¹ For the distinction between accent and stress, see Bolinger, A theory of pitch accent in English, Word 14.109–49 (1958), esp. 145.

² Intonation and emphasis, Miscellanea phonetica.

³ An outline of English phonetics §1047 (New York, 1956).

⁴ Newman, writing earlier—Word 2.172 (1946)—said, ‘Phonetically, the contrastively accented word is pronounced with an especially strong force of articulation.‘

⁵ The intonation of American English 45.

⁶ Lg. 25.278 (1949). Haugen takes Newman to task for presumably making a categorical distinction between contrastive and noncontrastive accent. Newman, however, did not separate the two so sharply: ‘The heavy stress which functions as the nucleus of an intonational unit or as the locus of an expressive accent will be termed the nuclear heavy stress’ (Word 2.175)—nuclear heavy includes contrastive.

⁷ Cf. Chomsky, Halle, and Lukoff, who exclude from their analysis—For Roman Jakobson 1956 78—‘all forms of expressive accent, including contrastive accent’. Also R. B. Lees, Word 16.123 (1960): ‘any under contrastive primary accent’. (Again I substitute accent for stress.)

⁸ Personal letter, 10 April 1956.

⁹ The field of Yiddish 2 (New York, 1954).

¹⁰ The ‘shift ‘ noted by Trager-Smith 73 refers to a change in the position of the accent within the utterance. It is not the same as the shift noted here, which affects the position of the potential for accent, i.e. the stress, within words.

¹¹ The obvious exception is the suffixes -less and -ful. But most pairs carrying them have ceased to be—if they ever were—true contraries: we speak of a cheerful person, but a cheerless room; the person himself may be soulless, but only his eyes are generally soulful; mindful means ‘remembering’, mindless ‘idiotic’. Many, perhaps most, morphs carrying one of the suffixes do not carry the other: beautiful, hapless, penniless, spiteful, gainful. Finally, -less is productive, -ful is not: footless, eyeless, coined penless, etc.; this may reflect the homonymic competition of -full: eye-full, drawer-full, cupful, etc.

¹² We would be less surprised by regimént than by regimén, because of the support that regimént would have (even though in a different category, as verb rather than noun) from the numerous pairs in which a verb is distinguished from an adjective or a noun by reason of having its final syllable either lengthened or stressed. This is to say that a stress shift can occur more readily if the result is some morphemic shape already familiar than if it is relatively unfamiliar. Examples of distinctive length are affiliate [-ejt] verb versus affiliate [-ṭ] noun, separate [-ejt] verb versus separate [-ṭ] adjective, compliment [-ment] verb versus compliment [-mṇt] noun. Of stress, a number of examples have already been cited. A. A. Hill, in his note on Stress in recent English as a distinguishing mark between dissyllabics used as noun or verb, American Speech 6.443–8 (1931), was of the opinion that new nouns could be derived from old verbs by the process, but not new verbs from old nouns. While it is true that examples of nouns from verbs are overwhelmingly more numerous, there are a few recent creations of verbs from nouns exhibiting both the length and stress features already illustrated. Certificate [-ejt] comes from certificate [-ṭ], and a favorite among linguists is the verb to segment [-ment] from segment [-mṇt], where I waver between [ségment] and [segmént] in my own speech and have heard both pronunciations from others. I would also readily use the verb to fragmént. The area of -ment is perhaps the most active one, with new pairs supported by old ones like augmént-áugment (only stress distinguishes these, as both have two long syllabics), régiment [-ment] - régiment [-mạnt] (length only), etc.

¹³ Word 12.252 (1956).

¹⁴ This is partly the result of the close match between the stressed -bil- of intelligibility and the subdued but nevertheless prominent -bil- of unintelligibility, which reduces the contrast.

¹⁵ 14 April 1956, p. 36.

¹⁶ I have since heard the identical shift on the stalactite-stalagmite pair, under circumstances similar to those of the cartoon. (Conversation recorded May 1957.)

¹⁷ E.g. Station WRCA, New York, 8:28 p.m., 18 December 1956.

¹⁸ It might be said that the accent here is emphasis on the meaning of the utterance as a whole rather than any of its parts. Cf. Bolinger, English stress: The interpenetration of strata 301–2, Study of sounds (Tokyo, 1957).

¹⁹ An instance of how these two pressures can meet head-on occurred to me when I was about to voice the following complaint : The other day those peanuts were only nineteen cents a pound; today they're thirty-níne; I fumbled the accent on the last word. The aim was a contrastive accent on thirty-nine, such as might easily have worked in contrasting it with twénty-nine, but this would not work against the background of the already stress-shifted níneteen (shifted by reason of the phrase níneteen cénts, as against an unshifted They cost ninetéen), carrying an accent on the element that was not in contrast.

²⁰ Unless stress is defined as prerequisite to “word” status.

²¹ In the latter case it is virtually interchangeable with Why don't you make me a présent of it?—no contrast on me in the other sentence, and no contrast on present here.

²² Cf. Newman 178: “the degree of emphasis is actualized not only by the intensity of the nuclear heavy accent itself but also by the contrasting weak intensity of the subordinate heavy accents occurring in the same prosodic context.” (Accent substituted for stress.) In an utterance like [Hey, Joe!] I want you to hélp me tomorrow with no accent on tomorrow, there is no reason, in the absence of more context, to infer a contrast; but if generally unaccented tomorrow receives the accent, we-almost necessarily do infer contrast. (Not with absolute certainty, however. In

the accent merely covers something previously unspecified. Its function is the general one— which includes that of contrast—of marking the point of least redundancy.)

²³ A contrary example : in the following, the main accentual prominence falls elsewhere than on the contrasted words : It would be a terrible blow to American unions if the conspiracy device again comes into common usage in the prosecution and persecution of unions (Frontier, Sep. 1960, p. 12).

Trager and Smith, if their application by Stockwell represents their thinking, incorporate a contrastive accent within other primary accents (I substitute accent for stress to conform to my definitions here) so long as not more than one occurs in a given macrosegment, but ascribe additional contrastive accents to an overlaid voice-qualifier. See Lg. 32.376 (1956), and see Trager and Smith, An outline of English structure 73 (Norman, Oklahoma, 1951).