Amelia Louise Susman Schultz, Sam Blowsnake, and Ho-Chunk Auto-Literacy1
In June 103 years ago, Amelia was born in New York City into a German Jewish family, sharing the Brooklyn home of tante, a great aunt who was a seamstress. At age six, she started grade school at PS 44, skipped 2 grades, and graduated as valedictorian of James Marshall High School at 14. Joseph Greenberg shared her subway ride, sometimes conversing in Latin with a school mate. Since she was just under five feet tall, she was ineligible to work at Macys, the usual graduation option for women until marriage. Instead, she enrolled in Brooklyn College, 1931-1935, gaining an AB degree in Psychology with Solomon Asch while earning $20/month through National Youth Administration (NYA) of the CCC. Seeking a graduate degree, she was warned by classmate Irving Goldman that only Columbia anthropology under Franz Boas willingly accepted women into its graduate program. Most other disciplines were strictly men only. Accepted at Columbia University, 1935-1939, she wrote two dissertations. The first was a study of acculturation at multitribal Round Valley in northern California, relying on documents in the San Francisco archive more than interviews with these intertribals whose traumas included genocide, girl slavery, and land loss to the ranching family of their BIA agent. Her work was part of a comparative team study in eight tribes, eventually published as Acculturation in Seven American Indian Tribes (Marian Smith = Puyallup; Jack Harris = White Knife Shoshoni; Marvin Opler = Southern Ute; Henry Elkin = Northern Arapaho; Natalie Joffe = Fox; Irving Goldman = Carrier; and William Whitman = San Ildefonso). These seven thereby earned their PhDs since publication was a final degree requirement.2 Since Amelia also devoted a paragraph to a bogus land claims scheme, collecting coins from natives, growing suspicion of a law suite led Ralph Linton, who had replaced Boas, to put pressure on Ruth Benedict to ask Amelia to withdraw her chapter, which she did in favor of her classmates. Why no one thought to excise the paragraph remains a mystery, except Amelia’s great loyalty to Boas was probably involved. Boas had always been Amelia’s advisor of record, so, from retirement, he directed her second PhD project, a study of Ho-chunk ~ Winnebago grammar to improve on a 1937 PhD, published by King’s Crown Press of Columbia University in 1945, by William Lipkind,3 advised by Paul Radin and working with Marguerite Hittle at the Dutch Reform Mission in Nebraska. Informally advised by Radin and working closely with linguist George Herzog, Amelia’s speaker was none other than Big Winnebago, alias Crashing Thunder, Sam Blowsnake, the subject of a very early native autobiography. The Blowsnakes were leaders in the Ho-Chunk community, moving between the Nebraska reservation and ancestral lands in Wisconsin. Preparing to work with him at his home in Green Bay, she bought down clothing but quickly returned her winter outfits when she learned that he was dancing and singing with his wife
1 # Ho-Chunk is now the preferred term among tribal and degreed members in Wisconsin and Nebraska, while Winnebago is what their Algonkian neighbors call them. Amelia and Sam Blowsnake wrote of the Winnebago Syllabary; it is more properly herein called Ho-Chunk Syllabary. 2 Marian was the last Boas student while he was working full time, Whitman also worked with Oto and famously died when his furnace blew up while he was teaching at Harvard. His wife, to whom he was unfaithful, edited his Oto for publication at Columbia. 3 Lipkind (1904 - 1974) then worked in central Brazil with the Carajá and Javahé under the auspices of the Social Science Research Council, eventually becoming author of children’s books as Will Lipkind.
i Evening Star and daughter Whirling Eagle on the Boardwalk at Atlantic City. Relying on his more reliable fee of 35¢/hour, she moved the family to Brooklyn and gathered her data into notebooks now at the American Philosophical Society in Philadelphia and onto aluminum disks now at Indiana University and Wisconsin Historical in Madison. With her first paycheck from the Women’s Army Corps she had copies of The Aspectual System of Winnebago offset printed and distributed in 1943, which by then counted as publication to qualify for her final PhD. Confident of her linguistics, in 1941, Boas, with American Council of Learned Societies funds, sent Amelia to work with William Beynon, a fluent Tsimshian speaker, high school graduate, and hereditary Wolf chief at Prince Rupert, BC. Enamored of the working partnership of the Curies in Paris, their marriage was briefly considered. Instead, Amelia put down roots in Seattle, just as WW II began, serving in Women’s Army Corps, 1943-1946 (after the inductor “gave” her a ½ inch to qualify for the legal 5 feet). At first, she was editing bilingual material for troops (phrase books, dictionaries, language courses) at the Army linguistic office at 165 Broadway along with Morris Swadesh, Mary Haas (Thai desk), Stanley Newman (Persian desk), and other American linguists under the direction of Haxie Smith. Later she became a psychiatric social worker for neuropsychiatric discharges at Mason General Hospital, Brentwood, Long Island, NY. Mustered out, she returned to Seattle, where, hoping for an academic career, she renewed Tsimshian linguistics with Mrs Louise Mertz, a bus ride away in West Seattle. When no academic position materialized, Amelia found a job, 1942-1943, in social work, state department of Public Assistance, at Raymond and Bremerton, Washington. She met and married Elias Schultz, nicknamed Dutch, through Melville Jacobs, a fellow Boasian linguist and folklorist long at UW, and his wife Bess, a psychiatric social worker. To gain necessary credentials, Amelia enrolled in Social Work at University of Washington, earning an MSW, 1946-1947, with a thesis on Indien unmarried mothers, abstracted in Social Work Review. A labor organizer, Dutch used his GI Bill to study woodcarving in Switzerland guilds, while Amelia offered classes in English. Unpleasant interactions with Swiss locals, especially landlords, led them to move to London, 1950-1952, where Dutch studied at Guild Hall while Amelia took anthropology and linguistics classes at London University through a 1951 grant from the Foreign Service Institute, State Department, during stringent post war conditions. Back in Seattle, Amelia worked as Child Welfare Worker, State Public Assistance, Seattle, 1952-1953; Case Worker, Children’s Home Society, 1953-1958; Case Worker, Jewish Family & Child Service, 1959-1960; Social Worker, Research Associate, Clinical Research, Center and Department of Medicine, University of Washington, 1960-1965; Social Worker, Research Instructor, Clinical Research Center and Department of Medicine, University of Washington, 1965-1973; and Social Worker, Assistant Professor, Clinical Research Center and Department of Medicine, University of Washington, 1973-1977, when she retired. She also served on committees for human subjects review and medical school admissions. She has been a member of Linguistic Society of America, American Anthropological Association, National Association of Social Workers, and the short lived Seattle Anthropological Society. Throughout her careers, she relied on three of the four fields of anthropology: linguistics, physical, and cultural, but not (yet) archaeology, as shown by her own publications:
Word Play in Winnebago American Anthropologist 1941 Accent in Winnebago, PhD Thesis Columbia University 1939, published 1943 Brienzer Deutsch, Studies in Linguistics 9 (23) 1951
ii The Impact of Genetic Disorders. Social Work 11: 29-34. The Round Valley Indians of California, PhD 1938, Department of Anthropology, University of California, Contributions of the Archaeological Research Facility 31 1976. Boas on Phonemics and Dissertations. International Journal of American Linguistics 43 (10> 56-57) 1977 January. Review of Autobiographies of Three Pomo Women by Elizabeth Colson. Journal of California Anthropology: 144-145 1977. Swiss Swear Words; Epithets in the Alps Maladicta 3 (2): 261-274, 19xx. Amelia also coauthored with Arne Motulsky, the famous geneticist at UW whose archive is also at APS:
Motulsky, AG, Schultz, A, and Priest, J Warner’s Syndrome: Chromosomes, genes, and the aging process. Lancet I: 160-161 1962 Epstein, CJ, Martin, GM, Schultz, AL, and Motulsky, AG Warner’s Syndrome, A Review Of Its Symptomatology, Natural History, Pathologic Features, Genetics, And Its Relationship To The Natural Aging Process. Medicine 45: 177-221 1966. Schultz, AL, and Motulsky, AG Medical Genetics And Adoption. Child Welfare 50:4-17 1971. Schultz, AL, Pardee, GP, and Ensinck, JW. Are research subjects really informed? Western Journal of Medicine 123: 76-80 . Lipe, Hillary, Bird, Thomas, and Schultz, AL Risk Factors for Suicide in HD. American Journal of Medical Genetics 48: 231-233 1993. Since 1967, she has been actively concerned with the Huntington’s Disease Society of America as a national trustee for several years in 1970s, NW Chapter Board to 1999, Area Contact (currently), and coordinator of support groups for those diagnosed by DNA. She has long been involved with the families of Woody and Arlo Guthrie, famous musicians who carry this gene. She also works with Citizens For Improvement Of Nursing Homes And Long Term Care, and Caroline Kline Galland Home, where her younger sister, hindered by a botched delivery during WW I, lived after she moved out from New York with their parents. Amelia maintains her health through yoga, tai chi, and diet, along with community services. Her professional involvements are sustained by proofreading academic manuscripts and a wide range of family and friends of all ages, interests, and educations, including a lively book group. She was honored by a special session when Applied Anthropology met in Seattle (sfaa2011-s33).
