Quick viewing(Text Mode)

Plains Cree: a Grammatical Study Author(S): H

Plains Cree: a Grammatical Study Author(S): H

Plains : A Grammatical Study Author(s): H. Christoph Wolfart Reviewed work(s): Source: Transactions of the American Philosophical Society, New Series, Vol. 63, No. 5 (1973), pp. 1-90 Published by: American Philosophical Society Stable URL: http://www.jstor.org/stable/1006246 . Accessed: 10/07/2012 14:55

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

.

American Philosophical Society is collaborating with JSTOR to digitize, preserve and extend access to Transactions of the American Philosophical Society.

http://www.jstor.org TRANSACTIONS

OF THE

AMERICANPHILOSOPHICAL SOCIETY

HELD AT PHILADELPHIA

FOR PROMOTING USEFUL KNOWLEDGE

NEW SERIES-VOLUME 63, PART 5 1973

PLAINS CREE: A GRAMMATICALSTUDY

H. CHRISTOPH WOLFART Associate Professor of , University of

THE AMERICAN PHILOSOPHICAL SOCIETY INDEPENDENCE SQUARE PHILADELPHIA

November, 1973 This book has been published with the help of a grant from the Humanities Research Council of , using funds provided by the Canada Council.

Copyright ? 1973 by The AmericanPhilosophical Society Library of CongressCatalog Card Number 73-79574 International Standard Book Number 0-87169-635-5 US ISSN 0065-9746 PREFACE

Although the goal of linguistic description is well original manuscript; in many cases he pointed out known, the complexity of is such that a the more general implications of specific problems. complete grammar does not exist of any language, Warren . Cowgill was kind enough to send me a long ancient or modern. Since the linguistic study of a and detailed list of comments. Thanks are also due particular language is a cumulative process, however, to Charles F. Hockett who first introduced me to the it is useful to provide progress reports at certain "marvellous complexity" (Bloomfield) of the Algon- points: neglected areas will then become obvious and quian ; he has read and extensively criti- research plans can be integrated into a coherent cized all my efforts in Algonquian linguistics, including scheme. the present one, and where I have disregarded his The present study outlines an integrated statement advice, I alone am to blame. The greatest debt, of the morphological structure of the Plains of course, is to the Cree speakers who with consider- of Cree, an Algonquian language. While its primary able patience and generosity taught me some under- aim is to provide a framework for further investi- standing of their language. gation, this work may also serve as an introduction It is a pleasure to acknowledge the field support to the study of Cree texts. during 1967-1968 of the National Science Foundation The exposition of the grammatical categories of (GS-1535), the Wenner-Gren Foundation for Anthro- Cree is based on a detailed morphological and semantic pological Research, and the Phillips Fund of the analysis of the inflectional paradigms. The sketch American Philosophical Society; and from 1969 to of formation, which is necessarily less compre- 1971, of the Research Board of the University of hensive, is intended to highlight some of the more Manitoba. The writing of the original dissertation productive and characteristic patterns of derivation was supported by Yale University which, together and composition. with the Studienstiftung des deutschen Volkes, also This study is a revised version of my 1969 Yale largely financed my graduate studies. University dissertation, "An Outline of Plains Cree From Oblate Fathers to sound technicians, acknowl- " whose basic orientation remains un- edgments are due to more people than can be named. changed. It is primarily based on data collected in This preface would be incomplete, however, without in 1967-1968; more recent and continuing special thanks to my wife Juliane whose part in it far informant work, mainly in Manitoba, requires only exceeds that of the perfect Cree typist. one or two additions which are specifically noted in the text. are also drawn from the nikdwiy e-wi-petamawak. Examples pub- H.C.W. lished texts of . University of Manitoba I am grateful to Floyd G. Lounsbury for his , Canada criticisms and suggestions during the writing of the December 1971

3

PLAINS CREE: A GRAMMATICAL STUDY

H. CHRISTOPH WOLFART

CONTENTS

PAGE PAGE 1. Introduction ...... 6 4. ...... 33 1.1. Aim and scope ...... 6 4.1. Pronominal paradigm I ...... 33 1.2. D ialects ...... 7 4.11. awa, ana, naha 'this, that, that yonder'. .... 33 4.12. awlna 'who' ...... 1.21. Cree and Montagnais-Naskapi...... 7 ...... 34 1.22. Cree ...... 8 4.13. tdni'which'...... 34 1.23. Plains Cree...... 10 4.2. Pronominal paradigm II ...... 34 1.3. Data and informants...... 11 4.21. tdniwa, ewakwd 'where is he', 'there he is'.... 34 1.4. Abbreviationsand conventions...... 12 4.22. 5ya 'that no longer here' ...... 35 1.41. Grammaticalinformation ...... 12 4.3. with endings ...... 35 1.42. Text references...... 12 4.31. kikway 'what' ...... 35 4.32. kikway 'something'...... 36 2. Grammaticalcategories ...... 12 4.33. kotak 'another' ...... 37 ' 2.01. Major dimensionsof contrast ...... 13 4.34. aya ... one ...... 37 2.02. Animate and inanimate 14 ..... 4.4. Isolated paradigms ...... 37 2.1. Person...... 15 4.41. ewako 'the selfsame' ...... 37 2.11. Personal o-...... 15 prefixeski-, ni-, 4.42. awiyak 'someone' ...... 37 2.12. Personal mi- ...... 15 2.13. Personal ...... 16 4.5. Personal pronouns ...... 38 2.14. Combinationsof ...... 16 4.51. Stem s ...... 38 person categories 4.52. s ...... 38 2.2. Obviation...... 16 Paradigm 2.21. Focus and spans...... 17 5. inflection ...... 38 2.22. Inflectional limitations and change of focus 17 5.1. Basic verb types ...... 38 2.23. Marked status of obviative...... 19 5.11. Transitive animate (TA) ...... 39 2.24. "Further obviative"...... 20 5.12. Animate intransitive (AI)...... 39 2.3. Gender...... 20 5.13. Transitive inanimate (TI) ...... 39 2.31. Shift of gender ...... 21 5.14. Inanimate intransitive (II) ...... 40 2.32. Gender classes of ...... 22 5.2. Paradigm tables ...... 40 2.33. Marked status of inanimate ...... 23 5.21. Sources ...... 40 2.4. N umber ...... 23 5.22. Empty positions ...... 40 2.41. Absence of contrast...... 23 5.23. Level of representation ...... 40 2.42. Marked status of plural ...... 24 5.24. Versions of the TA independent indicative 2.5. Direction...... 24 table ...... 40 2.51. Direction and focus...... 25 5.25. Summary of verbal paradigms ...... 41 2.52. Further implications...... 26 5.3. Modal categories ...... 41 2.53. Historical survey ...... 26 5.31. Subcategories ...... 41 3. Noun inflection...... 27 5.32. Independent order ...... 44 5.33. Conjunct order...... 45 3.1. position classes ...... 28 5.34. Imperative order ...... 46 3.2. The paradigm...... 28 3.21. Theme formation...... 28 5.4. Affix position classes ...... 47 3.22. Inflection...... 29 5.41. position 1: thematic obviative sign /em / 47 3.3. The number-obviationparadigm ...... 29 ...... 3.31. Suffixes...... 29 5.42. Suffix position 2: theme signs...... 47 5.43. Suffix 3: thematic obviative 3.32. /w/-alternation ...... 30 position sign /eyi/ ...... 47 3.4. Paradigm tables...... 30 5.44. Suffix 4: mode h- and 31 position signs: ht-pret- 3.5. The /epan/ suffix...... erit, delayed imperative ...... 47 3.6. Locative ...... 31 5.45. Suffix position 5: non-third person suffixes... 47 3.61. Simple...... 31 5.46. Suffix positions 6 and 7: mode signs: p-pret- 3.62. Distributive ...... 31 erit, dubitative ...... 49 3.7. Vocative ...... 32 5.47. Suffix position 8: third-person suffixes...... 49 3.71. Singular...... 32 5.48. Suffix position 9: third-person plural and 3.72. Plural...... 32 obviative suffixes ...... 49 3.8. "Quasi-nouns"...... 32 5.49. Suffix position 10: mode signs: subjunctive, 3.81. Locative...... 32 iterative ...... 49 3.82. Vocative ...... 32 5.5. Stem shapes ...... 49 3.83. Plural...... 32 5.51. Transitive stems ...... 50 5 6 WOLFART: PLAINS CREE [TRANS. AMER. PHIL. SOC.

PAGE The inadequacy, on a variety of counts, of the 5.52. Animate intransitive stems ...... 50 present work is recognized. Thus it is not meant 5.53. Inanimate intransitive stems...... 51 as a museum of whatever but as a tool 5.6. The transitive animate 51 piece, merit, (TA) paradigm...... for reference and research which is 5.61. Morphological structure and syntactic func- bound to be tion ...... 51 modified by the results of its very application. 5.62. Direction...... 52 It is well known that linguistics remains far re- 5.63. Meaning and reference in the third person moved from the goal of providing a complete, formal subparadigm ...... 53 of a with reference to a 5.64. You-and-me forms...... 55 grammar particular language 5.65. Independent order endings ...... 56 unified theory of language. In spite of the important 5.66. Conjunct order endings...... 56 achievements of the last seventy years, there is still 5.67. Imperative order endings...... 58 no language for which such a grammar is even approxi- 5.7. The animate intransitive (AI), transitive inanimate mated. In this respect, Cree differs only by degree (TI), and inanimate intransitive (II) paradigms 58 from a like which has been under 5.71. TI theme 58 language English signs...... intensive for several 5.72. AI and TI non-third and imperative endings 58 study centuries.' On the other 5.73. AI and TI third person endings...... 59 hand, the study of "exotic" languages has differed 5.74. II endings ...... 59 from that of the more widely spoken languages in an 5.8. Marginal and suppletive paradigms...... 59 important point: in almost all cases, linguistic 5.81. The relational paradigm...... 60 has been the work of an outside observer 5.82. The ...... 61 analysis paradigm with control of the under 5.83. The TA inanimate actor paradigm...... 61 incomplete language study. 5.84. The TA indefinite actor paradigm...... 62 Such an observer, whose role as participant is limited, 5.85. The AI and TI indefinite actor paradigm.... 62 obviously cannot be expected to have the same kinds 5.86. The AI inanimate actor paradigm...... 62 of insights as a native speaker, and a pessimistic 6. Word formation...... 62 conclusion might be that "beyond a relatively super- 6.1. Derivation ...... 62 ficial informant in will fail 6.11. formation ...... 63 level, techniques general Primary and further if will on 6.12. Secondary derivation ...... 63 work, any, depend linguistically 6.13. Derived (deverbal) suffixes ...... 63 trained natives" (Postal, 1966: p. 93). However, 6.2. Roots...... 65 while the basic issue cannot be disputed, Postal's 6.21. classes ...... 65 pessimism seems exaggerated. Nor should the tactical 6.22. ...... 66 of the outside observer be overlooked 6.23. Relative roots ...... 66 advantage a team effort of and 6.3. Medials...... 66 entirely. Thus, linguist linguistic- 6.31. "Body-part" medials ...... 66 ally trained native speaker would appear ideal. 6.32. "Classificatory" medials...... 67 In practice, however, little would be gained by 6.33. Extended medials ...... 68 restricting our attention to those languages where 6.4. Finals...... 68 native are available. 6.41. Noun finals ...... 69 speaker-linguists Unfortunately, 6.42. Particle finals...... 70 the salvage aspect cannot be overlooked in the field 6.43. finals...... 70 of "exotic" languages. Even granting the inadequacy 6.44. Transitive verb finals...... 73 of many grammars, it may be well to remember the 6.5. Compounds...... 75 concern with which we treasure some scrap of ancient 6.51. Nominal compounds ...... 76 literature or some brief 6.52. Verbal 76 poorly recorded, vocabulary. compounds...... In the field of American Indian at Appendix: linguistics, any A: Morphophonology...... 79 rate, the salvage efforts of the last one hundred years B: Sample text ...... 83 can hardly be regarded as useless or irrelevant. References cited...... 89 The present study of Plains Cree is in a technical sense informal. 1. INTRODUCTION A fully formal theory of a natural language will 1.1. AIM AND SCOPE remain an ideal for a long time to come, if indeed it is achievable at all. he who to This study is an attempt to describe the structure Consequently attempts of in Plains Cree. Its immediate is construct a formal theory of part of a natural language purpose will have to content himself with a small twofold: first, in covering a relatively area, it is very part. large But even to achieve this limited a intended as a framework for further investigation to goal, fairly which reference can be made in more detailed studies. substantial knowledge of the language under study is required, so that the will not be It may also serve as an introduction to the fairly partial grammar large body of Cree texts. Two volumes of texts have 1 In view of this fact, the angry claim (however distorted) that been published by Bloomfield, and there are several it is a "mass of . . . almost completely superficial and inexplicit manuscript collections. To make full use of these linguistic descriptions which make up our linguistic literature texts for a today" (Postal, 1966: pp. 92, 93) cannot be dismissed out of hand. linguistic, literary (in wide, Jeffersonian To make matters worse, the and other many grammars convey annoying sense), purposes, presupposes both a impression, intentionally or not, that they constitute a definitive (which is in progress) and a reference grammar. statement of fact, observed and described once and for all. VOL. 63, PT. 5, 1973] INTRODUCTION 7 entirely out of tune with the remainder which is 1.21. Cree and Montagnais-Naskapi to be constructed. In eventually practice, therefore, The use of "Cree" for the entire of Cree such are constructed complex partial grammars usually only and dialects is even for which are well some Montagnais-Naskapi widespread languages known; languages, it has no means been established that there such as have not centuries of though by English, only undergone is a chain of dialects.3 but have also been studied mutually intelligible investigation by large This identification of the as numbers of the Montagnais-Naskapi native-speaker linguists during present "Cree" back at least to 1849 when McLean In the so-called "exotic" goes John century. languages, by wrote: "The Indians . . . of are a tribe of informal accounts need to be written Ungava contrast, yet the Cree nation Their lan- before more detailed studies can be undertaken designated Nascopies. and, a dialect of the Cree or exhibits a formal theories be constructed.2 guage, Cristeneau, perhaps, considerable mixture of words . . ." (cited A formal account, then, does not come into being after Michelson, 1939: p. 87). Michelson, however, in a vacuum. Not only that, however; semi-formal probably did not mean to imply mutual intelligibility descriptions may have the additional advantage of when he stated (1912: p. 247) that "excluding avoiding the straitjacket of premature formalization phonetic changes, Montagnais is practically the same or of one particular formalism. Data need to be language as Cree." Unfortunately, this statement observed from a variety of vantage points and with as seems to have been over-interpreted by later scholars. few a priori restrictions as possible (such as, for Michelson himself speaks of "very sharp boundaries" example, the familiar stricture against the mixing of (1939: p. 73) and definitely seems to imply a language levels). when he that "it cannot The framework of this is intention- boundary says (1939: p. 70) linguistic study be too that east of Hannah eclectic. The basic orientation be strongly emphasized Bay ally may perhaps Cree leaves off and described as since it is influenced (Ontario-Qu6bec line) Montag- praeter-Chomskyan 1 in some of the other which have taken nais-Naskapi begins." (See map 1.22.) by developments Recent dialect distance the Summer with the of testing by place simultaneously emergence generative- Institute of transformational to one Linguistics (Irvine Davis, personal theory. Thus, give just would also indicate a break of mutual the semantic of communication) example, aspects paradigmatic between eastern and western dialects. are intelligibility analysis emphasized. the break occurs at an The debt to the Bloomfieldian of describ- Curiously, however, entirely great way different between Nelson will be obvious. point, namely House, ing Although and Ontario. The tentative his sketch is the most known of Manitoba, Winisk, (1946) surely widely nature of the dialect lets it his on it is in the survey appear possible writings Algonquian, posthumous that at further locations in this area would Menomini that Bloomfield's testing grammar descriptive show the transition from the Manitoba dialects to is most style clearly expressed. those of northern Ontario to be much less Adherence to the Bloomfieldian model to abrupt. helps In that the scores northern make of different case, relatively high linking grammars Algonquian languages Ontario and dialects towards more More Bloom- Quebec might yet point easily comparable. importantly, a link of Cree with But until field's in to be Montagnais-Naskapi. descriptions general appear remarkably less and more detailed evidence becomes the to their ambiguous appropriate (in Hjelmslevian sense) the term "Cree" should be used in its matter-in of unanswered available, spite many ques- narrow sense. tions. Aside from relatively minor matters, we The reflex of Proto Algonquian *0 is usually taken depart from his model only where there is a special as a convenient diagnostic in determining the language reason and adequate evidence to propose a different affiliation of the Algonquian dialects of eastern and approach. central Canada. Such isoglosses gain special signifi- 1.2. DIALECTS cance where a large number of dialects are spoken small and mobile bands. While the The appellation "Cree" is commonly used in at by fairly hunting reflex of Proto *0 is n in most of the least two different senses. (1) It may refer to the Algonquian it to t in Cree-Montagnais- complex whose Algonquian languages, corresponds Cree, stretches from the Labrador coast to the e.g. atim 'dog, horse'; cf. Fox anemwa, Menomini territory etc. This set Mountains; or it refer to Cree as anE.m, Ojibwa anim, correspondence Rocky (2) may dialects of Cree from the opposed to Montagnais-Naskapi. clearly distinguishes great variety of dialects of the Ojibwa- (Saulteaux)-Ottawa- ' To avoid any misunderstanding it might be emphasized that Algonquin complex.4 "informal," as the opposite of the technical term "formal," is not intended to imply or to excuse lack of internal consistency, ele- 8 For writings on Montagnais-Naskapi see Lemoine, 1901; gance, comprehensiveness, economy, etc. These are properties Rogers, 1960; Hisey and Fiero, 1969; MacKenzie, 1971; and which any account, formal or informal, strives to attain in some McNulty, 1971. degree. But even when considered collectively they are different 4 For the application of such criteria to early missionary sources in kind from formalness, and they do not add up to it. see Hanzeli, 1969. 8 WOLFART: PLAINS CREE [TRANS. AMER. PHIL. SOC.

