¹ From discussions on alternative morphological procedures in general (and specifically as applied to the organization and reorganization of the approach to the problem treated in this paper), I have gained much from those who specialize in methodology. I am glad to acknowledge my indebtedness to Z. S. Harris, W. D. Preston, Bernard Bloch, R. S. Wells, Morris Swadesh, Roman Jakobson, George Herzog, Y. R. Chao, and F. K. Li. The methodology reflected in the standard descriptive works on Siouan is curiously homogeneous; compare, for example, Franz Boas and Ella Deloria, Dakota Grammar 1–183 (Memoirs of the National Academy of Sciences, Vol. 23; Washington, 1941); Franz Boas and John R. Swanton, Siouan (Dakota) 875–965, including Paul Radin's treatment of Winnebago (Handbook of American Indian Languages, BAE Bull.40, part 1; Washington, 1911); William Lip-kind, Winnebago Grammar (New York, 1945); William Whitman, Descriptive Grammar of Ioway-Oto (IJAL Vol. 13, No. 4. 1947), hereafter cited as Whitman.

² Bernard Bloch, Studies in Colloquial Japanese I, Inflection (JAOS 66.97, 1946): ‘An inflected form, in Japanese as in other languages, is a member of a closed group or set of words called a paradigm.' Footnote to this passage: 'Stated more pedantically, a paradigm is one of a number of closed sets of words, the words in each set being related but different in both form and meaning, in such a way that the differences are parallel from set to set.’ Notable also is Bloch's definition of zero in Japanese (98): 'Zero counts as an ending if it has the same function as an overt ending in another paradigm.'

³ Leonard Bloomfield, Algonquian, in Linguistic Structures of Native America, 85–129 (VFPA 6; New York, 1946): Bibliography of Algonquian according to Language Groups (123–9). Bloomfield's own work is less given to statements in paradigm form than that of other Algonquianists.

⁴ There are of course other difficulties besides those concerned with the distribution of person markers; I have tried to show how Boas and Deloria (op.cit.) and also Robert H. Lowie (in his sketch, The Crow Language) approach these other problems in Siouan (review in Lang. 20.264–70 [1944]).

⁵ There is difficulty in correlating form and person marked by form in Cherokee; and phonemicization, which is relatively simple in Siouan, is not simple in Cherokee. See Ernest Bender and Zellig S. Harris, The Phonemes of North Carolina Cherokee, IJAL 12.14–21 (1946).

⁶ Thus, in Seneca to find morpheme boundaries of person markers requires extensive lexical materials.

⁷ Floyd G. Lounsbury has devised an ingenious chart for separate Oneida bases: he plots as many as 64 person markers on his chart for each base.

⁸ C. M. Barbeau, Classification of Iroquoian Radicals with Subjective Pronominal Prefixes (Memoir, Canada Department of Mines, Geological Survey, No. 56; Anthropological Series, No. 7; 1915). Franz Boas, Notes on the Iroquois Language, Putnam Anniversary Volume 427–60 (New York, 1909).

⁹ Oto, a Siouan language; our knowledge of Oto is derived from Whitman, op.cit.; from the Smithsonian Institution, Bureau of American Ethnology manuscript 1371 entitled Oto Vocabulary of Max [Prince Maximilian] of Wied (corrected by J. Owen Dorsey); and from impressions gained in checking this manuscript during a rainy week near Perkins, Oklahoma, when we stayed at the Oto home of A. M. Springer until the roads became passable again. Our casual field inquiry, incidental to other work in Oklahoma, was to satisfy a curiosity concerning the diversity of person-marking affixes in Ms. 1371.

¹⁰ Our corpus of Oto materials consists of that part of Ms. 1371 (see fn. 9) which we were able to check in the field; additional forms cited by Whitman are not given in this paper.

¹¹ Compare the treatment of Tübatulabal, in which many morphophonemic symbols are used to represent morpheme alternants: Morris Swadesh and C. F. Voegelin, A Problem in Phonological Alternation, Lang. 15.1–10 (1939).

¹² Both main stress and compound stress are phonemic, and a third type of subphonemic stress is syllabically spaced in reference to the main stress (see Whitman). Ms. 1371 is generally blank here, and our field record tends to confuse the three types of stress for base X and -X; the notation adopted for this paper omits stress indication for X and -X, but shows the main stress on affixes preceding -X; at times our informants would humorously exaggerate such stress preceding -X which marks first person in certain alternants of affix 1.

¹³ The notation adopted for this paper is in general agreement with Whitman's statement for the phonemes of Ioway–Oto (there are dialect differences which may be phonemic, as the status of r, d, l, and the vowels o, u). We write a cluster of two consonants, stop + glottal stop, where Whitman sets up a single phonemic series, glottalized stop.

¹⁴ Whitman gives, in his §14, affixes corresponding to our 4, 6, and 7 in combination with a dozen ‘active verbs’ and half as many ‘passive [stative] verbs’ which may also appear without the affixes in question.