APS holds her Ho-Chunk linguistics corpus:
Amelia Susman ACLS Committee on Native American Languages 11 notebooks Winnebago 1938-39 Li-Rite Stenographic Notes No 101 unbound top spiral # 1-7 #1 phonetics May 18 1938 – Aug 2 1938 inside cover index by date 49-61 Autobio bilingual
iii #2 May 27-30, June 17, Aug 2 Winnebago Texts Sam Blowsnake p102 twins both sides payments inside cover #3 Aug 30 – Oct 10 1938 78 medicine dance a few notes inside cover #4 Oct 7 – Oct 25 (36) p81 songs inside cover > payments $ #5 Oct 25 – Nov 2 to Dec 2 1938 732 E21 St Brooklyn payments $ inside #6 (Nov 23) Dec 5 1938 – Jan 4 1939 dates inside cover #7 Dec 9 1938 – Jan 13 p29 songs 96 word order syntax #8 Notebook Hi-Grand line Turns quickly stands upright lies flat Jan 13 – Jan 17 songs 8, 9, 34, 109, 73 #9 The Spiral Gregg Pub Co No HG-627 Jan 17 songs p 10, 134 medicine dance #10 The Spiral Gregg Pub Co No HG-627 [con’t] 1-7 medicine D 76 songs 84 Curtis 100 education recorded [Natalie Curtis songs ] #11 The Spiral Gregg Pub Co No HG-627 syllabary education 3-31 The Winnebago Syllabary 36pp. ms.
Ho-Chunk ‘Winnebago’ Syllabary
Though unpublished, Amelia’s paper remains the touchstone for studying the workings of the Ho-Chunk syllabary. More to the point, the syllabary text “Child Teaching” has been repeated as a key example by both Willard Walker and Kathleen Danker. It first appeared as the sample text provided by Sam Blowsnake at the end of Amelia’s dissertation. At 103 (Miller 2018) she recalls while working for months in Brooklyn on her grammar, she watched Sam mouth out the words before he decided how to spell them, providing linguistic examples for study as well as song texts for an aluminum disk of him singing these songs. He wrote routinely in the syllabary, including the hundreds of pages of his autobiography done for Paul Radin, who published the English translation. Yet scholars have made wrong assumptions of where and how Sam provided his examples, beginning with Willard Walker. In an endnote, Walker indicates that a year end 1973 letter from Frances Perry, a Ho-Chunk woman who learned to write the syllabary from Blackdeer in 1915-16, identified the source of these texts was Sam and his wife (Walker 1974: 403, 414).
The letters and sounds appear in the order in which they were written for Susman by a Winnebago informant [speaker], almost certainly Sam Blowsnake, himself.
Sam Blowsnake, the celebrated author of the original Winnebago manuscript which, in English translation, was to become Paul Radin's 1920 The Autobiography of a Winnebago Indian, wrote a number of syllabic texts in 1939, when he was 68 years old, old enough to have learned the syllabary at its inception. His texts were subsequently analyzed by Amelia L. Susman (n.d.), whose analysis incorporated the results of a single interview with Blowsnake [!] and included the sample text reproduced below. It was sent to Franz Boas on October 8, 1940, under the title The Winnebago Syllabary
iv and survives in the possession of the American Philosophical Society (Walker 1974: 403).
In his analysis of the Ho-Chunk system, Walker concluded Siouian Ho-Chunk took Algonkian
Fox orthography that was, and still remains, essentially phonemic and converted it into a morphophonemic system which represented 36 phones with 18 letters and a number of spelling conventions based on an ordered set of 16 context restricted rules (Walker 1974: 397, 411).
Winnebago writing system can best be described in term of an ordered set of context restricted rules which generate surface strings of graphic symbols from underlying strings of morphophonemic representations (Walker 1974: 413).
Most recently, making their linkage clear, Walker (1996: 173) noted
The spelling conventions described above were used by Sam Blowsnake, whose version of the syllabary was well known to Paul Radin and Amelia Susman.
Kathleen Dankers compares three examples of syllabary scripts over most of a century, to trace spelling improvements, including the marking of nasal vowels. She begins with that provided to Alice Fletcher in 1890 by the actual (sadly unnamed) Nebraska Ho-Chunk who, on a visit to the Fox in Iowa in return for their stay as guests in Nebraska the year before, started adapting the Fox system to Ho-Chunk. Next is “a syllabary text composed in 1938 by Sam Blowsnake in Wisconsin [!] that was included in an unpublished manuscript by Amelia Susman in 1939,” and finally a late 1970s text by Felix White, Sr., who taught Ho-Chunk at Little Priest Community College at Winnebago, Nebraska. Her bibliography indicates Danker was aware of both Amelia’s dissertation and the 1939 draft, but otherwise does not connect the two.
Lastly, something needs to be said about the intuitive importance of syllabaries for natives. They emerged in the 1800s among Algonkians themselves in the Great Lakes, using French script letters, and in the Far North, using at first English letters handmade by Methodist missionary James Evans (1801-1846), and now used for Cree and Inuit. Among Athapaskans, Chipewyan uses a ‘syllabarium’ by Anglican Rev. William West Kirkby (Walker 1996: 174, 176). Of course the most famous and ubiquitous example, now including a digital font called Digohweli, is the syllabary of Sequoia ~ George Guess. Ellen Cushman – raised in a Catholic - Southern Baptist Cherokee home, rhetoric professor at Michigan State, and enrolled Cherokee citizen – explores the intuitive popularity of the Sequoya syllabary, which she calls Sequoyan:
Within the span of a few years, Cherokees were able to read and write in Sequoyan, without print or mass education (Cushman 2011: xi).
The system was taught in manuscript version and developed into shorthand by Cherokees … Sequoyah apparently intentionally developed this writing system to be separate and separable from English and the Roman alphabet ... As an artist himself,
v trained in observation of form, he found a way to organize eighty-six distinct shapes into the seven forms ... (Cushman 2011: 214).
Sequoyan has been a flexibly adaptive tool that codifies meaningful units of language in script, print, and digital media…. Whenever the syllabary is present, the four aspects of peoplehood can potentially be present as well: language, sacred history, religion, and place (Cushman 2011: 187).
As interest grows in issues of native literacy, making Amelia’s manuscript more widely available will serve many purposes. Socially, the Ho-Chunk syllabary provided the means of exchanging letters among Wisconsin and Nebraska communities. It is an example of human ingenuity comparable to that of Sequoia himself, with the added complication that two unrelated native languages were involved. Algonkian Fox differs from Siouan Ho-Chunk in significant ways, so over a century modifications were made in the syllabary to conform to actual Ho-Chunk sounds and usages. By making it available digitally, it will reach a much wider, and much younger, audience of readers both foreign and native.
Works Cited
Ellen Cushman The Cherokee Syllabary ~ Writing the People’s Perseverance. UOP 2011
Kathleen Danker Ba-be-bi-bo-ra: Refinement of the Ho-Chunk syllabary in the nineteenth and twentieth centuries. Advances in the study of Siouan languages and linguistics. Catherine Rudin & Bryan Gordon, eds. Berlin: Language Science Press. 2016: 83-103.
Jay Miller, ed. Acculturating Amelia. Amazon. 2018
Willard Walker 1974 The Winnebago Syllabary and the Generative Model. Anthropological Linguistics 16 (8): 393-414. 1996 Native Writing Systems Handbook of North American Indians. Languages # 11: 158- 184.