In its eastern manifestations this set of corre- comes /c/ before a front " (Rogers 1960: p. 94). spondences is known as "Roger Williams's sound shift" Thus, from sohk- and -isi- there is a verb stem sohcisi-; (cf. Haas, 1967a); Roger Williams was first, in 1643, cf. Plains Cree sohkisi- 'be strong.' to observe the regular substitution of n, 1, and r in the word for 'dog' in Coweset, Narragansett, and Quinni- 1.22. Cree Dialects piac. The usefulness of the diagnostic for Cree is The dialects of Cree not affected by the fact that the reflex t also occurs proper (excluding Montagnais- are to be described In in Blackfoot, Cheyenne, and the Atsina and Nawa- Naskapi) yet adequately.7 his well-known of Michelson a clear thinehena dialects of Arapaho; the distinction of Cree study 1939, gives of the data situation: "it is not . . . from these languages is no problem. Note further picture possible to the work that has been done on some that since *0 alternates with *s before *i, *, *y, only approach . . . as mor- its reflexes in other positions are available as European languages regards phonetic, or or distribution of diagnostics. phological, syntactic differences; words. A cannot even accumulate the It is thus an important piece of evidence for the single person to of close affinity of Cree and Montagnais-Naskapi that necessary materials, say nothing interpreting them" While the Summer Institute the same reflex t also occurs in Montagnais-Naskapi, (1939: p. 75). of referred to earlier covers e.g. atum 'dog' (Lemoine). Linguistics survey (1.21) a it is not a dialect in the usual Moreover, both Cree and Montagnais-Naskapi large area, study sense but is based instead on the of dialect differ from the great majority of the surrounding technique distance for languages by keeping distinct the reflexes of Proto testing (see, example, Kirk, 1970). test were used west of the Algonquian *0 and */ regardless of their eventual Moreover, only eight points Province of while some realization in the dialects (cf. 1.22); consider the Qu6bec. Thus, yielding information on mutual the examples below. interesting intelligibility, scope of this survey seems to have been too restricted Proto Algonquian *aOemwa'dog' *elenyiwa 'man' to match the diversity of Cree dialects. Cree atim iyiniw A convenient preliminary classification is provided Montagnais-Naskapi atum ilnu by the reflexes of Proto Algonquian *1.8 Fox anemwa ineniwa Howse seems to have been first, in 1844, to use this Menomini anE m eneni6w diagnostic: "The widely scattered tribes of this nation Ojibwa anim inini change the th, consecutively into y, n, l, and (vide so far stressed the common features of Cree Eliot) r, e.g. We-thi ('he'), W6-yd, W6-na, We-li, Having etc." and it be well to out (1844: p. 316).9 Montagnais-Naskapi, may point Lacombe further the locations where the at least one of the more differences. provided striking different dialects are and because of its wide Before the reflexes of Proto *i and *i, spoken, Algonquian his table deserves to be in full *k in Montagnais-Naskapi appears as c (alveo-palatal acceptance given the 'I'; Lacombe, 1874a: ); consider the second person prefix ci- (with p.xv): (Plains Cree ki-) or the cZ 'past' (Plains Cree kt; cf. 6.522).5 Cris proprement dits (Plains Cree) niya This palatalization is attested at least as early as Cris d'Athabaskaw (northern ) nira 1696 when Fabvre listed the alternant forms miki8aps, Presque tous les Cris de Bois ( mitchi8aps for 'cabane, maison' 6; cf. Plains Cree of Rupert's Land) nitha mikiwahpis. Whether the k-alternant represents Cris du Labrador (Montagnais? Moose Cree, as Hanzeli has it, or whether it reflects a sound Cree?) nila change in progress has to remain open. Mask6gons () nina Valuable evidence on relative chronology is pro- Algonquins et Sauteux [sic] (Ojibwa) nin vided by the occurrence of palatalization in cases where the conditioning environment has subsequently 7 Several fairly detailed dialect studies are now in progress at disappeared; consider the third person animate plural the University of Manitoba, the University of Toronto (under the Proto *-aki for which auspices of the Odawa Language Project), and elsewhere. For a ending Algonquian Montagnais- preliminary statement on methodology and a detailed study of a Naskapi shows -ac and Plains Cree -ak. At least particular population, see Wolfart, 1971b. in the Mistassini dialect of Montagnais-Naskapi, this 8 Note, however, that the reflexes in Cree and in Montagnais- palatalization is also reflected in synchronic morpho- Naskapi are entirely independent, however much they resemble each other. Similar have taken phonological alternation: "-final /k/ be- developments place elsewhere; cf. Michelson, 1939: p. 75. 6 Palatalization apparently does not take place before *y; con- 9 The reference to Eliot shows that Howse failed to distinguish sider Proto Algonquian *kydtdwa 'he hides it', Montagnais- fully this intra-Cree variation from Roger Williams' (and John Naskapi katau.) Eliot's) sound shift; it is noteworthy that he gives no example for 6 Fabvre, 1970: p. 152; Fabvre's 8 correspondsto our w. Cf. r. Howse's description is of the th-dialect of Rupert's Land (see also the discussionof this manuscriptin Hanzeli, 1961: p. 126 ff., below); we follow him, Lacombe, and Michelson in writing th for a section which is not included in Hanzeli, 1969. what generally seems to be a voiced . VOL. 63, PT. 5, 1973] INTRODUCTION 9

MAP. 1. Distribution and interrelationsof the Cree and Montagnais-Naskapidialects. From Michelson, 1939.

Michelson's treatise of 1939 reflects the hetero- the dialect of Montreal Lake and Stanley and Pelican geneous nature of his data; it is best represented by Narrows in northern Saskatchewan also shows y. his map (map 1). (Plains Cree is also spoken on Rocky Boy's reserve Map 2 and the summary which follows are based near Havre, and, at least during the earlier on Michelson's study but include other data as well part of this century, also at Turtle Mountain and (cf. Wolfart, 1971b); they may be regarded as a Fort Totten, North Dakota.) working classification. R: Isle A la Crosse, Saskatchewan and areas to the Y: Plains Cree and the dialect of northern Alberta; north of there; Tete-de-Boule Cree, Quebec.

HUDSON _?\ -BAY

N

''[N

ONT-R^ '-f,? NT Rt RQuE QUE

MAP. 2. Cree dialects according to reflex of Proto Algonquian *l. 10 WOLFART: PLAINS CREE [TRANS. AMER. PHIL. SOC.

,'p % 19-^ 70.B%

MAP. 3. Distribution of Cree groups. --- Approximate distribution of Cree-speaking grolups 1970. - Plains Cree (culturally defined; after Mandelb;aum) 1860.

TH: Woods Cree at Lac La Ronge, Saskatchewan wan.'2 These figures reflect only "treaty Indians" and in Rupert's Land (between the lower courses of not all of whom necessarily still speak Cree; on the the Nelson and Churchill Rivers). other hand, they ignore the large iumber of non- N: Swampy Cree, in a broad belt from Cumberland treaty Indians and Metis many of whom do speak House, Saskatchewan (just west of The Pas, Mani- Cree, so that they may be useful as a gross estimate toba) to the coast of Hudson Bay and James Bay, after all. from the Nelson River in the north to the Albany The historical movement of the Plains Cree from River in the south.10 their seventeenth-century location in an area bounded L: Moose Cree, spoken at Moose Factory (Mooso- by Lake Superior, Lake Winnipeg, and Hudson Bay, min), Ontario and in the lower portion of the Moose to the western Plains and the foot of the Rocky River drainage. Mountains has been described in great detail by Mandelbaum Mlandelbaum's chief concern 1.23. Plains Cree (1940). is the dramatic change, brought on by the fur trade, The Plains Cree dialect which forms the basis of from the aboriginal woodlands culture to that of the the present study is spoken primarily in the central Plains; but the texts even of today bear eloquent part of Alberta and in the central and southern parts witness to the woodlands heritage, as for instance in of Saskatchewan. Since language-based figures do the Windigo stories. not exist, we can only cite Canadian government Map 3 shows the range of the Plains Cree (cultur- sourcesll which indicate a total of about 26,000 Plains ally defined; after Mandelbaum, 1940) in the 1860's; Cree and 12,000 other Cree in Alberta and Saskatche- it also indicates the approximate distribution of Cree- speaking groups at present (after Canada, 1970). 10Ellis's Spoken Cree (1962) and 'Cree Verb Paradigms" (1971) are composite treatments of the n-dialect of Fort Albany and the 12 In 1874 Lacombe had estimated the number of Plains Cree i-dialect of Moose Factory; in Ellis, 1962 the text shows I and the as 15,000 to 16,000 (1874a: p. x). These figures would seem to tapes, n In the present study, the dialect(s) described by Ellis is fit the estimate of Chafe (1962: p. 165) who gives a figure of referred to as James Bay Cree and cited with n. 30,000 to 40,000 for all Cree (and 5,000 for Montagnais-Naskapi). 11Canada, 1970. These figures are based on cultural affiliation According to government sources for 1970 (Canada, 1970) there and other criteria and are not to be relied on as guides to dialect are about 16,000 Cree in Manitoba and another 16,000 in Ontario or even language assignment. and Quebec, bringing the total to about 70,000. VOL. 63, PT. 5, 1973] INTRODUCTION 11

