¹ Grateful acknowledgments : Navaho data were collected with the support of the Southwest Project in Comparative Psycholinguistics, of the Social Science Research Council, in 1955 and 1956; a research training fellowship of the Social Science Research Council, in 1959–60; and the Department of Public Health of Cornell University's Medical College, which maintains clinical research facilities at Many Farms, Arizona, on the Navaho Reservation, where I lived during most of 1959–60. I have exchanged views about these data at a meeting of the Yale Linguistic Club in 1957; at the Allerton House Conference of Southwest Project members and others at the University of Illinois in 1957; and at the Northwest Anthropological Conference at the State College of Washington in 1958. I have profited also from private discussions and correspondence with J. B. Carroll, J. B. Casagrande, E. H. Lenneberg, D. French, G. Motherwell, B. H. Smeaton, and others. Responsibility for the views presented below, however, is my own.

References (excluding most dictionaries and word lists): F. Boas, Kwakiutl grammar 220–2, 224 fn. 14 (H. B. Yampolsky, ed., Transactions of the American Philosophical Society 37, Pt. 3, 1947); L. Bloomfield, Language (1933); C. Cherry, On human communication 279 n., 294 (1957); M. M. Clark, Left-handedness 41–2 (1957); P. Garvin, On the relative tractability of morphological data, Word 13.12–23 (1957); E. F. Hahn and E. S. Hahn, Stuttering (1956); B. Haile, A stem vocabulary of the Navaho language 1.33, 104, 262 (1950), 2.278 (1951); C. F. Hockett, A course in modern linguistics (1958) ; R. Jakobson and M. Halle, Fundamentals of language 56 (1956); F. Kainz, Psychologie der sprache 3.105–6 (1954), 4.150, 484, 491 (1956) ; H. M. Kaplan, Anatomy and physiology of speech 42–3, 339 (1960) ; C. Kluckhohn in G. Lindzay, Handbook of social psychology 944 (1954); D. C. Leighton and C. Kluckhohn, Children of the people 14 (1947); H. Maclay and C. E. Osgood, Hesitation phenomena in spontaneous English speech, Word 15.19–14 (1959) ; J. G. Miller, Information input overload and psychopathology, American journal of psychiatry 116.695–704 (1960); S. S. Newman, The Yawelmani dialect of Yokuts, in H. Hoijer et al., Linguistic structures of native America [LSN A] 235, 237 (1946); C. E. Osgood and T. A. Sebeok, Psycholinguistics (1954); W. Penfield and T. Rasmussen, The cerebral cortex of man 87–108, esp. 93, 94, 96 (1952); L. Roberts in Cecil and Loeb's Textbook of medicine 1442 (1959); S. Saporta, Morph, morpheme, archimorpheme, Word 12.9–14 (1956) ; G. Stein, A novel of thank you 5, 249, 248 (1958) ; J. L. Stewart, The problem of stuttering in certain North American Indian societies (Journal of Speech and Hearing Disorders, Mono. Supp. 6, 1960); M. Swadesh, Problems of long-range comparison in Penutian, Lg. 32.17–41 (1956); B. L. Whorf, The Milpa Alta dialect of Aztec, LSNA 386; G. L. Wyatt, Stammering and language learning in early childhood, Journal of abnormal and social psychology 44.75 (1949); R. W. Young and W. Morgan, The Navaho language 57–8, 206, 207 (1943).

² The term stuttering is used in this note as a cover term for a variety of phenomena, and some readers may feel that these phenomena have been lumped together wrongly. This lumping, however, is convenient, and it does not affect conclusions reached later about morphological elements. I express no opinion about speech pathology. In fact, use of a cover term enables me to avoid involvement in certain problems which are not germane to linguistics, e.g. stammering vs. stuttering (some authorities see no difference ; others do—see Wyatt, Hahn and Hahn, and Kaplan) or stuttering vs. normal disfluency (some authorities regard as touchstones of stuttering an awareness by speaker and hearer of disfluency as a psychological problem, overt anxiety, muscular contortions, and anguish on the speaker's part; cf. Stewart 6). My interest is in unconventional allomorphs which have statistically rare phonemic shapes. In the present section, my purpose is to suggest that such allomorphs occur in a large number of languages.

³ Cf. Clark, who cites Hildreth's belief that ‘stuttering is only found in civilised races where reading and writing are taught, and where there is rigid training in social manners involving manual acts’ and Chrysanthis' belief that racial factors affect stuttering and that stuttering is unknown among the Chinese.

⁴ Wakashan (Nootka, Kwakiutl) and Salishan (Cowichan, Nanaimo, Tcil'Qē'uk, Siciatl, Stlatlumtl) words are given by Stewart 42, 62, in an interesting attempt to relate the supposed absence of stuttering in Navaho and Ute groups to the absence of 'genuine' verbal reference to stuttering, and cultural lack of concern about stuttering, and to relate the presence of stuttering in Wakashan and Salishan groups to the presence of words for stuttering and concomitant cultural concern about stuttering.

⁵ Cf. Young and Morgan, and Haile: A stutterer is called ^?alt'aniЗihí ‘the one who stutters’; the noun is derived from a third-person neuter verb ^?att'aniЗih ‘he stutters’ (stem -Зih 'breathe [in or out]') and an enclitic í 'the one (who or which ...)'. A Navaho may confess ^?att'anisЗih 'I stutter'; or he may be teased with nicoo^? yee yánítti^? 'you're talking with your tongue!' Cf. Leighton and Kluckhohn: some Navahos observe a pregnancy taboo against the breaking of a pot by a prospective parent, lest the child 'stutter or have other speech difficulties', and they maintain the prophylactic custom of breaking a pot above the newborn baby's head ‘in such a way that the pieces fall together in a pile beneath him’.

⁶ A morpheme as understood here is a class of semantically similar and noncontrastive allomorphs. For some implications and difficulties of pertinent criteria see Saporta and Garvin.