Editing
Working with Amelia, we recovered her carbon copy of her paper, with annotations by her linguistics advisor, George Herzog. We integrated these into the text along with other varieties of brackets. Thus { } are Herzog comments, < > are written in additions, and [ ] track original page #s as well as provide exclamations for counterfactual assumptions. Amelia proof read and commented on the finished version. Because this is a historic document, Winnebago has been retained in lieu of Ho-Chunk, the ethnonym preferred in Wisconsin, while Winnebago is still used in Nebraska. Winnebago is their name among Algonkian speakers, meaning . In past usage, a sonant is voiced, while a surd is voiceless. The major change throughout has been making a series of the footnotes, with #s continuous instead of restarting on each page as in the original.
vi The Winnebago Syllabary Amelia Susman 1940 Page Introduction 1 Background 1 Algonkian Syllabary 3 Early Winnebago Syllabary 4 Modern Winnebago Syllabary 1. Letters and Symbols 5 2. Vowel repetition in writing as related to historical problem 8 3 Relation of the syllabary to the phonemic structure of Winnebago 10 a. Consonants 13 b. Intervocalic consonant clusters and the syllabication problem 18 c. Vowels 21 4. Use of the syllabic writing in making difficult phonemic identifications 27 5. Syllabic writing and the analysis of compounds 29 6. Inefficiencies 32 7. Miscellaneous items 34 Conclusion 34 Text 36
THE WINNEBAGO SYLLABARY
The syllabic writing4 – which the Winnebago Indians use at the present time5 – has an internal organization which partly conforms to, and partly ignores, the phonemic structure of the spoken language.6 Certain inconsistencies, such as duplication of symbols for a single phoneme, ambiguity of symbols, and variant transcriptions of a single word, are also of interest.7 [2]
4 1. The Winnebago call their "alphabet" "Babebibora" (or Gagegiora, etc.) pronouncing a set of syllables that represent one consonant with four of the five vowels {vowel phonemes not letters}, and adding the suffix -ra, "the". 5 2. The material on which this paper is based was obtained {in 1939} in the course of studying the accentual system of Winnebago, with Sam Blow Snake, a sixty-eight year old man, brother of one of Dr. Radin's chief ethnological informants. He transcribed the Babebibora, and wrote the texts of about thirty-five Medicine Dance songs of one sentence each. In addition, he transcribed phonograph records of a portion of a connected text, and several word lists. When he wrote the "alphabet" he pronounced each syllable aloud. I discussed the symbols with him on one occasion but there was no opportunity to check with him any of the points that came out of an analysis of his transcriptions. He did not offer any comments on the system, but acknowledged that one of the letters (the k) was superfluous. 6 3. Dr. Morris Swadesh first suggested that the syllabary should be studied for the light that it might throw on phonemics. 7 4. Often, in writing. Blow Snake pronounced the syllable, either aloud or to himself, and
1 It seemed that light might be thrown on all these features, by comparing this syllabary with the Algonkian syllabary8 from which it is known to have been adapted, and with the earliest recorded Winnebago version.9 The published material on the last two is inadequate for extensive comparisons.10 The short text in script11 which Jones gives, serves to bring into relief the unusual and apparently original features of the present Winnebago form, and to show, on the other hand, how closely the script has been preserved. Miss Fletcher's discussion shows that symbols were being added by the Winnebago from the very beginning, and permits inferences concerning the adaptation of certain letters. However, the transliteration she gives is not clear, nor very helpful.
Background
seemed to be deciding on the spur of the moment how it should be written. Another indication of this approach lies in the inconsistency of syllabication, which will be described below. Such a problem could hardly have troubled him, had he learned to write words as such. Blow Snake is intelligent, and well versed in tribal matters, and his way of writing is probably representative. There may, of course, be errors due to a personal idiosyncrasy, but this does not seem likely. He was careful to correct errors when he read back his material in preparation for phonographic recording. In addition, there are many features in which he is perfectly consistent. Such inconsistencies as occur, therefore, may be {tentatively} assumed to be those of the system of writing. {Not too cogent, esp., in view of next sentence. If you really have a chance to check with someone else soon, why not delay the article?}. 8 1. "An Algonkian {Algonquian} Syllabary" by William Jones, is an article in the Boas Anniversary Volume, 1906. It describes the syllabary used by the Sauk and Fox tribe. He says that there was no trace of old figurative writing in it at the time of its “recent introduction” to these Indians. {He says that it had no trace of development from the old representation which was still used by the older people. “The jump from the old to the new must have been sudden.”} There may have been some change in the Sauk and Fox syllabary after the Winnebago made their adaptation of it in 1884-5, but the maximum time interval between Miss Fletcher’s description of Winnebago usage (1890) and Jones' paper is not great, and the latter can be taken to describe in general the form in which the Winnebago found the syllabary. 9 2. The article by Alice Fletcher in the Journal of American Folklore, 1890, Volume III, {“Winnebago Syllabary”} contains a brief sketch of the written letters, with transliteration. (Jones gives the script.) He describes how one Winnebago learned the syllabary on a visit, and then other Winnebago literally "put their heads together" in adapting it. 10 3. A comparison of the Sauk and Fox syllabary with the phonemics of that language would be interesting, but has not, so far as I know, been made. 11 4. The script is an old-fashioned (perhaps slightly modified) English type. I can not recognize all of the letters. By inference from Miss Fletcher’s transliteration, the letter for č as in church) is represented by capital D. The letter for kw may be a form of g. There is a modified form of Algonkian writing which employs dots in different positions, instead of letters, for the vowels, and omits the dot in representing the vowel a . Still another alphabet modifies consonants and vowels, and uses a peculiar set of symbols, partly geometric. Both of these, however, seem to have a one-to-one correspondence with the regular script. The only feature that has any bearing on the Winnebago is the omission of a sign for a in the dot system for vowels. [3]
2
The Algonkian Syllabary
The following are the phonemes represented in the Sauk and Fox syllabary, together with the script forms and English words to illustrate the sounds.12 [Monotype Corsive]
Sounds: p ℓ pen t ð ten s s see š d she č ðt church y y you w w we m m me n n no k k kill kw 8 quit a13 u what, all, hut (The letter u represents this sound in medial and final position, but in initial position it is represented by hu.14 e e let, late, alley i i it, see o o fellow, full, rose, loon
There is no mark of accent. Words, and word-sentences, are separated from each other by a period, or a dash. The end of a sentence is marked by + or x. In the appended text with which Jones illustrates his points, each syllable appears as a vowel, or consonant plus vowel, (run together usually). There are no consonant clusters. Diphthongs, it is stated, are represented by the first vowel.
Early Winnebago Syllabary
12 1. The series at the margin are phonetic symbols, which may facilitate comparison of this with the Winnebago script. The others are from Jones. 13 2. How correctly these letters, particularly the vowels, represent the phonemes of Fox, I do not know. Vowels in Fox are short and long, but length is not indicated in this writing. 14 3. This may indicate an aspirate attack, but it is not used with other vowels in initial position. The handwritten text shows five vowel letters: a, e, i, o, u. In the absence of transliteration I do not know how these are to be interpreted. Probably a and u stand for the same thing. [4]
3 Script equivalents K d w ҳ t m n £ g π Th y b Transliterationx g j w x td15 m n R gw S Th y p Phonemes+ (?) g ĵ16 w γ t m n r s č17 y b Script equivalents a e i o d A Transliteration ä e ï o sh h Phonemes (?) a e i o š h x As given by Fletcher + This is not a complete list of Winnebago phonemes. It simply gives what I assume to be the sounds represented. Cf. p. 3, 5 [herein]
There are a number of obscure points here. Why capital letters are used in the transliteration, and what the diacritics mean there is no way of knowing. The order is that in which Miss Fletcher’s informant gave the letters, and she considered it significant. w, m, n, and y correspond to the Algonkian letters. The stops correspond, but are transliterated as (sonant) g, (surd) p, and (perhaps intermediate) td. gw as transliterated, seems to be derived from Jones’s kw, but what it could represent in Winnebago I do not know.18 π in the script series probably corresponds to Jones’s s.19 The spirants are transliterated as surds, x, s, š.20 h in this series is a regular consonant, whatever it may have meant in Jones’s description. The affricative č is probably what is meant by Th, though whether the symbol dt given by Jones and used in modern Winnebago writing, is derived from script Th I am not sure. Interpretation of vowels is also confusing. Whether the [5] diacritic " refers to vowel quality, such as openness, or nasalizatian, and to length, perhaps is matter of conjecture. Another usage which I can not attempt to explain, is the presence of a small π in transliterated syllables, after a vowel, or after a vowel and before k.21 The script equivalents of transliteration for vowel plus r are simply vowel. Syllables transliterated as ending in πk, and for that matter in k after a vowel, are given script equivalents ending in m.22 The vowels in themselves seen to represent the same sounds as the Algonkian letters. Unlike the Algonkian syllabary, there seems to be
15 0. There is only one t sound in spoken Winnebago, unaspirated, fortis, with little voicing. 16 1. As in judge, probably č as in church. 17 1. As in judge, probably č as in church. 18 2. This combination of sounds occurs rarely in Winnebago. Where it does, as in a word like higana, the u is part of a rising diphthong. 19 3. In modern Winnebago script, the phoneme r is written with a small open s, an s not completely closed. Perhaps the pre-emption of a letter like π for the s phoneme accounts for the use of the letter L to represent the tongue-flapped r phoneme. This capital L is interesting also in view of the fact that Dakota l corresponds to Winnebago r, and the Winnebago substitute an 1 sound for the English r in pray. 20 4. x is the palatal spirant, an addition to Jones’s series. š is as in she. Presumably some Winnebago, including Miss Fletcher's informant, spoke English. 21 1. These syllables are listed with no translations. 22 2. This feature ties up with one in modern writing. (See p13 #4)
4 Even taking into account the brevity of Miss Fletcher’s paper, it is significant that she makes no mention at all of the characteristic features of the modern way of writing.