In Alberta, speakers of Cree are found as far west textual data. In addition to some direct elicitation, as Duffield and as far south as Hobbema and Rocky their help was employed mainly in the extensive Mountain House-allowing, of course, for a scattering grammatical analysis of texts, in paraphrasing, etc. of even beyond these limits; this is especially JV who has since died was in his sixties. He was true for the foothills region north of the Edmonton- born and raised at Long Lake but later attended a Jasper highway. In the northern part of the province, boarding school in the south of the province. His the Peace River may be considered the western and wife is from Saddle Lake and since they lived in northern boundary. But even beyond this line, as relative isolation, some Saddle Lake influence may in the urban centers of the south, Cree is often used be expected to show in his speech. JV's unfailing as a lingua franca; as Lacombe put it (1847a: p. xi): patience and the kindness which both he and his wife "On peut dire que le cris est pour le Nord-Ouest ce extended to me are memorable aspects of my field que le frangais est pour les pays civilisEs." experience. Throughout the Cree-speaking areas of Alberta it is MC is in her early twenties. She was born and the first, and dominant, language of all those who raised on one of the Hobbema reserves; her slow grow up in the more remote areas. In the immediate speech proved particularly advantageous during the vicinity of urban centers, especially Edmonton, many early stages of field work. children today learn to speak English first, and some A large body of texts was collected in 1967-1968 never acquire a working command of Cree; the social mainly from the "southern" area. A representative pressures in favor of English are, of course, consider- selection (comprising approximately seven hours) of able. In the more remote areas, monolingual speakers these texts has been deposited in the Library of the are not uncommon among the older generation. American Philosophical Society at Philadelphia. The In addition to local differences, there is a relatively informants who appear in this selection are briefly clear distinction, however slight, in Alberta between introduced below. a "northern" and a "southern" variant (both y- AM is said to have come from Saskatchewan as a dialects). The boundary runs somewhere between child and ML was born at Riviere qui Barre, north- Edmonton and Lesser ; that is, it coincides west of Edmonton; all others were born, and spent with the cultural and ecological boundary of plains most of their lives, in the Hobbema area. and woodland. AM is over ninety, the patriarch of an important However, there can be no doubt about the full family; his repertoire of texts is known to be extensive. mutual intelligibility of these variants whatever the PO who has since died was seventy-eight years old at historical situation may be. We therefore tentatively the time, and was considered to be one of the few extend the of the domain linguistic term "Plains Cree" "pure-blooded" Cree still alive; the full meaning of the limits of the cultural unit beyond with which it this term could not be ascertained. He speaks very is primarily associated. slowly and clearly but his repertoire, at least as The isolated dialect differences which have been displayed towards me, seems to be limited. WW observed are noted in the relevant sections of this appears to be in his eighties, he is rather weak and there is also some lexical study (e.g. 5.33); variation his enunciation is none too clear; he is apt to get between the northern and southern areas. One confused in his story-telling. observation of a more is that the general range speed The other informants represented in the selection of utterance seems to increase as one considerably all appear to be in their sixties. PL's Cree differs moves north. from that normally heard at Hobbema by being much faster; his family seems to back 1.3. DATA AND INFORMANTS ultimately go to Rocky Mountain House where some "Saulteaux" The present investigation is based primarily on influence is said to exist. CL is his wife, and MY and data collected in central Alberta in 1967-1968.13 How- JY are her brothers, all of them living close to each ever, examples from Bloomfield's published texts other. These last four informants are all very (1930, 1934) are also used liberally so that the in- traditionally minded people, observing the old ways terested reader may examine the wider context. and openly longing for the day when the White Man Bloomfield's texts were recorded at Sweet Grass will be gone and the world in its proper state again. Reserve (near Battleford, Saskatchewan) in 1925; The Cree themselves classify all texts into two the , sometimes down to minute details, categories, dtayohkewin and dcimowin. The first of between Bloomfield's texts and those recorded in these is translated as "sacred story," and while it 1967-1968 is indeed remarkable. may contain fairy-tale elements even of European or Two informants provided the bulk of the non- Oriental origin, and the like it more properly refers to stories about the earlier state of the world and the 13 Concordances and indices of Cree an inverse stem texts, of the culture lexicon, and a variety of statistical information are among the exploits hero, wtsahkecdhk. These preliminary results of a computational project which is described sacred stories in the narrow sense are highly conven- in Wolfart and Pardo, 1972. tionalized, down to the linguistic structure of particu- 12 WOLFART: PLAINS CREE [TRANS. AMER. PHIL. SOC. lar stretches of the story. New stories may be made inverse: -ik TA (3')-3 up on the existing pattern. -ikoyiwa TA (3')-3' Several of these sacred stories have been recorded in more than one version. "Wisahkecahk and the For transitive inanimate (TI) , only the actor is indicated TI since the number and obviation Shut-eye dancers," for example, was recorded in at (e.g., 2p) least two versions Bloomfield 34-40 distinctions of the inanimate goal are not morpho- by (1930: pp. reflected in the verb and 1934: twelve versions, of logically (cf. 5.13). pp. 282-284); varying When cited in verbs are in- quality and length, were recorded in 1967-1968. isolation, generally The term acimowin is translated in- flected for a third person actor; nouns and pronouns usually by are in the formants as 'true story,' which may be taken to given proximate singular. In Cree forms, or indicate imply that texts of the sacred story type are not leading trailing historical in the usual sense of that term. An that a segment is not a free form; when a form is cited in dcimowin may concern event, it may morphophonological representation, leading any everyday or are omitted. In be an anecdote, a story, or the like; but there trailing hyphens usually phonemic funny a within a word marks it as is an important subclass of historical narratives, called representation, 'old-time These deal with (6.5). kayds-acimowin story.' We use Bloomfield's A military exploits of the horse-raiding days, or with (cf. appendix other historical be recollec- and Bloomfield, 1930: pp. 2-6) except for the purely topics; they may personal mechanical substitution of and c for his and tions, or "recollections on down o, e, u, d, by proxy," passed ts. Phonemic is indicated italics. the generations. That they contain magical experi- representation by ences does not disturb their status as true stories. This mode of representation14 is used throughout, even when smaller than words are cited. A final, non-native, subclass has for the moment strings Morpho- enclosed in been labeled "exhortatives"; they usually contrast phonological notation, namely strings slashes (and the additional characters the of the buffalo economy with today's /e/, /0/, /L/), golden age is used where it is relevant to the misery and, especially, alcoholism. only immediately discussion; cf. appendix A. 1.4. ABBREVIATIONSAND CONVENTIONS 1.42. Text References 1.41. Grammatical Information Most of the illustrative material is quoted from The use of technical has been kept the texts (see below). Examples which are not to a minimum. specifically identified are taken from my field notes. The verb classes are occasionally referred to by the An identification included in parentheses, e.g. following symbols: (T55p62), means that the form is not cited verbatim TA transitive animate but regularized or in a citation form. Glosses are not TI transitive inanimate necessarily uniform throughout. AI animate intransitive Although unpublished, the texts collected by II inanimate intransitive myself are identified by a number preceded by T, e.g. T105. Since they are in varying states of editing, The abbreviations of the person-number-gender-obvi- reference is sometimes made to paragraphs (by a ation categories (indf, 1, lp, 21, 2, 2p, 3, 3p, 3', hyphen) and sometimes to pages (by the letter p). 0, Op, 0', O'p) are defined in table 1 of section 2.01; Thus, 105plO refers to page 10 of text 105, and the choice of numbers should be largely self- T91-6 refers to paragraph 6 of text 91. explanatory. Bloomfield's published texts are identified by S for In discussions of transitive animate (TA) verbs, a Sacred Stories of the Sweet Grass Cree (1930) and by P combination like 1-3 is to be interpreted as the first for Plains Cree Texts (1934). Citation is by page person acting on the third; an inverse action, with 3 and line, e.g. S247-34. acting on 1, would be indicated by 3-1. Passages from the texts are left exactly as originally Transitive animate (TA) verb forms which involve printed except for obvious misprints and the me- third persons exclusively, show only one of the refer- chanical replacement of certain symbols; see 1.41 ents morphologically expressed (cJ. sections 5.622 ff.). above. Note especially the frequent writing of final h Textual examples, however, will be more intelligible in Bloomfield's texts; cf. appendix A and Bloomfield if the syntactic referents (rather than merely the 1930: pp. 2, 3. significative and morphological status of the verb form) are indicated. That referent which is not 2. GRAMMATICAL CATEGORIES expressed morphologically, is enclosed in parentheses, The categories of Cree are e.g. major , number, person, and obviation. The cate- direct: -ew TA 3-(3') -eyiwa TA 3'- (3') 14Cf. appendix A, footnote 85. VOL. 63, PT. 5, 1973] GRAMMATICAL CATEGORIES 13 gories are present in nominal and pronominal as well as in verbal inflection. The further categories of direction and of verbal order and mode are present in verbal paradigms only and are discussed in 5.62 and 5.3, respectively. Direction, however, not only plays a role in the construction of diagrams which would display the grammatical categories; it also involves problems which are of a sufficiently general nature to be treated here (2.5) rather than in the context of verbal inflection alone.

2.01. Major Dimensions of Contrast The major dimensions of contrast are displayed in table 1. The table also defines the abbreviations for the person-number-gender-obviation categories. Table 1 shows the most salient feature of the inflectional categories of Cree: that the third-person FIG. 1. category not only contrasts with the first and second persons but also is the domain of the contrast of proximate and obviative. Thus, it functions in both shows neither the various combinations of the non- these dimensions simultaneously. third persons (especially 21) nor does it reflect the The diagrams which follow are attempts to represent wide meaning of the third (non-obviative) category graphically the "fulcrum" function of the third-person (cf. 2.23). Note further that Hockett does not category. Figure 1 is a slightly modified version of interpret the relation between first and second person Hockett's diagram of 1966; cf. 2.24. as a manifestation of direction (cf. 2.5); otherwise the The usefulness of the diagrams is limited because one grammatical dimension of direction would be of the multi-dimensionality of the categories. Both diagrams disregard number entirely. Both are also Proximate misleading in suggesting some but not all possible paths of action between categories. Further, figure 1

Inanimate 0 TABLE 1 A GRAMMATICAL CATEGORIES DIRECTION) (OMITTING ^~~~\ ~ Obviatlve Gender Person Obviation Number Code

Animate Indefinite (sg/pl) indf First and pl 21 second V Animoat First sg 1 pl lp Second sg 2 pl 2p Third Proximate sg 3 pi 3p Obviative (sg/pl) 3' ' Third Inanimate Proximate sg 0 pl Op Obviative sg 0' pl O'p Non-third FIG. 2. 14 WOLFART: PLAINS CREE [TRANS. AMER. PHIL. SOC. represented by two dimensions of the diagram. This feature is yet to be found; however, there is at Figure 2 (which ultimately derives from figure 1) may least one set of contexts where a link might con- be interpreted to reflect the wide and narrow meanings ceivably be established. of the non-obviative category. It purposely excludes (a) In a series of coordinate nouns which function the combinations of the non-third categories and the as the joint goal of a transitive verb,'8 animate direction of action among them. obviative and inanimate plural are obviously indistinguishable: 2.02. Animate Obviativeand Inanimate Plural S 57-17 ... otinam ekwah ospikekanah ekwah Throughout the grammar of Cree, the inanimate take(TI 3) and rib (Op) and (proximate) plural category and the animate obviative otihtihkosiwah . . . show the same forms. category (number-indifferent) kidney (3') This identity"5 is seen in the inflection of nouns for number-obviation (3.3) as well as in the independent 'he took the ribs and the kidneys' forms of the verb where a direct contrast of gender (b) If a noun which is usually inanimate functions exists; i.e. in the obviative forms of the transitive as the goal of a transitive animate verb of speaking, verbs with animate or inanimate goal (TA, TI) and of the noun ending is completely ambiguous: the intransitive verbs with animate actor (AI); and in the plural forms of the intransitive verbs with S 8-43 soskwdc kahkiyaw kakwecimew inanimate actor (II). In all these paradigms, the right away all ask(TA 3-(3')) inanimate plural and the animate obviative are both otdpacihcikanah. marked by the ending -(w)a. utensil (Op) Such a similarity may of course be accidental, just as the number-indifference of the Cree (animate) 'Without delay she asked all her utensils.' obviative is due to the in this in merger, position is inanimate but here stands Cree, of Proto Algonquian *h and *1; thus, the Cree otdpacihcikanah normally in construction with kakwecimew; for more details ending -a (morphophonologically /ah/) corresponds to the Proto see 2.3. Algonquian (obviative) singular ending same *-ali as well as to the *-ahi. (c) Another aspect of the phenomenon (obviative) plural ending animate verb forms But if such a were the cause of the involves a set of transitive (TA) development where an animate third is the of an identity, it would have to antedate Proto Algonquian, person goal inanimate actor which is otherwise unspecific (5.83). for even there the inanimate (proximate) plural and Another set ff.) also has an animate third person the animate obviative singular are marked by the (5.61 as and, in reference, has an obviative actor; in same ending, *-ali. goal and the actor is not specified. That the identity cannot be reduced to historical meaning morphology, Thus, unless the situation is clarified by the context, a accident (at least not of such superficiality) is evident form like may be interpreted from the inflectional paradigms of pronouns. In T60pll e-pZikiskwatikot as either 'it or to him (3)' or 'he (3') addition to the standard nominal paradigm (4.3) (0 Op) speaks to him and this would be one meeting there are two exclusively pronominal paradigms both speaks (3),' of which show the same in of point of two seemingly disjunct categories. identity spite great rise differences of One uses the In fact, the situation just described may give phonological shape. (4.1) individual -hi for both these another to striking ambiguities. Thus, not only ending categories, (4.2) forms or but eha.16 verbal nominal-pronominal phrases entire clauses be with If we rule out accident as the cause of the identity may truly ambiguous respect to the under of the animate obviative and the inanimate plural, categories investigation, e.g., we have to look for that semantic feature of Cree T103p8. . . nandtohk ohi which these categories have in common.17 various this (0p/3') 16 The usual term for such a phenomenon, syncretism, is avoided kikwasa 5hi e-wiyesimikot. because of its directional implication (of a merger of previously separate categories). something (Op/3') trick(TA (Op/3')-3) 16 Ojibwa shows the same identity in pronouns which greatly 'all these various little things used to trick him.' differ from each other in phonemic shape; see Bloomfield, 1958: p. 43. to both the feminine singularand the neuter plural. Cf. Schmidt, 17 The situation is somewhat reminiscent of the Indoeuropean 1889; Meillet, 1937: pp. 291-292; and Lehmann, 1958; on the case where in many languages the same endings occur in feminine semantic aspects of this case see Kurylowicz, 1964, especially pp. singular and in neuter plural nouns, e.g. via 'street': iuga 205-206. 'yokes'; Sanskrit send 'army': yugd 'yokes.' In some languages, The Cree data do not suffice to attempt a similar historical moreover, neuter plural nouns take a singular verb form, e.g. account (in spite of the stimulating suggestions of Warren Cow- Greek 7rrvra ,be 'all things change.' gill and Francis Pardo), and any directionalinterpretations (split, This state of affairs is taken as evidence for an earlier category merger, category re-alignment)have to be postponed. of individual vs. collective, with the original collective giving rise 18 For the choice of verb type cf. 2.33. VOL. 63, PT. 5, 1973] GRAMMATICAL CATEGORIES 15

