ (frequently heard in Ontario is /ge'rædʒ/, a pronunciation not recorded in the dictionaries); says /sεz/, /sez/; abdomen /'æbdamən/,/æb'domən/; secretary /'sεkrə, tεri/, /'sεkrεtrI/ (the secondary/'sεkən,dεri/ stress on the penultimate syllable of similar words ending in -ory, ary, dictionary, ordinary, obligatory, reformatory, etc. , is characteristic of American and Canadian English). Needless to say, this list could be lengthened considerably and no doubt another (probably shorter) could be compiled of words pronounced by Canadians after the British fashion. Such compilations would have value as a basis for a long overdue survey of Canadian pronunciation.

19 For discussions of this phenomenon in the United States see Oswald, Victor A. Jr., “Voiced t —A Misnomer,” American Speech, XVIII (1943), 18–25 CrossRefGoogle Scholar; and Lehmann, W. P., “A Note on the Change of American English /t/, “ American Speech, XXVII (1953), 271-5CrossRefGoogle Scholar.