Modern Winnebago Syllabary
Letters and Symbols Script ℓ d k ĸ ð ðt n m £ ҳ y t π w a e ei o u Sounds23 b ž g k h ĵ n m r γ y t z w a e i o u
I have preserved the order24 in which the syllabary was written for me, although it did not seem to me at the time that there was a memorized or preferred sequence. As a matter of fact, w was omitted, and put in after the other consonants, only after I remarked on its absence. The syllabary was given, not in letters, but in symbols,25 the syllable which each represented being pronounced aloud. Each consonant was pronounced with each of the vowels,26 just as the name, Babebibora, implies,
Phonemes ba be bi b'a b'e 'i Symbols b be bi ba bee29 i The consonant letters for the most part represent sonants. A final surd that 23 3. Sonants are represented for b, g, ĵ, z, ž, γ since that is clearly what the letters in script mean. k is a surd, t, "intermediate". These will be described in more detail below. For phonemes see p13. [6] 24 0. “Symbol” will hereafter be used to mean the representation of a syllable, consisting of one or more letters, and spaced off from what precedes and follows. 25 1. The Fox taught the syllabary in this order: first all the vowels, then each consonant with each and all of the vowels. 26 2. There are a number of facts that make it highly probable that u was a late introduction into the Winnebago syllabary. It does not appear in Miss Fletcher's lists. It is not part of the name of the syllabary, and stands last in order. Neither of these last two points is conclusive in itself, of course. The use of u is most important in this respect. See #3 > 22. 27 3. This is not the way this letter is read in all, or even most, of its occurrences in actual use. The relation of the letters o and u to the phonemes o, u, and y is fairly complicated and will be discussed below. It suggests strongly that u as a letter is rather recent, and that all three phonemes were earlier represented by the letter o. 28 4. Every syllable in Winnebago starts with a consonant; the glottal stop, as syllable-initial, or in any cluster, is a consonant. The syllabary has no letter to stand for it, but is still perfectly unambiguous in representing it. 29 5. The whole question of vowel repetition will be taken up immediately. Here, its use in indicating the presence of glottal stop is patent. 5 "sonorants">. Pre-vocalic surds and glottalized consonants are treated as clusters.30 However, the surd k sound may he written as a cluster, or with the single letter k. The general technique of writing is as follows: Each symbol, representing a syllable, may consist of one or more letters, written together as in longhand. Theoretically, as many as eight letters might fall into a single symbol, but the longest combination observed so far has been six. The pauses which mark the ends of phrases31 are indicated by periods (.). In the transcribed word lists, a period follows each word. The most striking feature of {this} syllabic writing is the alternation of consonant and vowel letters. All symbols, except those which contain the phonemes ' [¿] or a , which have indirect representation, show such alternation. So far does this tendency to avoid consonant clusters in writing go, that {almost} all32 consonant clusters are written with a vowel letter inserted, and syllables ending in consonants have a vowel added at the end.33 {Does this have a *com? with writing sibilant instead of final surd? Cf above} [8] Examples of syllable writing:34 1. ų£ to do u u 2. ba'ną he distrusts b n ℓ n 3. če•k new ĵehege tteteke 4. gų•'s to create gozo koπo 5. hi•šĵą eye hi žĵ ði dtt 6. kše• kše'k crooked geže gežege kede kedeke 7. haruwą'k eight h ru wk ð £u wk 8. t'ą•p to jump tab , tam35 taℓ tam 9. γo•pkį to make γo bigihi ĸo £iĸiti oneself holy 10. hakaraxe' he buries h k π γehe ð K £ ĸete** (his own) 2. Vowel Repetition in Writing as related to Historical Problems In 1883, J. Owen Dorsey pointed out36 that certain consonant clusters, permitted in related 6 Siouan languages, were replaced in Winnebago by alternating consonants and vowels in "dissyllables". "A triliteral monosyllable in *Chiwere37 (and often the corresponding ones in Dakota and Dhegiha)38 is changed into a quadriliteral dissyllable in Hotcangara (Winnebago), when the first letter of the monosyllable is a mute, a palatal spirant, or a spirant sibilant, and the second consonant is [9] a labial or dental mute, or a dental spirant. The first consonant of the Hotcangara dissyllable is always a surd; the second is, as in the corresponding Chiwere word, a labial or dental mute, or else a dental spirant; and each consonant (in Hotcangara) must be followed by the same vowel sound. In no case, as far as examples have been gained, can any mute stand next to one of the same order; e.g., a labial can not precede a labial." [10 < 9] The historical problem involved is that of determining whether this trait of Winnebago is a development peculiar to itself, or whether it was once a condition common to Siouan languages, from which some of the consonant clusters of the other languages have arisen by vowel loss. Dr. Boas recognizes the latter possibility, but does not consider it probable.39 It may be well to examine a little more closely the salient facts in Winnebago. First there are the "dissyllables" of which Dorsey speaks. Strictly speaking, these are not two syllables, but a cvcv structure which behaves in a unique way. The first consonant is a surd (stop, affricative, or spirant) and the second is r, n, or w,40 with the limitation, mentioned by Dorsey, that no two consonants with the same position of articulation appear together in the combination. The vowels are identical, and quite short compared to normal vowels in Winnebago. In regard to stress (and in Winnebago there are length-like stress patterns of definable form), cvcv behaves like a single long syllable. In syllabic writing, however, such combinations are written in two separate symbols. (See p 8 # 10, Kuπa i ; p 23 # 2, 6, 13.) Secondly, there is the behavior of non-initial consonant clusters,41 the second member of which is r, m, n, w, or y. These clusters bear a resemblance to cvcv forms, since they correspond in so far as the [10] second consonant is concerned. However, the first consonant may be any sonant (stop, affricative, or spirant) and there are no limits to the 36 3. J. Owen Dorsey, The Comparative Phonology of Four Siouan Languages, Annual Report of the Smithsonian Institution for 1883, p. 923. This paragraph is quoted in the section on Siouan languages in the Handbook of American Indian Languages, by Franz Boas. Bulletin 40, Bureau of American Ethnology, 1911, p.888. 37 4. 7 syllabary treats this situation in two different ways. Sometimes the break is made between the two consonants, in which case the vowel written after the first consonant is the same as the vowel that precedes it. In other cases, both consonants are put into the same syllable, and the vowel written between them is the same as the one that follows the second consonant. In neither case, is there any difference from the syllabary treatment of medial clusters of consonants which stand right next to each other, with no suspicion of vowel glide. (See **) {Not very cogent argument} This characteristic vowel repetition in writing in a system which was developed recently by the Winnebago themselves, seems to add support to Dr. Boas’s position that vowel repetition and the avoidance of certain consonant clusters in Winnebago reflect a strong tendency which need not, without strong evidence pointing in that direction, be assumed to have ever been the case in related languages. [11 < 10] 3. Relation of the Syllabary to the Phonemic Structure of Winnebago The largest unit of utterance rapid speech, (between stem and suffix they might, in the slow speech of a linguistically unsophisticated informant, remain as clusters) complicates the problem and makes it rather questionable to speak of these as intra-syllabic. 8 ways43 in writing, so that the only possible conclusion to be drawn is that native speakers are no better able to decide than we, and the distinction itself becomes academic. There are two types of clusters that contain consonant combinations which are phonetically different when they open a syllable from the way they sound when the syllable division is between them. These are glottalized consonants (stops and affricative), contrasting with syllable-final consonant plus initial glottal atop, and aspirated stops, contrasting with syllable-final consonant followed by initial h. In such cases, the same distinction is made in writing. In many other clusters where there is no phonetic reason for deciding one way or the other, factors of etymology seem to weigh heavily. Details and examples of syllabication will be discussed further under a Consonants {> center} The following is a list of the sounds which may occur in initial pre-vocalic, and in intervocalic positions. The large groupings are phonemic. They bring together the consonants which follow certain patterns of behavior, particularly as regards their role in consonant clusters. Group I, however, is a phonetic list. The subdivision into two parallel series is recognized by the syllabary and in phonemic analysis, but not in the same way by each,45 so it seems best to present the phonetic facts, arranging the sounds according to position of articulation, type of release, and voicing, in order to facilitate description and, a little later, application of the alternative interpretations. I II III A B Stops b p w m 43 1. This is true in two 9 r n t g k y h ‘ ¿46 affricative ĵ č spirants z s ž š γ x Consonants in Group II need no special definition. They represent the same sounds as the English alphabet {as they do in English} except that r is a single flap of the tip of the tongue against the alveolar ridge, like the New York pronunciation of the second consonant in [14] "butter". They are all unit phonemes.47 None of them appear in phrase-final position, or as first consonant of a cluster. They never enter at all into phrase-initial clusters. In the middle of a phrase, they appear as second member of a cluster, after one of the sounds shown in I A.48 Group III consonants do not occur in phrase-final forms 46 2. That the glottal stop is a full-fledged consonant, and not just a glottal attack or an initial vowel is indicated by its distribution, and by the fact that it is never lost in composition (reduplication, affixation, compounding). Syllabic writing recognizes this, despite the fact that it has no symbol for glottal stop as such. (See p.6 #4) 47 For data on possible relationships between n & π See below p 26. 