2.1. PERSON generality seems to be a much more important characteristic of this form's than indefinite- Distinctions of person are found in the actor and, meaning also 4.422. mi- is used when there is no within the transitive animate (TA) paradigm, also ness; cf. cross-reference or when it is irrelevant. in the goal of verbs. They further appear in the Thus, consider possession paradigm of nouns and in the personal The pronoun paradigm. third-person T 10p128 ekwa wiya oma are also involved in the inanimate intransitive (II) and (emphatic) this paradigm of verbs as well as in the number-obviation paradigm of nouns and of some pronouns. micihciy e-pe-otinahk. Because of its close parallelism to the person cate- hand (0) take(TI 3/0) the indefinite form of gories proper, possessor depend- 'and then this hand took it . . .' ent nouns is also included here (2.12). The indefinite actor of verbs, on the other hand, is more properly when the narrator is asked whose hand it was that the subject of section 2.5; it shows no prefix. reached in through the window, the explanation is, 2.11. Personal ki-, ni-, o- T 10pl29 tepiyahk awiyak ocihciy The personal prefixes ki-, ni-, and o--0 mark the just somebody(3) his(3) hand(O) basic in the of person categories possessive paradigm 'just somebody's hand.' nouns and in the independent order of verbs. ki- marks the second person, or addressee; ni- marks the awiyak 'somebody' is indefinite but specific (and is first person, or speaker; and o--0 marks the person therefore cross-referenced by the third-person prefix which includes neither speaker nor addressee, namely o- of ocihciy) whereas the prefix mi- indicates the third. Of the third-person alternants, o- appears generality. in the possessive inflection of nouns, in the personal- Further examples: pronoun set, and in the independent h- and ht-preterit T 34p7 mistikwdna ka-wdpahtamdn . . . of verbs; elsewhere the third-person prefix is zero. head (Op) see(TI 1) The combination of these prefixes with stems begin- ning in a vowel is described in appendix A: 6. 'heads I saw . . ni-, ki-, and o--0 constitute a position class of morphemes which are mutually exclusive; they are T 53p15 ekoni miyawa e-nahastdcik. not specific as to actor or goal. Thus, in the transi- this (Op) body(Op) bury(Al 3p) tive animate (TA) verb paradigm, a choice has to be 'These bodies buried.' made which person is to be expressed, and the members they of this position class can consequently be viewed as an The most extensive series of textual examples is found ordered set: ki- takes precedence over ni- and o-~0, in the sacred story of the rolling head (cihcipistikwan- o- That whenever a form and ni- in turn over -0. is, -atayohkewin); two children are being pursued by their involves a second person, whether as actor or goal, mother's severed head. The text is currently avail- the prefix is ki-; etc. able in three versions: Bloomfield, 1930: text 1; The ordering of the set of personal prefixes reflects Bloomfield, 1934: text 43; and T104 of 1968. a fundamental order principle of Cree: among the mistikwan 'a head' and ostikwan 'his/her head' occur which in turn person categories, second precedes first side by side in these texts, e.g., precedes third. This ordering principle is also manifest in the fixed order of in both noun and verb S 9-34 . .. 6ma mistikwan ka-tihtipipayiyik, head roll inflection. Non-third markers always precede third- this (0) (II 0') person markers, and among non-third markers, okawwdawa ostikwdniyiw. second-person markers precede first-person markers.'9 their mother (3') her (3') head (0) 'that their mother's head'; 2.12. Personal Prefix mi- rolling head, for further see also 2.31. Another prefix of the same position occurs with examples The o- form most characteristically occurs as the dependent noun stems only: mi- indicates a general base of further derivatives, e.g., otehimin 'strawberry,' possessor. The traditional term "indefinite possessor" nite heart.' will be used for the sake of convenience, even though otehipak 'cabbage,' etc.; cf. /niteh/ 'my mi- seems to be used primarily with reference to a 9 While it might be tempting to speculate about the socio- human possessor while o- typically occurs in nouns cultural of this Geary, 1943: implications ordering principle (cf. animal parts, as in slaughtering. However, p. 150), some writers (especially R. A. Logan) have made it the denoting basis of utterly fantastic claims about Cree personality structure. these are only tendencies rather than discrete distri- 16 WOLFART: PLAINS CREE [TRANS. AMER. PHIL. SOC. butions, and counter-examples are readily found: is of some theoretical interest, the choice of interpre- tation is of small to the T . . . oma relatively consequence 60p14 -manipitahk eyak description of Cree. We use the symbol "21" rather 3) same(0) this(0) pull(TI than "12" to indicate this category; but the really osit. wiy ekosi k-etwieyn, wiy important fact is its status as distinct from the two leg(O) (emphatic) thus say(AI 1) simple plural categories. misita ayisk. e-ydhkipitahk, moy, nayestaw 2.14. Combinations of Person Categories pull out(TI 3) no only leg(0') because All and their are he loose this duck's in person categories symbols given '... pulled leg (a leg roasting in table 1 of 2.01. fall into two sets, I it out They namely the ashes). Just as said, he pulled easily, the third and the non-third This no! for were person, persons.23 they only legs!' is evident not only from the semantic combinations A dependent stem denoting a body part may even already discussed but even more clearly from the be doubly inflected for possession; on the basis of morphological and semantic structure of the various ostikwan 'his head' (stem -stikwan-, prefix o-), a rare verbal paradigms. and interesting form shows double possession: S150-20 All non-third persons may act on, or be acted upon nostikwdnim 'my head' (said of a severed head which by, the third person; this fact is indicated in figures 1 is used as lodge-emblem). While the evidence is as and 2 of section 2.01. Action between first person yet inconclusive, this phenomenon might be inter- and second person categories is also viewed as in- preted as showing the derivational character of volving the dimension of direction (cf. 2.5). possession; see also 3.21. Simple reflexives are formed derivationally, e.g., niwapamison 'I see myself' (primary stem wdpam-, 2.13. Personal Suffixes derivational suffix -iso-; cf. 6.438). No forms appear to exist for the complex reflexivization which involves The personal suffixes combine with the personal the 21 category acting on, or being acted upon by, prefixes to mark the plural persons in the possession the other non-third categories. paradigm of nouns and in the non-third plural persons of the verbal order. independent 2.2. OBVIATION The following prefix-suffix combinations are com- mon to all nouns and verbs (3.22, 5.45): Within the third-person categories of either gender there are contrasts of OBVIATION. While the obviative lp /ni- -enan/ forms of Algonquian have occasionally been regarded 21 /ki- -enaw/ as fourth (and even fifth) persons, the Cree evidence -ewaw/ 2p /ki- (2.23) indicates three basic persons; the obviation The 3p possessor form of nouns also shows the suffix dimension functions within, rather than on a par /ewaw/; cf. 3.22.20 with, the third-person category. /ki- -enaw/ signals a plurality of referents which Contrasts of obviation are found in the inflection includes both speaker and addressee. This category of verbs, in the possession paradigm of nouns, and in has customarily been considered a first person plural the number-obviation paradigms of animate nouns inclusive category, and such a view agrees with the and pronouns. usual English translation of this category by 'we.' 21 But even where the dimension of obviation is The distribution of the prefixes, however, which covert,24 as in inanimate nouns, it is nevertheless follows from the order principle of 2.11, would lead present. This is easily seen when the noun stands in one to regard this form as basically a second person. construction with a principal clause, e.g., Thus, the contrast of inclusive versus exclusive S 58-1 . . ., kitahtawe kd-miskahk would be operative in the second person plural rather soon find(TI 3) than in the first. Comparative evidence seems to support this view; for in most of the Algonquian maskisinah. otinam, e-wa-wdpatahk ; languages22one of the suffixes (/enan/ or /enaw/) has moccasin(0p) take(TI 3) examine(TI 3) been generalized, so that the distinction of the forms rests entirely in the prefix. While this state of affairs eh-miywdsiniyikih. be pretty(II O'p) 20Note that the form /ewaw/ rather than /waw/ (cf. Bloom- field, 1946: p. 96) is due to paradigmaticleveling in Cree; cf. also 23For the non-third persons, Hockett (1966) uses the term 5.451 and 5.48. 'local.' 11But note that Bloomfield, whose translations are clearly u In James Bay Cree and in the Mistassini dialect of Mon- informant-inspired,translates two consecutive occurrencesof 21 tagnais-Naskapi, inanimate nouns have developed the contrast verb forms with 'you': S76-5. of proximateand obviative, using the thematic marker /eyi/; cf. 2 Cf. Goddard, 1967: p. 68. Ellis, 1962: p. 3-20 and Rogers, 1960: p. 110. VOL. 63, PT. 5, 1973] GRAMMATICAL CATEGORIES 17

. ., soon he found some moccasins. He picked cross-reference). We shall say that the proximate them up and examined them; they were very person, in any context, is in focus and the obviative pretty.' person or persons are not. The correlates of focus in terms of discourse are not The verb eh-miywdsiniyikih expresses the obviative analysis fully known, and Bloomfield's statement is status of its "actor," maskisinah. (1962: p. 38) only an approximation: "The proximate third person repre- sents the of discourse, the nearest the 2.21. Focus and Spans topic person speaker's point of view, or the person earlier spoken Within each CONTEXTUAL SPAN only one third person of and already known." 25 is proximate; all others are obviative. Thus, in There are only few environments where focus is a When a noun is niwdpamdw atim. assignment predictable priori. inflected for the is nearer than see (TA 1-3) dog(3) possession, possessor who or what is possessed; the latter is necessarily 'I see the dog.' obviative. Focus is or in assignment largely expectable, but not in a technical sense predictable, in another type of pimohtew ndpew. context. When a main clause involving a third walk(AI 3) man (3) person is modified by an inanimate clause indicating a state of the physical environment (climate, time of 'The man is walking along.' day, season, etc.), the latter is usually obviative: the third persons are identified as proximate in the S 53-31 ekwah verbs as well as in the nouns. In eh-tipiskdyik, then be night(II 0') atimwa. pakamahwew napew dcimostaww, . . . hit(TA 3-(3')) man(3) dog(3') tell (TA 3-(3')) 'The man hits the dog.' 'Then, when night came, he told him tales . . . on the other hand, or in niwdpamdw napew e-pakamahwat S243-24 eh-otdkosiniyik iyikohk, see(TA 1-3) man(3) hit(TA 3-(3')) be evening(II 0') at that time atimwa. kiwew. dog (3') go home(AI3) 'I see the man hit the dog.' 'When it was evening, she went home.' napew is and atimwa and the proximate obviative, S254-2 kitahtawe verb form indicates the same relation. peponiyikih mdna then be A basic exception to the above statement occurs winter(II 0') always when two nouns are in close parataxis; then both may kanawyimik . . . be proximate: take care(TA (3')-3) T 49-9 e-kN-notinitocik, ayahciyiniwak 'Then in winter-time he would take care of him. .' fight each other(3p) Blackfoot(3p) ekwa nehiyawak . . . 2.22. Inflectional Limitations and Change of Focus and Cree(3p) The inflectional machinery may not always be 'they used to fight each other, the Blackfoot and the sufficient to accommodate all the referents to be Cree, . . .' discussed.2 When such a situation arises, two possi- S264-33 mistahi kitimdkisiwak bilities exist: either there are several obviative referents which are very be pitiable(AI 3p) kept apart by non-inflectional means such as the context; or the span is ended and a kohtawiy kikdwiy new span, with new focus assignment, is begun. This your father(3) your mother(3) 25 For furtherexamples and discussionsof the semantic function kistes. of focus see Hockett, 1966; Frantz, 1966, and the literature re- your older brother(3) ferred to in the latter. Other systems of referent-indexingare describedby Jacobsen, 1967, who seeks to establish a typology of 'Very pitiable are your father, your mother, your such systems. older brother.' 26 Such a situation is more frequent in Cree than in those Al- as for The dimension of obviation thus marks a gonquian languages, instance Potawatomi or Blackfoot, semantic which show yet another opposition whithin the dimension of system of FOCUS (as well as the syntactic linkage of obviation; cf. 2.24. 18 WOLFART: PLAINS CREE [TRANS. AMER. PHIL. SOC.

second possibility seems to be preferred in conversa- P 98-11 piyisk kahkiyaw awiyak tional style, but there is a great deal of variation in finally all someone (3p) this matter and the details of stylistic variation are nisto yet to be explored. miyweyihtamwak; |1mina ayahciyiniwah be also three It may safely be said, however, that focus changes glad(TI 3p) Blackfoot(3') are frequent, i.e. that spans are relatively brief. This eh-nipahat, awa ndpesis, I is not true, however, in at least one text (Bloomfield, kill(TA 3-(3')) this(3) boy(3) 1930: text 10) which in its entirety constitutes only one span; i.e., it has constant focus assignment miyweyihtamwak ok dyisiyiniwak, . . . be throughout. glad(TI 3p) these(3p) people(3p) 2.221. The of obviation combined with machinery 'Finally everyone was glad; 11also because this boy that of the of nouns possession paradigm may provide had slain three Blackfoot, II those people were full cross-reference. Examples: glad, P 98-38 ohi napewah kd-nipahdyit 2.223. Change of focus (11) may be indicated that kill by man(3') (TA 3'-(3')) the same referent being assigned to different obviation owikimakaniyiwah itohtahew. categories, e.g., his(3') wife(3') take there(TA 3-(3')) T125-1 ekwa anihi ostesa, || 'He took to that place that man who had slain his and these (3') his(3) brothers(3') (own) wife'; ekonik esa ki-wiwiwak; the crucial cross-reference is that of the possessor of these thus marry(AI 3p) owikimdkaniyiwah 'his own wife' which is marked as 'And his had 3' by the prefix-suffix combination o- -iyi-. brothers, 1] they married'; S 36-39 . . . ekwa e-mzcisot, wisahkecdhkwa T10pl17 ekwa mina mana anihi then eat(AI 3) Wisahkecahk(3') and this(3') e-kimotamawdt onawaczwiniyiwa. k-acimdt 1] kayds ka-kz-kimotit rob(TA 3-(3')) his(3') roast(Op) tell about him(TA 3-(3')) long ago steal (AI 3) . . . then he Wisahkecahk of his ate, robbing anihi soniydwa. roasts'; this (3') money (3') Wisahkecahk's roasts: o- -iyi-. Contrast i.e., 'And then he told about this one [1 who long ago S 37-1 namoya kiskeyihtam asay stole this money.' not know(TI 3) already T 58pl9 sdsay kd-kwdskwwewepchomiht oniskima. eh-kitamwdyit already knock up(TA indf-3') eat up(TA 3'-(3')) his(3) goose(3') ohi ocemisisa; mitoni ispimihk 'He did not know that the other had eaten 1[ already this(3') his(3) horse(3') really in the air his geese.' kwdskwewepahok awa misatim. i.e., the geese belonging to the first-mentioned person, knock up(TA (3')-3) this(3) horse(3) indexed by o- as proximate third, and not the eater (3'). 'Already his little horse had been knocked up into 2.222. Change of focus (11) may be indicated, even the air; 1Ireally high up he (buffalo) knocked this within a , simply by indexing different horse.' referents with the same obviation category, e.g., T konta T 55p35 ninayomaw e-sakihak, 58p9 -pehtakosit; 11oma carry(TA 1-3) love(TA 1-3) just be heard (AI 3) nikdwiy e-mamitoneyimak ka-pehtakosiyit e-itdpit . . . my mother(3) think about(TA 1-3) be heard (AI 3') look around (AI 3) e-wz-petamawak. 'he (buffalo) just made a big noise; 1] when he want to bring it to him(TA 1-3) (buffalo) made a big noise, he (man) looked around . . .' 'I carried it (sc. a kettle(3)) on my back, I prized it, [I I thought of my mother and wanted to bring it 2.224. When several obviative referents are in- to her.' volved, may provide some clues; in the VOL. 63, PT. 5, 1973] GRAMMATICAL CATEGORIES 19 main, however, the identification of referents then (a) One such context is in the personal pronouns rests on meaning and context. where the third-person pronouns wiya and wista are used for both proximate and obviate referents. T124-4 okosisa sakdpekipahdyit Similarly the personal prefixes also show no distinction lead 3'- his son (TA (3')) (3) (3') in the third-person prefix. otma , othhtapiwina. (b) Close-knit nominal phrases where one noun his(3) horse(3') his(3) mount(3') shows possessive cross-reference with the other are often used with non-obviative verb forms even though 'His son was leading his horse, his mount.' they are inflectionally obviative: P 98-27 . . . . ohi kd-miyosiyit S 75-8 ekwa awa nikdn pimotam this (3') be beautiful (3') then this(3) first shoot (TI 3) oskinikiwah, ". . ." eh-itat macihkiwis ohtdwiya. (3) his father young man (3') say to him (TA 3-(3')) Silly-Fellow (3) 'Then father shot first.' owikimdkana. Silly-Fellow's (ohtdwiya: morphologically 3', syntactically 3) his (3) wife (3') '. ... of that handsome youth, "..." he would T124-4 kd-wdpamyadhk otitwestamdkew say to his (own) wife.' see(TA lp-3) interpreter (3) okosisa e-ati-pimipayit ekote T103p5 ekota es dnima mahihkana his (3) son start riding (AI 3) there there wolf (3') e-ispayit. otihtineyiwa ohi witimwa go there(AI 3) grab (TA 3'- (3')) this (3') his (3) sister-in-law (3') 'We saw the interpreter's son start riding to go e-ndnistipitamdtoyit, e-mowdyit. there.' tear eat him 3'- up jointly (AI 3') (TA (3')) (okosisa: morphologically 3', syntactically 3 as goal 'There the wolves grabbed these sisters-in-law of his, of kd-wdpamdydhk and actor of the remaining two tore them to pieces among themselves, and ate verbs; otitwestamdkew: morphologically 3, syntactic- them.' ally disregarded except as possessor of okosisa.)

S 8-5 poti ekwa kd-wdpamdt mistikwa T123-6 ota ki-pimipayit nanos and then see (TA 3- (3')) tree (3') there ride(AI 3) Nanos (3) e-pahpakamahwdyit, ekwa kd-pe-wayawiyit okosisa, wdhyaw nakasiwew. beat(TA 3'-(3')) and come out(AI 3') his(3) son far be ahead (AI 3) kinepikwa, kd-wdpamdt owikimdkana 'There Nanos's son was riding by, he was far ahead.' snake(3') see(TA 3-(3')) his(3) wife(3') (okosisa: morphologically 3', syntactically 3)

e-owicimosiyit. T114p3 . . . nimosom awa have as lover(AI 3') my grandfather(3) this(3) 'And then he saw her (his wife) beating a tree, and nipdpa opapawa he saw his wife have it when a serpent came out, my father(3) his(3) father(3') for a lover.' nisipwehtahikondn. 2.23. Marked Status of Obviative take him away(TA 3-1p) In the opposition of proximate and obviative, 'this my grandfather, my father's father, he took obviative is the marked member. us away.' One reason for this assertion is purely morphological. (opdapwa: morphologically 3', syntactically 3) In the morphological analysis of Plains Cree, present (c) The most typical context of neutralization is the obviative is marked a always by morpheme-/em/, provided when a verb has two adjuncts one of which or is added to the non-obviative /eyi/, /h/-which is proximate, the other obviative; the verb is then more of is that in form. Much important, course, inflected for a non-obviative plural referent, e.g., contexts of neutralization we find the proximate category which is thus clearly characterized as S 53-32 eh-kiskeyimdt, eh-nohtehkwasiyit, unmarked. Examples in (a) through (c). know (TA 3-(3')) be sleepy (AI 3') 20 WOLFART: PLAINS CREE [TRANS. AMER. PHIL. SOC. ekwa kawisimowak. 2.24. "Further Obviative" then go to bed(AI 3p) Most recent studies of Cree grammar assume a 'When he knew the other to be sleepy, then they further binary opposition within the obviative cate- went to bed.' gory, resulting in a "nearer" and a "farther" (Bloom- field) or "further" (Hockett) obviative. Such a P 98-40 wdpam esi-miyosicik system may well be historically justified for Cree. see(TA 2-3p) be so beautiful(AI 3p) Synchronically, however, there is strong evidence to that this further distinction is nisim ohi iskwewah. suggest spurious. The only point in Cree grammar where such a my brother(3) this(3') woman(3') further distinction would be expressed is in the third- 'Look how beautiful are my brother and this person forms of the transitive animate (TA) paradigm. woman.' The reasons which presumably led to the received analysis are explored in 5.631. One of its more salient T124-3 nete tahkohc-dyihk e-otihtdt weaknesses is that it would skew the TA verb paradigm there at the 3-(3')) top reach(TA and require an inordinate amount of paradigmatic kiyomdndkosiwak; mergings. be in full sight(AI 3p) In an argument (5.63) based on the overall struc- ture of the TA verb paradigm and on the nature of 'when he reached them at the top there, they the proximate-obviative opposition we attempt to were in full (including him) sight.' show that the present interpretation more adequately Another example involves a pronominal predication: portrays the facts of Cree. Certainly no evidence for such a distinction is found T 10p91 tdniwd etokwe in the inflection of nouns and pronouns.2 where is she(3) I wonder Thus, Hockett's description (1966) of the total omamdwawa. system of person indexing in the central Algonquian their (3p) mother (3') languages is too wide for Cree. While it may well fit 'Where is she, I wonder, their mother?' the referential system of Cree, it over-accounts for the morphological and significative patterns. Need- Thus the non-obviative category, being unmarked, less to say, of course, Hockett's system is easily has a wide and a narrow function and meaning. We adjusted to account for the Cree situation as viewed use the term "proximate" only of the narrow meaning, here.29 where it is opposed to "obviative." For the wide meaning, the term "third person" is obviously appro- 2.3. GENDER priate. These terminological conventions may be There are two gender categories, animate and diagrammed as follows: inanimate. The gender contrast is manifest through- out the inflection of and verbs. "thirdperson" nouns, pronouns, Although the evidence is inconclusive, the gender contrast is possibly neutralized in the identity of

obviation categories (symbolized as 3, 4, and 5) there is a non- indexed, i.e., general form as well. "This non-indexed form is not hypothetical but the 'name' of an item in a context where person indexing is not relevant, such as in response to the ques- tion, "What is the word for __? . . ." (Frantz, 1966: p. 51). Thus, the non-indexed form for 'man' is nina, the proximate, ninaoa, and the first obviative, ninai. 28 This is in striking contrast with the situation in Potawatomi, for example, where Hockett (1948: p. 72) describes the obviative marker/n/ as occurring twice in succession; e.g., mt'uk 'tree' (3), "proximate" "obviative"obviative mt'ukwen 'tree(s)' (3'), and mt'ukwenun 'tree(s)' (3"). How- ever, Hockett himself (1966: p. 64) calls the 3" forms "extremely It is interesting that the present analysis in terms rare" and says they are "perhaps avoided as 'awkward.'" of marked and unmarked members of an opposition Rogers (1963: p. 103) reports the same situation in Northern is in fact inherent in the system of abbreviations used Ojibwa, but without examples or further details. While Bloom- the tradi- field (1946: p. 94) asserts the distinction with reference to Cree, by many Algonquianists. (In practice, it in tional definitions of the abbreviations "3" is no mention of was found his descriptions of Menomini, Fox, differ; or Eastern Ojibwa. normally used only in its narrow meaning, 'proxi- 29In terms of Hockett's diagram (fig. 1; 1966: p. 60), only the mate.')27 "further obviative" node needs to be deleted. If the line leading to it is left to peter out, pointing to no specific node, this might a7The present analysis finds striking support in the Blackfoot provide a graphic conceptualization of the open-endedness of the situation where in addition to noun-forms marked for different Cree situation. VOL. 63, PT. 5, 19731 GRAMMATICAL CATEGORIES 21 animate obviative and inanimate plural endings which and is discussed in 2.02. A typical example is found where a inanimate noun functions as the of a T125-8 ..., mina kahkiyaw kikway usually goal and transitive animate (TA) verb of speaking (see 2.31 everything (0) below for further details and examples of disambigua- e-kZ-wayesimdt,. . . tion), e.g., trick him by speech (TA 3-(3')) S 8-43 soskwac kahkiyaw kakwecimew 'and everything he used to trick by speech.' right away all ask (TA 3- (3')) In most cases, however, one gender or the other is otapacihcikanah. obviously more appropriate, and so we find correc- her(3) utensils (Op/3') tions, e.g., T 58-3 sdkastenohkisi 'Without delay she asked all her utensils.' e-ayitdpit, look around (AI 3) in the east In this the verb stem indicates sentence, only clearly eyapic ka-wdpaht-- -, kd-wpamat the of the since the of gender goal ending otapacihci- further see(TI 3) see(TA 3-(3')) kana(h) is ambiguous as to animate obviative or inanimate plural; cf. section 2.02. However, when kikway; the direction of action is reversed even this indication something (0) of is removed because of the gender partial homonymy 'As he looked around, in the east in due course he (2.02) of the transitive animate inverse set with the saw (sc. kdawdpahtahk) - - - -, he saw something.' inanimate actor set of the same paradigm. Thus, a sentence like the following becomes completely The opposite situation occurs in ambiguous: T103pll e-mekwd-pimohtet, . .., dsay mina T100p4 "eha," k-etikot es ohi walk(AI 3) already again tell(TA these yes (3'/Op)-3) ka-wdpamdt ohi, kd-wdpahtahk osita. see(TA 3-(3')) this(3') see(TI 3) his (3) feet(0p/3') mikiwahpis. '"Yes," he was told by his feet.' wigwam (0)

2.31. Shijt of Gender 'As he walked along, ..., again he saw this one, he saw the wigwam.' Generally, all reference to speaker or addressee An is animate; witness the lack of first and second-person excellent example on a somewhat larger scale in forms in the inanimate intransitive (II) paradigm. is found the text of the rolling head, already However, that this is not a hard and fast rule but a referred to in 2.12. Consider the following consecu- tendency which involves competing pattern pressures, tive pair of sentences: is clear from the deal of variation that is en- great S 8-40 . .., kitahtawe tohkdpzmakan countered, e.g.,30 presently open eyes(AI inan(0)) T 87-1 kahkiyaw kskway pikiskwdtam. omah pisisik mistikwdnis. ekwah kitahtawe to everything (0) speak it(TI 3) this(O) mere little head(0) then presently 'to everything he spoke.' kd-pikiskwet om ostikwdn. speak(AI 3) this (0) her(3) head (0) Even the same narrator may use either gender in essentially the same context; contrast 'Presently it opened its eyes, that mere head. Then presently that head spoke.' T131-4 wiya kahkiyaw kikway for everything (0) However, even when speaking, the head may be inanimate: e-kZ-wayesihtahk awa wisahkecdhk. trick it by speech(TI 3) this (3) Wisahkecahk(3) S 9-4 . . . wdpahtam oma pisisik see(TI 3) this(O) mere 'For everything he used to trick by speech, this Wisahkecahk.' mistikwdn eh-pikiskwemakaniyik, . head (O) speak (AI inan (0')) 80 The situation of an otherwise inanimate referent speaking or being spoken to most typically occurs in sacred stories. 'he saw this mere head which spoke' 22 WOLFART: PLAINS CREE [TRANS. AMER. PHIL. SOC.