48 0. See p. 14. 49 1. There is no disagreement between the two interpretations of Group I on this score. In the clusters with Group II, the basic Group I phoneme is actualized as sonant, with a phonetic glide vowel. These clusters arise in suffixation, and juxtaposition of words, and the Group II phoneme changes the preceding stem-final or word-final surd to sonant, phonetically. 50 2. See p. 12, # l. 51 3 < 4. Glottalized consonants are recognized as clusters of Group I phoneme plus glottal stop in both systems. Syllable final p in symbol written l or m after nasal voiced phoneme ą see p 21. See also p. ** 10 Phrase-initial clusters t g ž s š ' ¿ p p' t t' k ks kš k' č čg s st sg sĵ s' š šg šĵ š' x xg xĵ x' These clusters52 have been written phonetically, to agree with the A-B split among Group I consonants, and for the same reason that that division was made. It is apparent that the spirants are always surd. The stops and affricative are surd (slightly aspirated except before spirants) when they occur as first member, in second position they are at least partially voiced, and definitely unaspirated. Glottalization follows spirants, but is practically simultaneous with stops and affricatives.53 One other type of consonant cluster54 must be mentioned here, since it has direct bearing on the interpretation of Group I consonants. These are the internal clusters of sonants (A) with Group II consonants, with glide vowel between.55 [16] It has been stated that the only consonants that may appear at the end of a phrase are surds. (B) The lack of contrast here, like the phonetic simplicity of initial clusters, suggests that perhaps there is only one basic phoneme series in Group I. This is the way the problem is handled in syllabic writing: There is one letter for each of the consonants in Group II, and for t and h in Group III. The glottal stop is represented by the letter for the vowel that follows it.56 Each of the consonants in Group A is represented by a letter, and there is an additional letter standing for surd k. These letters stand for sonants in initial and intervocalic position, and for Group I consonants of either sub-group in clusters. In addition, final consonants (phonetically surd) are written with these same letters.57 52 0. The clusters may occur at the beginning of a phrase. Within a phrase, they may, and in a few cases can be demonstrated to be, syllable-initial, but this is the only way they can be accurately designated as a class (See p 17) There will be, whatever interpretation we take, a series of basic stops plus h which will fall into this group, phonetically identical with Group A or B. These have been omitted until the merits of the different interpretations can be discussed. 53 1. Glottalized consonants are recognized in phonemic analysis too as clusters of basic phonemes with consonantal glottal stop. See p16 for facts of distribution of glottal stop. 54 2. Discussion of other, including bi-syllabic, internal clusters, will be reserved until later. 55 3. See p. 13 # 1. 56 1. See p.6. 57 2. This relegates the features of surdness (and aspiration in the case of stops and affricative) to the realm of definable positional characteristics, and identifies final surds with a basic sonant series. 11 Surds (B) in initial and intervocalic position, however, are written with the letter for the corresponding sonant, plus h, with regular vowel repetition, as in other clusters.58 This system is adequate, in the sense that there is no ambiguity, no difficulty in reading back what has been written (at least none that arises from the choice of letters to represent consonants). However, it does use two different techniques for writing what are almost identical phonetically, that is surds, in initial and inter-vocalic position (written as clusters of sonant and h), and in final position (written with letters for sonants), and, as is implied in the foregoing, uses one set of letters (sonant) for two very different phonetic phenomena, initial and intervocalic sonants, and final surds. [17] Phonemic analysis, along lines suggested by Dr. George Trager, takes as basic the surd series, which can then be said to occur, with very little, if any, phonetics difference, in initial pre- vocalic, intervocalic, and phrase-final positions.59 In initial clusters of Group I consonants spirants are always surd; stops and affricative are surd, with little aspiration, in first position, they are unaspirated, slightly voiced, in second position. A consonant of Group I is sonant, and has a glide vowel, in before clusters (only phrase-internal) before Group II consonants; similarly, a Group I consonant becomes sonant before h, which is not itself otherwise represented phonetically. This is the interpretation of Group I sonants. The advantage of this interpretation lies in the fact that it avoids the weakness of the syllabary system,60 and is simpler in the sense that it brings the clusters with h into line with those that have Group II consonants in the same position. Further support for this approach comes from the fact (not known to Dr. Trager when he organized the phonetic data in this way) that in composition of words, and in running words together, the main tendency of elision is that an initial h changes a final surd to the corresponding sonant.61 Glottalized consonants are clusters of consonant plus glottal stop, rather than independent unit phonemes, for the following reasons: The glottal stop is an independent phoneme which occurs in initial pre-vocalic, and in intervocalic positions. It is never phrase-final, or syllable final, however. Glottalized consonants are never in final position. Neither are any clusters, nor do they enter into combinations with other consonants in initial clusters. The phonetic difference between a glottalized consonant between vowels, Intervocalic Consonant Clusters and the Syllabication Problem62 58 3. Dr. Morris Swadesh suggested, before the phonemic analysis had been worked out, that this treatment in the syllabary warranted careful consideration of this interpretation, that is, that surds are phonemically clusters of sorants with h. This has been done, but it seems better to accept instead the alternative interpretation presented below, for the reasons to be given. 59 1. In those cases where we can be sure of syllable division, syllable final consonants are invariably Group I surds, with aspirated release. 60 2. See p. 15. 61 3. Efficient as this analysis appears, Dr. Trager does not think that it should be unqualifiedly accepted without a phonetic recheck of the language, so far impossible. 62 0. See p. 10. 12 The problem of syllabication 63 1. So strong is this tendency that in reduplication it sometimes overrides transparent etymological conditions, e.g. wa•zį 'to suckle (wa's breast; hį to suck) is reduplicated: wa•zįzį. 64 2. Similarly, intervocalic aspirated stop is phonetically different from stop plus syllable-initial h. There is not this criterion in the case of glottal spirants, since glottalization follows the spirant, and there is no phonetic difference between bi-syllabic and infra-syllabic clusters. 65 1. Final č plus initial t form long t•. Other combinations are geminate. 66 2. I was not at first aware of them, nor always sure that they were present. 13 {It is} initial-type clusters {which} can not be said unequivocally to be intra- or bi- syllabic when they occur in the middle of a phrase between vowels. The first member is always surd. According to phonemic interpretation, the second, if it is a spirant, is a unit phoneme (because surd) wherever the syllable division. If a stop67 or affricative (voiced) the second member is a cluster of basic phoneme plus h if syllable division immediately precedes it, but a unit phoneme if the preceding consonant is in the same syllable with it. If there is any phonetic distinction which might place the division, I have been unable to recognize it. It becomes an academic distinction. There is no need, phonemically, to tell exactly where one syllable ends and another begins. The rule that the second member of an initial cluster, if stop or affricative, is unaspirated and at least partly voiced, can be extended to cover these initial-type internal clusters, and left at that. {Arguments seem to clash.