S 9-6 ostikwan pikoh ka-pikiskwemakahk. differ by the gender of the actor, e.g. ohpikiw 'he head (O) only speak (AI inan(0)) grows up,' ohpikin 'it grows up'; cf. also 5.1. In most instances, then, there is a choice, as to 'It is a head which talks.' only gender, among verb stems. Nouns are sharply When an otherwise inanimate noun becomes different since most of them belong to only one gender see 2.323 Even when are temporarily animate, the ambiguity under discussion (but below). they tempor- with animate may be removed entirely by the presence of modifiers; arily animate, showing agreement their inflectional far as are consider the demonstrative pronoun awa in verbs, endings (as they unambiguous) remain inanimate. S 54-42 nama ciy awa 2.321. Nouns which denote humans, animals, not (question) this (3) spirits, or trees are animate. E.g., ayahciyiniw 'enemy, especially Blackfoot,' czpay 'dead person, kitastotin ewako, "nik-dydwik!" corpse,' tahkohci 'On-Top' (personal name; there is your this (0/3) possess(TA 3-1) headgear(0/3) also a particle tahkohci). mistatim 'horse,' mostos eh-iteyimit kitastotin? 'buffalo.' dtayohkan (a certain kind of spirit), think of him(TA 3-1) your headgear(0/3) kise-manitow 'God.' sihta 'spruce,' mdyi-metos 'black popular,' mistik 'tree.' 'Is not this headgear of yours thinking this of me, Also animate are extensions of these, e.g., ayisi- ' "Let him possess me!" yinshkdn 'effigy, doll.' The transfer discussed so far has a statable 2.322. Also animate is a variety of objects some of gender which constitute well-defined semantic and function. Furthermore, it is relatively largely predictable and the of a one-way transfer, from inanimate to animate. It groupings. By large, gender assignment is thus distinct from the lack of concord nouns in Cree seems to correspond fairly closely to quite gender that of MIenomini which Bloomfield has described in that is occasionally encountered in the texts, e.g., great detail (1962: pp. 28-36). T 87-6 e-nitoniket oma sihta Some body-parts are animate: nisakitikom 'my grope(AI 3) this(O) spruce(3) braid,' nitasiskitan 'my calf of leg,' nitihtikos 'my shoulder-blade,' 'he for that kidney,' nitlhiy 'my niyihk 'my groped spruce' gland,' etc. It may be relevant that such discrepancies seem to be Animal hides and garments made from them: particularly frequent with noun stems showing class- waposwaydn 'rabbit skin,' mostoswaydn 'buffalo-robe,' cleavage, e.g., mistikw- animate 'tree,' inanimate maskwaydn 'bear skin,' etc. 'stick'; note also the homonymy of 3' and Opdiscussed Certain plants and their products: mahtdmin 'grain in 2.02. of maize, ear of maize'; pahkwesikan 'bannock,' pzswehkasikan '(leavened) bread'; pikiw 'gum, rosin'; S 48-10 osdm miywdsin awa ayoskan 'raspberry' (but not otehimin 'strawberry'), indeed be good(II 0) this(3) pitikomin 'dried prune,' pakdn 'nut'; etc. Tobacco and other items from its cistemdw mistik .. sphere: 'tobacco,' ospwdkan 'pipe,' ahpihcis 'tobacco-pouch.' stick (0) Some natural objects (perhaps in relation to their 'indeed it is good, this piece of wood' function as atayohkan?; see T104): pisim 'sun, moon,' pisimohkdn 'clock, watch, "pseudo-sun,"' acdhkos T 76p7 kd-pasastehahk mistikwa ohi 'star'; asiniy 'rock, stone' (cf. 2.323 below); kona etc. whip(TI 3) tree(3') this(3'/Op) 'snow,' maskwamiy 'ice'; soniydw 'gold, money'; Some articles of personal or household use: askihk 'he whipped these trees' 'kettle,' napwenis 'little frying pan' (loan from French la asdm It remains to be seen whether such instances are poele), kwapahikan 'ladle'; 'snowshoe,' akwdndn 'shawl,' nitds 'my trousers' (but inanimate really accidental slips of the tongue (or lapses of the in the meaning 'my gaiter'); akwask 'knob-shaped record) or whether they to features of perhaps point arrow head,' pahpahahkwdn 'shield,' etc. the gender dimension which are not understood. Further and more specific groups may be set up, but their predictive value is obviously low. The 2.32. Gender Classes of Nouns above examples are given mainly to illustrate the Gender is one of the basic criteria for the inflectional wide range of animate nouns. In short, only a list and derivational classification of verbs. Transitive can account for the gender of Cree nouns. animate and transitive inanimate stems largely come 2.323. Some noun stems are subject to class- in pairs, differing as to the gender of the goal, e.g., cleavage, taking both animate and inanimate endings, otinew 'he takes him,' otinam 'he takes it'; animate e.g., akohp 'blanket,' askipwdw 'wild potato,' etc. intransitive and inanimate intransitive stems similarly The animate and the inanimate stem often have VOL. 63, PT. 5, 1973] GRAMMATICAL CATEGORIES 23 different meanings. Thus, for example, the animate environment whose internal categorization is less stem mistikw- means 'tree,' the inanimate, 'stick.' generally agreed upon. Thus, the skewed relation- Consider ship of gender and physical properties provides a doubtful basis for investigating the relationship be- S 8-8 mistikwah ih-pakamahwdyit tween the genders.3' tree (3') hit(TA 3'-(3')) Hockett very aptly characterizes the imbalance wwa, . . . of the gender categories by terming the animate his(3) wife(3') gender "absorptive" (1966: p. 62). The crucial feature, however, is not that it includes nouns whose 'when his wife struck the tree, . . .' denotata are lifeless from a Western point of view but that inanimate nouns may temporarily become S 48-5 kitahtawe miskam .... mistik, animate in their syntactic behavior (cf. 2.31) whereas presently find(TI 3) stick(0) animate nouns do not, in a similar way, become in- 'presently he found a stick, . . animate. "Thus, there are routes for a shift of gender from inanimate to animate, but not the op- It is T102p4 ekwa mistikwa, . . ., k-mihcenwa, posite" (1966: p. 62). this argument which would seem to indicate that the animate is and stick (Op) be many (II Op) gender more general than the inanimate. 'and pickets, ..., they used to be plentiful, ... However, the inanimate gender also occurs in contexts of neutralization. For transitive Another of semantic differentiation is example, example animate (TA) verbs take a cikahkwan which means 'lance' when double-goal (cf. 6.446) inanimate, and second of either but the most charac- as animate noun denotes a certain goal gender, gambling toy teristic type of these verbs, namely those in /am-aw/ like a Bloomfield, 1930: 281. shaped knife-blade; cf. p. 5.814), are derived from transitive inanimate as animate noun (cf. asiniy, finally, means 'stone'; stems. Consider dtotamawew 'he tells for consider also (of him/it) misasiniy 'big stone.' The inanimate him' which is derived from the TI stem dtotam 'he tells noun means 'bullet' and occurs in asiniy such combina- about it'. Further well show that tions as investigation may niskasiniya 'bird-shot,' moswasiniy 'bullet inanimate second are more or etc. goals typical, perhaps (for moose),' historically prevalent, in this environment; but Of course, there may also be dialect differences in synchronically, this is a context of neutralization the gender assignment of nouns. Thus, one text which shows the inanimate member of the from Fort Vermilion in Northern Alberta consistently opposition to occur. shows soniydw 'gold, money' as inanimate, e.g., e-nitawmyihtamdn oma soniydw 2.4. NUMBER want(TI 1) this(O) money Number is singular and plural. In general, 'I want this money' contrasts of number are found throughout the inflec- tion of and verbs. In 2.41 we even though soniydw is usually animate. nouns, pronouns, describe the limited contexts the number 2.324. It does not appear practical to look for where semantic groupings among inanimate nouns. contrast is lacking. One generalization, albeit a weak one, may be advanced on the basis of derivational and semantic 2.41. Absence of Contrast criteria taken together. Abstract nouns derived In the set of the transitive animate from verbs with the suffix -win are inanimate; such you-and-me the second is number- nouns are freely formed and very frequent of occurr- paradigm (5.64), person ence; cf. 6.414. Examples: nehiyawewin 'Cree-ness, indifferent in the environment of the first person Cree speech' (cf. nehiyawe- 'talk Cree'); dcimdwin plural (Ip). Thus, mdmitoneyimindn 'think of us!' 'story' (cf. dcimo- 'tell a story'); mdkweyimowin 'scare' may be addressed to one or several people; similarly, (cf. mdkweyimo- 'feel pressed upon'); etc. e-k1-papdmi-nitondtdhk 'we were looking around for you' is indeterminate as regards the number of the 2.33. Marked Status of Inanimate goal. In considering the gender categories of Cree we The number distinction is lacking in the indefinite encounter the familiar contrast of grammatical and possessor of nouns (2.12) and the indefinite actor "natural" Where gender. gender largely corresponds 31 to as in there will be little For this reason, Goddard's assertion that "the class of ani- sex, Indoeuropean, argu- mate includes a number that refer to non-animate ment over the of and nouns good discrepancy gender sex; e.g., things" (by whose criterion?) and his inference that "the animate German das Weib 'the woman.' In Cree, however, gender is the more neutral of the two" (1967: fn. 74) are of doubt- gender correlates with a feature of the natural ful validity. 24 WOLFART: PLAINS CREE [TRANS. AMER. PHIL. SOC. forms of verbs (5.84, 5.85) as well as in the inanimate Within the category of direction, DIRECT is the actor set of the transitive animate paradigm (5.83). unmarked member, and INVERSE the marked. This Since the goal of transitive inanimate verbs is not analysis is based on evidence from the morphology morphologically expressed (5.13), there is also no of the paradigms which fits in well with a much indication of number. Consider S48-5 miskam mistik more general phenomenon observed in Cree. Among 'he found (TI 3) a stick (0)' and S48-13 e-wapahtahk the morphological evidence, the highly productive onimdskwdkanah 'he looked (TI 3) at his (3) weapons theme sign /ekw/ (see 5.422 for further references) (Op).' is the most obvious case. By far the most striking and pervasive instance of The unmarked status of the direct member of the number-indifference, however, is that of the animate direct-inverse opposition finds strong support in the obviative forms. This number-indifference is due to fundamental order principle which holds among the the historical development of Proto Algonquian *l person categories of Cree (cf. 2.11) and in turn and *h both of which in this position correspond to emphasizes the generality of that phenomenon. The Cree h (which in turn is non-distinctive in word-final relative position of the person markers within a position). Thus, while there are two obviative end- two-referent verb form is fixed (cf. also 5.62). In the ings in Proto Algonquian, namely singular *-ali and direct forms, the actual linear sequence (in time or plural *-ahi, Cree only has the one ending - (w)a "left-to-right") of the prefixes and suffixes corresponds (morphophonologically /(w)ah/) for both numbers.32 exactly to the priority of second over first, and of second or first over third. In the inverse forms, the 2.42. Marked Status of Plural actual linear sequence remains unchanged but the reversal of the fundamental priority order is indicated Singular is regarded as the unmarked member of theme the number This is based not by signs. opposition. analysis The direct set, then, consists of those forms whose only on the morphological fact that a plural morpheme action is is added to singular forms. More important, a collective singular occurs in statements of general (a) from a second person onto a first person, e.g., applications, e.g., kitasamin 'you feed me (2-1)'; from a non-third onto a third .... (b) person person, e.g., ..,. ek k-tosket nitotemindn nitasamaw 'I feed him not (1-3)'; work(AI 3) our(lp) kinsman(3) (c) from a proximate third person onto an obvia- 'our friends had no work . . tive third person, e.g., asamew 'he feeds him (3-(3'))'; T 62pl ohtitaw ta-pahpit aylsiyiniw, . . . (d) from an obviative third person onto another,33 always laugh(AI 3) man(3) e.g., asameyiwa 'he feeds him (3'-(3')).' 'people will always laugh, . . .' The forms of the imperative order are all direct, either with a second person acting on a first person, T 72p22 ekwa nandtohk e-pihtokwet dta e.g. pehik 'wait for me! (2p-l)'; or with a second person then all kinds enter(AI 3) here acting on a third person, e.g., pehihk 'wait for him! oma tipahaskdnihk aylsiyiniw. (2p-3).' The inverse set is the of the direct this (0) reserve (loc) man (3) exactly opposite (with the exception of the imperative and the in- 'Then all kinds of people, different nations entered definite actor forms which are not symmetrical). this reserve.' The action is 2.5. DIRECTION (a) from a first person onto a second person, e.g., kitasamitin 'I feed you (1-2)'; from the indefinite actor the inde- Apart forms, (b) from a third onto a non-third person, e.g., and orders of the transitive animate pendent conjunct nitasamik 'he feeds me (3-1)'; verb fall into two symmetrical sets distinguished (TA) (c) from an obviative third person onto a proxi- the of DIRECTION. These sets are charac- by category mate third person, e.g., terized by an in the actor-goal relation; opposition asamik 'he feeds him ((3')-3)'; e.g., (d) from another onto an obviative third person,33 (1) kiwapamin 'you see me'; e.g., (2) kiwdpamitin 'I see you.' asamikoyiwa 'he feeds him ((3')-3').' 82 Note that in inanimateintransitive (II) verbsthe obviative Direction is morphologically expressed by theme is markedonly by the suffix/eyi/; the -waof that paradigmis the pluralmorpheme. 33For details on these two types see sections 5.61 to 5.63. VOL. 63, PT. 5, 19731 GRAMMATICAL CATEGORIES 25 signs (5.42) which also indicate the agreement of the more neutral, and the other would not occur without verb with its nominal complements. appropriate textual environment. The extreme case of direction being completely 2.51. Direction and Focus determined by focus is rare; an example occurs when a possessed noun acts on its possessor: Semantically, direction serves to specify actor and goal. In sentence (3), for instance, the direct theme can otema ki-mdkwamik. sign /a/ indicates the noun atim as goal, whereas the John (3) his (3)dog (3') bite (TA (3')-3) inverse theme sign /ekw/ in (4) marks the same 'John's dog bit him (sc. John).' noun as actor. Where the choice of direction is not pre-determined, (3) nisekihandn atim. as it is in the above case, the difference between direct scare (lp-3) dog (3) and inverse sentences which are paraphrases of one 'We scare the dog.' another is clearly a matter of focus and the attending semantic-syntactic emphasis. (4) nisekihikondn atim. Direction is a completely independent category scare(3-lp) dog(3) only if we exclude those forms which involve two 'The dog scares us.' third person referents. In the latter forms, direction is subordinate to focus. Where both referents are third persons, the inter- 2.511. The extensive symmetry of the transitive play of the categories of direction and focus (obviation) animate (TA) paradigm and the reversibility of gives rise to a somewhat more complex situation: direction in many forms are highly reminiscent of in the the (5) sekihew napew atimwa. Indoeuropean languages. However, tempting similarity of the verbal forms must not be scare(3-(3')) man(3) dog(3') allowed to obscure the fundamental differences.35 'The man scares the dog.' very Direction reflects the actor-goal relationship and the "actual" of the sekihik atimwa. (or "logical") relationship (6) napew referents. scare((3')-3) man(3) dog(3') Voice in the Indoeuropean languages, by contrast, 'The dog scares the man.' is primarily a matter of emphasis and stylistics.88 There are, of course, pairs of individual verb forms sekihew napewa atim. (7) which appear to reflect opposite actor-goal relation- scare(3-(3')) man(3') dog(3) e.g., Latin amat 'he loves' vs. amdtur 'he is 'The scares the man.' ships, dog loved'; but as soon as nominal complements are added, the misleading nature of this example becomes (8) sekihik napewa atim. obvious: scare((3')-3) man(3') dog(3) 'The man scares the dog.' puer canem terret. 'The boy scares the dog.' If a pair of sentences shows the same choice in one of these categories, it necessarily differs in the other, canis terretur a puero. and they describe the opposite event34. Sentences 'The dog is scared by the boy.' and are identical with to focus: (5) (6) respect Whatever the differences in emphasis, etc., may be, is atimwa obviative. the ndpew proximate, However, the direction of the action is not affected by the change direction of the verb forms indicates a reversal of the of voice. in the actor-goal relationship: (5) proximate ndpew In fact, if we want to indulge in cross-language is the in it is the obviative atimwa. actor, (6) comparison, the Indoeuropean voice category func- Sentences (5) and (7), on the other hand, are tions in a manner similar to Cree focus (obviation) by identical with to direction: both show a respect providing emphasis, stylistic continuity, etc. actor, i.e., the verb forms are direct. The proximate Since glosses can acquire great importance in reversal of the actor-goal relationship is indicated by linguistics, a practical matter deserves to be empha- the difference in focus assignment: ndpew is proximate in (5), atim is proximate in (7). 85 Further implications of this problem are discussed in 5.664. 86 Sentences (5) and (8), finally, are paraphrases of The fact that certain English verb phrasesseem to be revers- one another. describe the same "actual" event ible is merely a red herring: They I saw John. I was seen John. but differ both in focus and in (a) (b) by assignment direction; For if we consider (b) more closely, it is not the passive of (a) but the same relation holds between sentences (7) and (6). of another sentence, (c): In any such pair the direct sentence, e.g., (5), is the (c) John saw me. The passive of (a), then, is not (b) but a fourth sentence, (d): 4 Word order is irrelevant to the present discussion. (d) John was seen by me. 26 WOLFART: PLAINS CREE [TRANS. AMER. PHIL. SOC. sized here: from a Cree point of view the voice of The semantic status of the indefinite actor forms English glosses is absolutely irrelevant. Whether is difficult to define in terms of the direction category. sentence (4) is glossed 'The dog scares us' or 'We are The examples given above can be described as actor- scared by the dog' has no bearing on the meaning of less verbs which only have a goal; in this they are the Cree sentence.7 somewhat similar to the agent-less passive of the 2.512. The use of "actor" and "goal" rather than Indoeuropean languages, e.g., Latin canis videtur 'the "subject" and "object" is an extension of Bloomfield's dog is seen.' But this interpretation is hardly usage which has become relatively standard for appropriate for the transitive inanimate (TI) and Algonquian linguistics.38 animate intransitive (AI) forms, neither of which The traditional use of the terms "subject" and shows a goal: is based on where "object" Indoeuropean languages T73p18 k-etamihk TI the does not coincide with the subject necessarily 'thus one calls it, thus it is called' actor; for example, in the Latin sentence, canis terretur d puero. T121-2 kzwdniwiw AI 'The dog is scared by the boy.' 'there was going home, one went home.' the goal canis is regarded as the subject by virtue of Finally, there are a number of derivational types, its case, agreement with the verb, etc. If "subject" many of them involving an /ek/-based suffix, whose and "object" were to be used in Algonquian, their meanings have been variously described as "middle" general function would correlate much more closely or "passive"; cf. 6.439 and the example quoted from with focus than with direction. In sentence (6), Bloomfield in 2.53, below. The analysis of such above, for example, while ndpew is the goal, it is formations would go far beyond the scope of the proximate and therefore in focus and might well be present discussion; for a detailed treatment in labeled the "subject." Menomini see Bloomfield 1962: pp. 280-298. A few Like many similar issues, however, this use of examples from Cree follow: kisisow 'he is cooked to "subject" and "object" is held in abeyance until completion,' cf. klsiswew 'he cooks him to completion'; further semantic and syntactic studies indicate a kitimakeyimow 'he feels pitiable,' cf. kitimdkeyimew 'he clear need for these terms. takes pity on him'; kipahikdsow 'he is obstructed,' cf. kipaham 'he closes, obstructs it'; T102p6 .. . .miy- . . 2.52. Further Implications weyihtamwak, miyweyihcikdtew . 'they are glad, it feels good . . '; etc. Among the implications of the direction contrast which remain to be explored more fully, indefinite 2.53. Historical Survey actor forms constitute the most obvious problem. In of Morphologically, the indefinite actor forms of the view the remarkable confusion which has come to surround the term in transitive animate (TA) paradigm cannot be grouped "passive" Algonquian with either the direct or the inverse set. The indf-3 linguistics, it may be of interest to briefly review some of the uses to which it has been In so form of the independent order, e.g., wdpamaw 'he is put. doing, we will also sketch the of the voice seen,' shows the same morphological structure as the history interpre- direct forms, notably the direction marker /a/. But tation of the direction category. Howse uses the term of the various all forms whose goal is a non-third person follow a "passive" (1) derivational which were hinted at different pattern (5.84). They are based on a suffix patterns above; (2) the actor forms which he also calls /ekawi/ whose connection with the inverse marker of indefinite "indeterminate and /ekw/ (cf. 5.422) remains to be clarified; e.g., subjective" (1844: p. 107); (3) some of contrast. e-wapamikawiyan 'I am seen.' of manifestations the direct-inverse The inanimate actor set of the transitive animate In one context (1844: p. 57) he uses "active- as with but (TA) 5.83) are based on the suffix passive" synonymous "direct-inverse"; paradigm (cf. elsewhere he that /ekw/ and its extended form /eko/ and (1844: p. 255) greatly emphasizes clearly belong the distinction to the "double to the inverse set, e.g., nipikiskwdtikon 'it talks to me.' active-passive applies third persons" only: "These in their direct and 37 Whatever stylistic carry-over might be found in translation inverse significations are active and passive, . . ., the would have to be restricted to third-person forms: if it were found other combinations of the pronoun being all expressed that Cree speakers consistently prefer the English passive in trans- actively." lating sentences like (6), this would lend added weight to the hypothesized similarity in function of English voice and Cree It is noteworthy that Howse also sees the direction focus. contrast in the you-and-me forms (1844: pp. 219, 220). 38 Of the choice of terms is a matter course, primarily of conven- Lacombe (1874b) uses "passif" of the mixed forms, tion and Bloomfield indicates as much when he says: "We prefer 'actor' to the term 'subject' which might be misleading ..." i.e., those involving both a third and a non-third (1962: p. 45). person referent. The indefinite actor forms he calls VOL. 63, PT. 5, 1973] NOUN INFLECTION 27

"passif ind6fini." Lacombe does not seem to recog- In the you-and-me set, even though the different nize the direction contrast among the you-and-me agreement of the prefix with actor or goal is made forms. explicit, the parallelism with the direct-inverse con- Hunter uses the terms "direct" and "inverse" trast is not commented upon. (1875: p. 16 et passim); some 250 pages of largely Bloomfield uses the term "passive" to refer to the unglossed paradigms are yet to be fully evaluated. indefinite actor forms; for these and for the inanimate In his sketch of Fox, William Jones develops actor forms, the agreement of the prefix, if any, with essentially the same view as that indicated for La- the object is also specified. But there is no explicit combe. "The use of the proper is mention whatsoever of an "active" to which the confined to an agent in the third person" (1911: "passive" would be opposed. p. 846). Lacombe's "passif indefini" recurs as the It is interesting that in treating derivation Bloom- "indefinite passive" (1911: p. 847). field is very explicit about the meaning of passive Jones also recognizes a middle voice (corresponding reflexives and carefully distinguishes them from the to the derivational patterns mentioned in 2.52): middle reflexives which show considerably more The middle voice represents the subject in close relation variety. In Menomini, with the action of the verb. It is a form of construction of passive reflexives are freely formed . . .; they are ex- which the dialect is especially fond. The form of the common. In verb is and of a intransitive tremely meaning, they border upon passive active, mainly predicative inflectional forms and upon middle reflexive derivatives. character; but the meaning is passive (1911: p. 845). Thus from we-hnew 'he names him,' the inflectional For Algonquian in general, Michelson listed no passive we hnaw means 'his name is spoken; he is men- fewer than five voices: tioned by name; he (say, a newborn child) is given a active, middle, passive, one or several some actor or and These name,' upon occasions, by reflexive, reciprocal (1926: p. 370). actors not specified in the immediate context; the middle "voices" seem to include both inflectional and deriva- reflexive we hsow means 'he bears (such-and-such) a tional patterns; thus, "the last two are formed by name; he gives himself such-and-such a name,' with no special suffixes" and the middle voice apparently is other person involved; the passive reflexives we hcekasow also considered derivational since it is formed with the and we.hcekate-w mean 'he, it is named or called so' by people in general (Bloomfield 1962: p. 282). "instrumental particles." Both types then show the usual endings of intransitive verbs. Hockett's description of Potawatomi in the matter On the other hand, of direction explicitly (1948: p. 141 fn.) follows Bloomfield's treatment of Fox. Since Potawatomi at least two passives are common, one [1] where the agent is either expressed or understood, the other [2] where the has no indefinite actor form of verbs (nor an indefinite agent is not expressed and is indefinite. The pronominal possessor prefix in nouns; see Hockett, 1966: pp. elements of the last, in the case of the independent mode, 63, the term occurs in the context are allied to 64), "passive" only partially the ordinary intransitive verbal of derivation In Bloomfield's pronouns. Other passives [3] apparently exist, but their (1948: p. 67). prefacing exact function is not accurately known. One appears to posthumous grammars, Hockett argues against the be very indefinite and to occur only with an indefinite term even in its limited application to the indefinite subject. actor forms: "Algonquian 'passives' are not like those Leaving aside this last, "very indefinite" passive, we of Latin or Greek; rather, they are special inflected can clearly identify [2] with our indefinite actor forms for indefinite actor, showing the same inflectional forms. [1] apparently refers to the inverse forms, at indication of object shown by other inflected forms least those involving both third and non-third of the same kind of verb" (in Bloomfield, 1958: p. vi). referents. In fact, Michelson seems to be somewhat Hockett also gives much weight to the syntactic hesitant when he says: "The forms of the independent parallelism of the indefinite actor forms and the mood with the third person animate . . . as subjects indefinite possessor form of nouns (1966: p. 64; in and the first and second person . . . as objects are Bloomfield, 1962: p. ix). really passives in construction" (emphasis supplied). It might finally be noted that Voegelin in his brief In describing the inflectional morphology of Fox, sketch of Delaware apparently regards the "direct- Eastern Ojibwa, Menomini, and of Proto Algonquian inverse" pair as synonymous with "active" and as well, Bloomfield strenuously avoided any reference "passive" (1946: p. 145). to a voice contrast in the transitive animate paradigm. Instead, he used the terms "direct" and "inverse" 3. NOUN INFLECTION which he defines as follows (1962: p. 141): Nouns are inflected in two separate paradigms Directforms. The first or second person acts upon a third which are realized in different layers of affixation. person,or a proximatethird personacts upon an obviative. The affixes of the POSSESSION paradigm (3.2) which If there is a prefix, accordingly, it agrees with the actor: with nouns is constitute an inner of . . .Inverse forms. The third person acts upon the first optional, layer or second person, or an obviative third person acts upon affixation. The NUMBER-OBVIATION paradigm (3.3) a proximate third person. If there is a prefix,accordingly, indicates the categories within which the noun itself it agrees with the object: . . . functions in the system of anaphoric reference. The 28 WOLFART: PLAINS CREE [TRANS. AMER. PHIL. SOC. number-obviation paradigm and a few other suffixes 3.2 THE POSSESSIVE PARADIGM form an outer of affixation. (3.6, 3.7) layer The possessive paradigm provides anaphoric refer- Nouns are of either animate or gender, inanimate. ence to a person other than that denoted by the noun Simple noun stems end in a non-syllabic or a cluster itself. Possession in a narrow sense is, of course, only of non-syllabics, e.g. maskisin- 'shoe,' ihkw- 'louse.' the primary or focal meaning of this morpheme class The usual citation form of nouns is not the stem but (and the label "possessive" is chosen just because of the inflected form for proximate singular which is its concreteness).4 Besides this narrowly possessive identical with the stem except where the latter is function, as in nitem 'my horse,' there are other, more monosyllabic; see 3.31. attenuated meanings, as in ninehiyawewin 'my Cree- ness, my speaking Cree,' or in otocikaniwdw 'their 3.1. AFFIX POSITION CLASSES doing'; (cf. the use of my in my motheror in my going to York The position classes of the nominal affixes corre- New tomorrow). noun stems show a spond closely, as far as applicable, to those of the While certain special possessed theme the occur verbal affixes; cf. 5.4. The present section may also (see 3.21), possessive paradigm may with stem. some serve as an index of morphemes. any noun Conversely, however, noun stems are inflected for The prefixes are described in 2.1; but see also 3.22, obligatorily possession; these bound stems are called DEPENDENT noun stems. for the indefinite possessor prefix mi-. especially nouns include kin and terms The suffix classes and their order are Dependent mostly terms, position for and a few summarized below. The brief labels used in this body-parts personal possessions.42 In general, the categories of the possessive paradigm list are intended as approximations only. are independent of those of the number-obviation 1 possessive theme sign paradigm. Thus, a stem inflected for a first person 2 thematic obviative sign possessor may be either proximate, e.g. nisit 'my foot,' 3 /epan/ 'former, absent' nisita 'my feet,' or obviative, as nitema in ndpew 4 possessive person suffixes wdpamew nitema. 'The man sees my dog.' If the 5 third person, locative, and vocative suffixes possessor is a third person, however, the usual rules 6 plural and obviation (animate) suffixes of obviation operate. The possessor is always nearer (more in focus) than who or what is possessed, so that theme Suffix position 1 is occupied by the possessive any third person possessor automatically causes the sign /em/ (3.21).39 noun itself to be obviative. The obviative sign of position 2, /eyi/, marks an obviative possessor (3.22).40 3.21. Theme Formation The suffix /epan/ 'former, absent' is tentatively Possessed themes are formed with the theme to suffix position 3; see 3.5. sign assigned which follows the stem. In suffix 4 there the /em/ immediately However, position appear personal the formation of themes is to a suffixes of the possessed subject possessive paradigm (3.22). deal of which further 5 the third great irregularity requires study. In suffix position there appear person The absence of the theme is of the special sign typical suffixes of the number-obviation paradigm; stems in n is no means restricted to animate suffix is the inanimate suffixes ending (but by /wa/-/a/, these). It is found, for example, with a number of are and /wah/-/ah/ (3.31). /wi/-/i/ noun types derived from verbs, such as the abstract The locative and vocative markers are (3.6) (3.7) nouns of 6.41; e.g., nipimttisiwin 'my life,' otocikan exclusive with the of mutually morphemes position 'his doing, fault.' Possessive forms without special 5 and 6. theme sign are also common with dependent stems, In position 6 there appear the animate plural and e.g., nistes 'my older sibling'; but contrast nisim obviative markers of the number-obviation paradigm, 'my younger sibling,' nitotem 'my kinsman,' etc. and namely /k/ /h/ (3.31). Other than that, the distribution of /em/ cannot even be the obvious 39 Note the homonymy of /em/ with the thematic obviative tentatively indicated; hypothesis sign of verbs; cf. 5.41. of an alienable: inalienable category has been explored 40Although in a strict positional analysis positions 2 and 4 without success. sSsip 'duck': nislsipim be are here for reasons of overall Examples: might merged, they kept apart 'fire': nitiskotem ihkw- patterning; cf. 5.4 and 5.43. 'my duck'; iskotew 'my fire'; In eastern dialects the correspondingsuffix marks the obviative 'louse': nitihkom 'my louse'; etc. in the number-obviationparadigm of inanimate nouns (see Ellis, As an of the erratic distribution 1962: 3-20 and 8-13 for and also example seemingly p. especially p. James Bay Cree; from Rogers, 1960: pp. 110, 112 for the Mistassini dialect of Mon- of /em/ consider nitaskiy 'my country,' askiy, tagnais-Naskapi); however, since the data are incomplete and in contrast to nipimsm 'my lard,' from pimiy. lack certain crucial examples (e.g., "he saw our canoe(s)" and "his son saw our canoe(s)") the relevanceof this evidence cannot 41 Hockett (e.g., 1966) uses the term "allocation." be assessed and the issue must be left open. 42 There is no evidence for an alienable: inalienable dichotomy. VOL. 63, PT. 5, 1973] NOUN INFLECTION 29