} Syllabic writing, however, demands a decision. Blow Snake lets etymology guide his choice in the sense that an initial-type cluster within a stem (and the vast majority of stems are monosyllabic and even those that are not almost always contain only initial-type clusters) is never divided, but the same cluster, when it arises in juxtaposition of words, is always split. When the cluster is the result of affixation, his writing is again inconsistent.68 Not even etymology, though it is seldom unclear, is followed. The syllables are separated before the cluster, or in the middle of it, or the offending consonant is written at the end of one syllable- symbol, and at the beginning of the next. (See p 32 # 56) The same difficulty arises in the case of intervocalic clusters of Group 1 sonant, with glide vowel, and Group II consonant. Phonemically it is doubtful whether syllabication can, or need, be made. These clusters occur in normal speech, as the [21] result of suffixation (See p 32 #4) or word juxtaposition. The latter kind disappears as soon as the tempo of speech decreases at all. The former does too, in Blow Snake's speech, though it is not a foregone conclusion that this would be true of a less sophisticated informant who was less aware of etymologies. At any rate, a full pause between the two members leaves syllable-final surd, and syllable-initial consonant, but then we are dealing, by definition, with two phrases. Full pause could not come before the cluster, since such clusters are not admitted in phrase-initial position. Again, it can be left with the conclusion that in rapid speech it is of no significance where the limits of syllables fall in relation to such consonant clusters. Syllabic writing, again forced to put the break somewhere, is inconsistent, and occasionally redundant, in its transcriptions. The suffices -ra "the" and -re "this, " "these" frequently seem to draw the final consonant of a stem into the syllable with them. (**See Text p36 l.1. čegregi geregi – nagre or gere l.8. hataginajra nj ) Vowels Vowel phonemes, and the letters that represent them in syllabic writing are as follows:69 67 0. Stop is sonant if second in initial cluster. Phonetically, hi•šĝą. Phonemically, hi•ščhá. 68 1. It must be remembered that such inconsistency or redundancy, while it no doubt hampers efficiency and speed, is in no way phonetically ambiguous. "Sonant" letters are used in either case. The difference lies in the choice of vowel letters. If division precedes the cluster, the first consonant will be followed by the vowel that follows the cluster – otherwise by the one that precedes it. 69 1. It must be remembered that a and ą phonemes are represented only when they follow 14 Phonemes a ą e i į o u u Letters a e i o o u Any of these vowel phonemes may follow any consonant or cluster of consonants, of Group I. After some of the other consonants, the distributions of some of them are limited: m and n may not be followed by a, i, u ; r and t may not be followed by the nasalized ą, į, ų . [22] There might be some doubt as to whether the nasalization of vowels after m and n may not be automatic, rather than part of independent nasalized vowel phonemes. However, since nasalized vowels do occur after most other consonants, where they must represent independent phonemes, they can most simply be recognized as such here. An additional factor is that e and o may follow m and n, apparently without nasalization. The syllabary ignores nasalization as such, merging the phonemes a and ą in the letter a, etc. The optional use of the letter m to stand for the phoneme p after ą has been mentioned as an indirect recognition of nasalization, arising out of the combination of the vowel release with the position of articulation of the p. However, there is only one stem in the sample of syllabic transcription where such a vowel-consonant sequence occurs, and the use of m is not required. There is, as far as I know, no attempt to represent an alveolar nasal for post-vocalic č , and certainly no letter for a velar nasal to stand for k70 after a nasal vowel. The treatment of the phonemes u and l. šųk žohogo dotoko "dog" 2 pųnųx boho noγo ℓoto noĸo crunchy 3 hišĵasu hi žĵ zhu did t πtu eye 4 re•xhučų reγe hu ĵohu £eke ðu ttok bucket with legs glottal stop [¿]. 70 1. It seems, in Miss Fletcher's transliterations, that m is used to represent k at the end of some syllable-symbols, but what exactly the use of this letter (m) is, is most unclear. 71 2. See p 6, # 2, 3.) 72 1. Whether this combination is used in other syllables that have vowel repetition I do not know. If there is a tie-up between this device, (the use of letter o for the first, or both vowels) and the choice of letter to represent phonemes u and ų after h and ' [¿], it* is difficult to say. These two consonants seem to be followed by either letter. In one word, the interconsonantal vowel was omitted, and u written for the other. 15 5 gų•zra gu zr gozo r ku r£ koπo £ the creator 6 kuruxe ko ro γehe ĸo£oĸete to run after (one's own) 7 mą•γožu m γo žo m ĸo do cultivate plants. 8 honįk'u•ną ho ni gou n he gave you 9 ų•xįnįra o γihi ni π the charcoal 10 hu•hu to start coming 11 hu•nįre ho ni πe Don’t come! Group II consonants followed by phonemes u or ų are represented in the following ways: nų is written no73, wų is written wu74, rų is written πo or πu, the former more frequently than the latter. Eight haruwa'k h πu wg you ran after suruxe (your own) zoho πo πeha to look at horuyu'8 ho πo voĵo Group III consonant h and ' [¿] may take either letter.75 [24] Whether this over lapping in syllabic writing of occurrences of the phonemes o, u, and ų has any bearing on phonemic interpretation is uncertain. The letter u seems to be a late addition to the alphabet, and one whose use is not altogether standardized. The three phonemes concerned are undoubtedly independent of each other and after most consonants, contrast with each other. On a morphophonemic level, however, there seems to be an obscure relationship among them which fuller material on the syllabary might clarify. For instance, the second vowel o in a cvcv stem76 becomes u before certain suffixes which start with (h) and some vowels, e.g. hoišoro "right hand side", hoišorua'nąga " " " Contraction of the two prefixes, hį•me , and ho in gives hų•- , as if the nasalization in the former, attaching to the vowel timbre in the latter, throws it into the ų phoneme. The diphthongs which have been identified in Winnebago are these: Falling77 ei eį ui ųį oi(oį) ai ąį ui ųe oe Rising ie įe io įo eo uo ia įą ią78 ea eą ua ųą oa oą 73 2. mų is a phonemic combination in which has not so far been observed to occur. 74 3. Only two examples of this occur in the sample. 75 4. The phonemic combination tu has not been found in writing, tų is not phonemically admissible. 76 0. See. p. 8. 77 1. Vowels in Winnebago have different degrees of strength relative to each other, in the decreasing order, a, o, u, e, i. If the first in a combination is the stronger, the result is a falling diphthong, that is, the second vowel is less prominent and shorter, and vice versa. 78 2. These diphthongs are an exception to the rule that except where e and o are one of the members, both vowels in a diphthong must be oral, or both nasal. Neither, however, has 16 They are interpreted as diphthongs (unisyllabic vowel clusters) on the basis of the basis of the stress patterns of the words in which they occur, which either permit or demand that they be so understood.79 The syllabary treats diphthongs in a way that bears out the analysis that has been outlined in a general way. Whether specific interpretations would also be carried out as the criterion of stress pattern requires in phonemic analysis, there are not enough examples to ascertain. Obviously, the fact that the written letter for any vowel except when it precedes a written consonant (and a always) stands for consonantal glottal stop 1. tai'sak I am cold ty zhg ty πtĸ 2. xetei'žą a big one γehe tey ž Kite tey d 3. hui'č foot, broad end hoyĵi86 Aoytti 4. hųįšą busy huy žh Auy dt 5. ho•ka'rakųįra•nąga87 ?? ho k π gohoyi π n g A K£ kotoyi £ n k been fully attested. 79 3. In other cases, the same two vowels may appear, but under stress conditions which demand that they be recognized as the nuclei of two separate syllables, and the glide between them a consonantal w or y. 80 0. The semi-vowel written is usually the one phonetically present. See however, the next paragraph. 81 1. In both examples at hand, the i itself has been omitted, which may be the rule. This causes ambiguity with the next type of diphthong mentioned, e.g. ea, written ey. 82 2. In at least one instance, in writing of eo , the o was omitted (See ex 7), and the written combination ey might, ambiguously, stand for eo or ea diphthongs, as well as for ei , mentioned in note # 1. 83 3. As in simple syllables, where w and y stand for consonants, these letters, written without following vowel, stand for wa, wą, ya, yą. 84 4. If, however, o, u or u precede i , the combination is written oy or uy, but not owi, uwi, See ex. 3, 4). 85 5. See p 22. 86 1. The vowel repeated after the final consonant in writing monosyllabic hui'č and gia's (see no. 8) is i , though this is second in one, first in the other. No other monosyllables containing diphthongs have been found so far. 17 6. hit'eo'kirač different languages hi teeyo ki πĵ Ai teeyo Ki £tt 7. hą•'heo'kahi every night h hey k hi A ðey K ti 8. gia's to flee giysi88 kiyπi 9. hįgua'ną right now hi gow n Ai Kow n 10. nųą'wąk he ran now wk now wĸ 11. hožua'kšąną he (moving about) ho žowk žh n Ao dowℓ dt n put it in 12. wioire the west wiyo πe89 wiyo £e [27] 4. Use of the Syllabary Writing in Making Difficult Phonemic Identifications The syllabic writing offers no help in making phonemic identifications or distinctions. There are a few points where it is very difficult to determine from the phonetic data how certain sounds are to be treated phonemically, and it seemed that the informant's writing might show that he feels certain distinctions which my ears are not sensitive enough to catch. Not only has this not turned out to be the case, but where phonemic analysis is difficult, syllabic writing shows confusion. Not only is it unsystematic, but it contradicts as often as it supports what seem to be the probable etymologies involved. For instance, the relationship between long stressed e and the diphthong ei90 needs to be clarified. Phonetically there is hardly any difference. The phonemic interpretation hinges on whether the vowel combination ei brought together in composition, is reduced to e (in contrast to all other similar combinations which, if reduced, lose the first member) or remain ei and can not be reduced. Both possibilities exist, although the first does not follow the ordinary behavior of eliding vowels. If this point were consistently treated in syllabic writing, it would help also in grammatical analysis, but several examples can be given to show that it is not: 1. hagore(i)žą sometime h goπey ža 2. here(i)reže they were he πe πe že 3. 'e•(i)gi here e gi 4. 'e•(i)ĵa there e ĵ [28] 5. he(i)pši to sneeze hey biži 6. ze'ze'terižainigenąki a big mountain anywhere ž ni ge The first example is of a compound, containing hižą, which as I remember, the informant could not himself analyze, but which is written so as to indicate the presence of i. The second 87 2. See appended text, line l 2. 88 1. The vowel repeated after the final consonant in writing monosyllabic hui'č and gia's (see no. 8) is i , though this is second in one, first in the other. No other monosyllables containing diphthongs have been found so far. 89 3. The third member of a triphthong is omitted in this, the only one transcribed. a i o "and uia have been found in the language, but very rarely. 90 1. This diphthong, if it does occur, does so only in cases of elision, not within simple words or stems, which means that the sound can be heard only in normal speech, and will be changed, or lost, if the informant is made to speak slowly. This makes it very difficult to tell if it does differ from e•, which occurs frequently in stems. 18 word is a combination of the stem here' = to be with the suffix -(h)ire = they, but is written without the i. The adverbs here (3) and there (4) are by inference analyzed to contain hiĵa and higi, but are written with only e. The fifth word is an unanalyzable stem in which I had no reason to suspect that a diphthong was present. The writing of ey suggests a line of analysis which was not, however, useful in this case. In the last example, which should be considered a phrase in all probability, there is, in rapid speech, an į after the first ą. There is nothing in the etymology to account for it, and it is not written in the transcription by the informant. Difficulty arises also in connection with the relationship between the two consonant phonemes n and r. Although the former can be followed only by nasalized vowels, the latter only by oral, they can both precede e and o and are thus demonstrated to be distinct phonemes. After a nasalized vowel, r changes its type of release, and sounds like a slowly released d, preceded by nasalization (n), and is altogether not easily distinguishable in rapid speech from ordinary n. When it happens, as it not infrequently does, that a nasalized vowel precedes, and an e follows, the questionable sound might be n or r.91 [29] Examples: 1. ne•wį•re•ną I am ne wine n (ne I | wį he to me; here' to be ?) 2. wą•kši'k-'regiži if they lived w gižigi i ne gi ži (wą•kšik- į to live, | (h)ire they) 3. waųįreže they did it w oy πe že ('u• to do | -(h)ire they) 4. hokipį•regiži they suffered ho ki bihi πe gi ži 5. -kĵąne ~ -kĵąre will be gĵ ne (future suffix) In the first two words n is written. The etymology of the first is not certain, but that of the second is. A perfectly parallel word is written with π in the third example, and another in the fourth. The future suffix (5) is unclear etymologically as well as phonetically. Blow Snake equates this suffix with a longer suffix, -kĵąnąhe , which refers, he says, to a more distant future than the one to which this elided form does. He writes it consistently with n. On the other hand, I often hear a distinct r (-kĵąre) which suggests that the derivation of the suffix might be the future suffix -kĵe , the vowel of which shifts to ą before the suffix -re = to become. If the writing were consistent on other points, it could be accepted as substantiating the etymology which Blow Snake claims for this element. As it is, the question must be left open. 5. Syllabic Writing and the Analysis of Compounds In Winnebago, close compounds must be distinguished from loose compounds, and both from mere habitual juxtaposition of words in ordinary or idiomatic expressions. These distinctions are not always easy to make, since there are several criteria, not all of which apply to all, or even many, types of the cases. Stress features and the rather elusive semantic factors (the latter always difficult to apply) are two of the criteria. The third, which is often crucial, is the 91 1. Even where there was no indication of doubt in my phonemic transcriptions, I had reason later to wonder whether my knowledge of the etymology had not influenced my writing of r wherever the suffix -(h)ire = they was involved, even though the preceding sound might have been a nazalized vowel, which lent nasalization to the i. 19 question of whether elision or contraction has caused the loss or [30] shift of phonemes92 characteristic of both compounding and juxtaposition. Presumably Contraction is a more permanent phenomenon, whereas elision is present only in normal rapid speech and would disappear in slow speech. Syllabic writing of it representing slow speech,93 should show the difference clearly. Here again there is a considerable degree of inconsistency. Extreme forms are written as they can, without much trouble, be understood by other than phonemic criteria. Unanalyzable words are, of course, written as units, though by inference they may seem Examples: l. čio'kisa'gonąžį'ra stands-in-the-Center-of-the-House ceremonial term for fire. written > ĵhiyo ki žh go n ži π (či house | hokisak center | honąžį' to stand in | -ra the) 2. wą•kši'go'į life (wą•kši'k person, human, man, ho'į' to live ?) written > w giži go i 3. wi•ra' ro•čąĵeipa noon (wi•ra' the sun | ro•čąĵe standing straight | hipa toward) written > wi π πo ĵh ĵey bh 4. hagorei'žą some time in the future (haga- time turn | hore' time of going ? | hižą one) written > h go πey ž [31] 5. nį•kĵą'grižą a child, any child (nį•kĵą'k child | -ra the | hižą one) written > ni gĵ gπ hi ž 6. wą•gią a man (wą•'k man | hižą one, a) written > w giy ~ wg giy 7. hit'eo'kirač different languages (hit'e' to speak | hokirač different) written > hi teeyo ki πĵ 8. hą'heo'kahi every night (hą•he night | hokahi every) written > h hey k hĵ 9. 'oziokšųkšų wrist ('ozi forearm | ho- in | kšu pivoting, jointed) written > o zi ho gožo The first four examples seem, by semantic criteria, to be compounds, although the first was the easiest for the informant [speaker] to analyze. The others seem to be composed of separate 92 < 1. A final surd becomes sonant before initial (eliding) h ; final Ø vowel drops or merges into diphthong with the same following the h. > 93 2. The difficulty with simply using Blow Snake's slow speech to decide the point arose from the fact that he very quickly learned (if he did not know) how to identify elements which less sophisticated speakers of Winnebago would probably not have recognized, or at least not have presented spontaneously as he seemed to do speaking slowly. 20 except the last are written as merged. Particularly, the use of hižą,94 the article, might seem to be a suffix, rather than a modifying word, since it has in all cases that have been found been written with the preceding word, and the phonetic shifts recorded. On the other hand, the more idiomatic usage, in which the suffix -ra plus hižą give the meaning of "whatever, any", and is even in slow speech pronounced -rižą, is transcribed in analytic form. (5) In contrast to what seem to be separate words written by Blow Snake, the last example is a compound, which he found quite difficult to analyze, but which nonetheless has the ho syllable written out. One factor in writing might be the case with which a particular eliding syllable that starts with h can be recognized by the informant as a locative prefix, since in a number of the examples given this could not be done.95 (2, 3, 4, 5, 7, 8). In the first example, however, the second ho- is perfectly clear, but is not separated out in transcription. [32] Syllabic writing throws little, if any, light on the problem of compounding, and can certainly not be relied on where other criteria fail to distinguish two degrees of compounding, and syntactic juxtaposition. 6. Inefficiencies There are three types of inefficiency in the Winnebago syllabary, all of which have been mentioned in various connections. They are: 1. duplication of symbols for a single syllable Duplication occurs in the use of k as well as the cluster g plus h, to stand for pre-vocalic surd k sound. Writing the letter m for final p96 after nasalized vowel might be a roundabout, but still serviceable device, except for the fact that it is optional, and the regular way of writing p is in use for the same thing. The use of letters o and u for phonemes ų and u involves duplication in that one would suffice, unless the letters distinguished between oral and nasal vowels, which they do not. The letter o is ambiguous because it also stands for the distinct vowel phoneme o. A different type of duplication occurs when troublesome syllabication of an internal cluster is handled by repeating the first consonant. (See examples) 1. ki•ši'reže they ordered (Text. line 6) ki žihi πe že 2. ho•ka'rakųįra'nągą they advise and ho k π gohoy π n g 3. t'ą•p to jump tam 4. t'ą•bre' Jump down! tam πe ~ tam beπe 5. t'ą•pge' because he jumped down tab bege 6. tą•pgni he jumped down already ktam gi ni ~ tab bigini97 [33] 94 1. hižą is also the numeral "one", and has not been analyzed. 95 2. Locatives are ha- , hi-, ho-, meaning, respectively on, with, by means of, in. 96 1. No other cases of final p after nasalized vowel besides this one of the stem t'ą•p have so far been observed, so it is impossible to say how far-reaching this, or related phenomena, might prove to be. 97 2. This merging of two syllables in one is probably accidental, though there is some similarity to the cvcv setup and similar words must be watched. 21 7. wą'gią' a man w giy ~ wg giy Two minor ambiguities have been pointed out: The use of the latter o for three different phonemes, o, u, and ų , and the writing of diphthongs starting with e so that ey can stand for ei, ea, and, if the o is omitted from written eyo, for that as well. The lack of sign for nasalization means that there is only one letter for a and ą , etc. The choice of the sonant series as the basic one for consonants of Group I means that as far as strict representation of the phonetic facts is concerned, both duplication and ambiguity exist,98 but not in a form that can create any difficulty in reading back what has been written. The really troublesome ambiguity of symbols arises precisely from the most pervasive99 feature of Winnebago syllabic writing, the repetition of vowels in ccv and cvc syllables.100 If the vowels are the same, and the consonants are the same in two such syllables (provided the first consonant is a sonant), it is impossible to tell them apart. This sometimes caused noticeable delay in Blow Snake's rending back of his own transcriptions, and very often his first interpretation was the wrong one. Of course, context, and the fact that there are not a great many pairs of stems with the same unit phonemes in these two combinations, lessens the confusion.101 The lack of consistency in syllabication is inefficient in several ways. Actual duplication of a consonant is treated above.102 What is important is that wherever a cluster of consonants is brought together by affixation, the writer must decide which one of three possible types of syllabication to use.103 [35] 7. Miscellaneous Items There are a few points which come out of an analysis of the sample of transcriptions in syllabic writing which may be of significance although they can not now be tied up with anything else. 1. The syllable pa was found written bhh with a small superscript h < in one example >. 2. The word howe•gawire' was written with weye for the second syllable. This might have been a slip of the pen for weyi. On the other hand, perhaps y once functioned in non-initial clusters in the way that group II consonants still do. 3. The repetition of a vowel after a consonant that checks a diphthong has been mentioned, particularly the fact that in two such monosyllabic words, the repeated vowel was i, although this sound (i) was the first member of one of the diphthongs, the second of the other.104 4. The word ru•s has been found, written πusi 98 0. See p. 6. 99 1. The only exceptions to this rule found so far are sg and (ĵh) in the appended text (p 36). 100 2. See p. 7, including #3. 101 3. It will be apparent that pre-vocalic surds will be written as clusters with h and ' [¿] as such immediately recognizable, whether a consonant closes the syllable or not, since an h or a ' cannot end a syllable, nor are there final clusters. For the latter reason, ccvc can cause no confusion, since cvcc is impossible. 102 4. See p. 32, including ex. 4, 5, 6. 103 5 See p. 20, 21, 23 ex. 5; p.31 ex. 6. 104 1. See p 26, ex 3, 8. 22 All of these forms would have to be checked for correctness of the written form, by Blow Snake's standards, and similar ones elicited, before anything further could be done with them.105 Conclusion Winnebago writing is a fairly consistent method of representing syllables by symbols composed of letters derived from English script. It was adapted from the Sauk and Fox syllabary in 1884, and developed further in certain original directions. The classification of sounds implicit in it parallels phonemic analysis to a large extent. It is not of any use, however, in offering criteria for phonemic or grammatical analyses where internal criteria fail. There are several kinds of ambiguity and duplication of symbols, particularly the complete ignoring of vowel nasalization, and the repetition of a vowel so that [35] ccv and cvc look alike under certain conditions. Length and stress, highly significant features in spoken Winnebago, are completely left out. The repetition of vowels in written Winnebago is related to certain facts concerning consonant clusters or the avoidance of consonant clusters, in spoken Winnebago and in related languages, and has bearing on an historical problem in Siouan. Of larger units than syllables, only phrases are recognized, and enclosed in periods. There is no marker for the end of a word or a sentence as such, although syllabication in the majority of cases does follow etymology, within a word, and, except between closely related words, (possibly compounds) does keep words distinct. The connected test which follows demonstrates the use of the syllabary. The typed version is phonetic. Spaces separate words, which can be identified by a combination of grammatical and phonemic (stress) criteria, but not by any features of unit phonemes. The end of a phrase, usually with sustained pitch, is marked with a comma to distinguish it from the end of a sentence, or complete predication. (A sentence, like a word, can not in Winnebago be distinguished by any features of phonemes or even intonation, from a phrase, but must be regarded as a grammatical unit.) The end of a sentence is marked by a period. Each sentence in the text, being declarative, has falling pitch at the end. [36] 105 2. The limitations of the sample account for the fact that most words were written only once. 23 Test I Appednix O0 p 216 | 36 nį•kĵą'k ho•ka'πakų' ni kitt d ĸ L ĸo child teaching e•gi če•gregi wą•kšik , wą•kši gną•gre ĵa•sge wą•kšik į•regiži e ki dte kefe ki• w kidiki w kidiki n kbe tt πke kidiki i ne ki di . and in-the-beginning Indians Indians-these how (men)-live-they when nį•kĵą•grižą gixeteregiži ho•karakųįranągą e ĵas•ge e• čo•weĵa ni kett ks did kiHete te Lkidi ðo £Jkotoy J n k • tt πke ttto we tt child-the-a-big-become-when advise-they-and this now in-the-future wą•kšik- į•rekĵąregiži , wą•kšiagoįra pį•ų•kĵe že•sge karagiguzire w kidiki i nek kidi. W kidi ko iℓ ℓiti o ttete de πke £J ki ko πi ℓe (men)-live-they-would-when life-the good-make-will thus their-own-to teach-they kĵąregi če•kĵįra , waųįreže , čiokisagonąžįra , himągiragireže ktt ne ki ttte kittiℓ w oy le de. ðthiyo ki πt ko n diℓ ði m ki ℓede could-then at-first do-they-it-is said Lodge-Center-in Stands (fire) point-at-they it is said ų•xįnįra , ru•s, že•sge ki•šireže. hą•taginąč hakikaraži•reže kiti niℓ ℓoro de πke kiditi•lede ð ki £i£e di ℓ de. Charcoal-the take thus order-they-it is-said to-be-blessed struggle-they-it-is said -to-earn (to fast) če•kĵįra, xųnųnįk ha•ginąĵ•iregi;106 wi•ra ro•čąĵeipa , že•ĵaįxĵį ĸoto no niki ð t ki ntt tti ℓe ki. wi ℓ ℓottt tteyℓt de tty itti107 First-the little-ones to fast-start-they-when-Sun-the-Straight end-very Stand-Toward (noon) hį•bire, že•ĵąįxĵį waruĵireže. Ųąrairegi , hiĵaira hiaraireže di ℓi Le. De tty ĸitti w lo tti le de. uw ly le ki ðitty l dy ly le de. Reach-they end-very eat-they-it-is said do-they-when more go on-they-it-is-said hahi hą•psereč , hą•taginąĵ•ra hokipįregiži hahahe hi•re hąhiokahi, ð ði tℓ πete ℓette ð t kintt ℓ to ki ℓiti Je ki di t ðe Ji t tey k ti reaching day-long fasting-the suffer-they-when to night do-they night-every hahi hagakirahąšge nąįreže , nųbąhąhi•nąįre , e•ĵaxĵį ð ði. ð k kiJ ð deke nyJede. No ℓ t ðinyℓe e tt ĸti reaching once sleep-they-it-is-said twice-slept-they about mąkerepąnąįžą e•ĵa nįge•reže ĵo•bąhą hakewehąšge nąįres’aže. m He ℓe ℓt ny d ett nike8ede tto ℓð ð ke we ð deke nye πa de. ten-once there somewhere four-times six-times-perhaps slept-they-accustomed-to-it-is-said 106 2 Phonemically ĵ can be broken down into čĵ. 107 1 Syllables omitted by Blow Snake, accidentally. 24 October 19th 1940 Dear Amelia, Enclosed I am returning your much improved version of the Winnebago paper. I think it is much clearer in organization now, and where I still have qualms, I made pencil notations on the right hand margin, and occasionally at the bottom of the page. Specifically: 1) If you have a fair chance of checking on the syllabary with someone else, soon, I think it were better to wait with publication for that check. If not, it would be better not to make remark on p. 1, since a number of things do hinge on such a check and question arises in the reader's mind why you publish the material as is, if checking is around the corner. 2) fn. 1, Jones. I think he expresses himself very badly, but all he means is that this type of writing has nothing to do with what the Indians had before. What they had was no writing proper, but picture representations. Have you checked his exact wording? If he says "writing" for the old stuff I would put it in quotes to indicate the term is his, and indirectly, not proper. 3) I always thought the word was syllabification and syllabication; better check in the dictionary. Marked many places but not all. 4) As indicated on the proper page, I do not quite see why the vowel repetition has much to do with the history of the Winnebago cvcv type and its relation to Sioux clusters etc. Just because the writing is so recent, it would seem it would have little to do with the phenomena whose historical connections go way back. As a result it hardly proves anything either way, its only bearing being on purely phonemic problems which have no historical perspective behind them. Next week is still very cluttered up; I must go down town a few times, possibly on Tuesday afternoon too; except that I have a long appointment beginning with 4 PM. If there is anything special or urgent, would you care to tell me in a line? I haven't heard from the Modern Museum as yet and will let you know whenever; it might take quite a while yet. Hope to see you soon, and am sorry my time has been so uncertain these days, and to some extent still is. I hope you'll publish this and/or the Winnebago pun paper soon. Best wishes, George Herzog 25