Some stems may even occur both with and without TABLE 2 /em/. ayisiyiniw 'human being' has been recorded POSSESSIVE PARADIGM: INANIMATE DEPENDENT both ways from the same speaker: T91-7 wic-dyZsiyinT- NOUN /-TEH-/ 'HEART' wdwa 'their fellow-people' and T49-5 otayisiyinzma 'his people.' 1 /niteh-/ 'my heart' A clue to the be contained in 2 /kiteh-/ 'your heart' /em/-problem may 3 'his heart' an rare form: the stem /oteh-/ extremely While dependent 3' /oteheyi-/ 'his (3') heart' -stikwdn- does not normally take /em/, e.g., ostikwtdn indf /miteh-/ 'a heart' 'his head, a head,' a secondarily possessed form does lp /nitehenan-/ 'our heart' in fact show the theme suffix: S150-20 ndstikwdnim 21 /kitehenaw-/ 'our heart' head' of a severed head which is used as 2p /kitehewaw-/ 'your heart' 'my (said 3p /otehewaw-/ 'their heart' lodge emblem). 3.22. Inflection nouns, this category is nevertheless present as shown The inflectional affixes of the possessive paradigm by concord with verb forms, e.g. S12-46 ekosi osihtdw correspond closely to those found in verb inflection. eh-misdyik osi. 'Thus he built a great canoe,' where The personal prefixes are described in detail in 2.1. osi 'canoe' is covertly obviative as shown by the ki-, ni-, o- occur with both nouns and verbs. mi- obviative ending of the verb eh-misdyik 'it (0') is 43 occurs only with dependent noun stems. It marks big.' an indefinite possessor, e.g., micihciy 'a hand, some- one's hand'; T34p7 mistikwdna kd-wdpahtaman . . . 3.31. Suffixes 'heads I saw ... The number-obviation ("third person") suffixes The thematic suffix /eyi/ (cf. 5.43) immediately have an alternant with initial /w/ and one without. follows the stem or the possessed theme; it marks an The distribution of these alternants is discussed obviative possessor, e.g., in 3.32. S 36-41 ositiyiwah pikoh ohi The animate third-person suffix is /wa/-/a/. It their(3') feet(0p) only these (3') is followed by the plural marker /k/ or the obviative marker /h/. Stem /sisip-/ 'duck': niskah ekwah sisipah, ekonih pikoh geese(3') and ducks(3') these(3') only 3 /sisip-a/ 3p /sisip-a-k/ iskotehk astdw; 3' /sisip-a-h/ in the fire place (AI 3) The inanimate third-person markers are /wi/ /i/44 the feet of those and those 'Only geese ducks, only in the singular and /wah/-/ah/ for the plural. did he put into the embers'; Stem /maskisin-/ 'shoe':

P 98-38 ohi napewah kd-nipahayit 0 /maskisin-i/ that(3') man (3') kill(TA 3'-(3')) Op /maskisin-ah/ Final in owzkimakaniyiwah itohtahew. which are found our morpho- his(3') wife(3') take there(TA 3-(3')) phonological representation are subject to (appendix A: 5.1). 'He took to that place that man who had slain his The final vowel remains in nouns whose stem is (own) wife'; cf. also 2.221. monosyllabic; thus, we find animate nouns like niska The possessor is pluralized by one of the following 'goose' and inanimate nouns like mihti 'firewood' or suffixes (which recur in the independent order of the verb): /enan/ if it involves the first but not the second 43 In other dialects, e.g., in James Bay Cree, and at least in the Mistassini dialect of inanimate nouns are /enaw/ if it involves both; and /ewaw/ Montagnais-Naskapi, person; inflected for obviation by means of the thematic sign /eyi/; cf. otherwise. (Cf. also 5.451 and 5.48.) Ellis, 1962: p. 3-20 and Rogers, 1960: p. 110. Table 2 shows the possessive paradigm only. 44The suffix is set up as /i/ rather than /e/ primarilyfor com- parative reasons; the fact that it does not take part in 3.3. THE NUMBER-OBVIATION PARADIGM (see appendixA: 4.5) is not consideredconclusive since exemption from this rule can be attributed to the special status of mono- Animate nouns have inflectional endings for proxi- syllabic stems; only three instances are available: wtiwi 'egg', mate singular and plural, and for obviative which is osk-dyi 'young creature,' and meyi 'dung.' number-indifferent. The absence of palatalizationin wati 'hole,' wzsti 'beaverlodge,' and mihti 'firewood' to be a matter of level- inflectional for appears paradigmatic Inanimate nouns have endings ing. (The dependent stem /-iwaO-/, as in niwas (sg), niwata singular and plural. While in Plains Cree there is no (pl) 'my sacred pack' seems to be the only stem to retain inflectional distinction for obviation in inanimate this alternation.) 30 WOLFART: PLAINS CREE [TRANS. AMER. PHIL. SOC. mihko 'blood' (stem mihkw-).45 A form like kona clearly loans from English, e.g., omamawa 'his mother,' 'snow' shows that the retention of the final vowel is from nimdma; meriwa 'Mary' from meriy; cimiwa not restricted to monosyllabic stems with a short 'Jimmy' from cimiy, etc. vowel (cf. appendix A: 5.1). 3.311. The addition of a prefix does not affect the 3.32. /w/- Alternation retention of the suffixal vowel, e.g., otosi 'his canoe' The number-obviation person") suffixes of or omihko 'his blood.' One interpretation of these ("third nouns appear with two sets of alternants. One has forms would require apocope of the suffixal vowel initial /w/, namely /wa, wi, wah/, the other lacks it, to precede the addition of the prefix (both synchronic- and this would conflict with the namely /a, i, ah/. ally historically); the distribution of the alternants is fact that the affixes of the are Generally, possession paradigm the after a closer to the stem than those of the number-obviation governed by preceding environment; the alternant after a vowel kitosinaw 'our canoe.' It seems more /w/-less occurs, paradigm, e.g., that with For consider the forms that stems without /w/. instance, likely monosyllabic possessive nitemak otema 'his but suffixes are from the rule due to the 'my horses,' horse,' etc., exempt apocope 'his horse.' 5.471 for the similar of the in the otemiyiwa (3') Cf. pattern pressure simple forms; fact, situation in the transitive animate verb historical sources show some fluctuation. paradigm. Clearly, A number of noun stems end in a cluster the whole issue of stems large monosyllabic requires Since the occurs not in the 3 and further /Cw/. /w/ only 3p investigation. forms but in the 3' form and in the locative or vocative The suffixal vowel of stems is retained monosyllabic forms as the to the stem even when function as the second member of a well, /w/ clearly belongs they and does not take in the deri- 'bear this is part any alternation; compound, e.g., wdkayosi-wdti den'; vational structure of the stem does not bear on the seen as evidence of the independent phonological at hand. mistikw-: mistik status of members problem Thus, 'tree,' compound (cj. 6.5). mistikwak with suffix the 'trees'; /eyiwa/: omistikoyiwa 3.312. Phonemically, proximate singular and 'his with suffix mistikohk 'on a stems (3') tree'; /ehk/: tree'; the obviative of monosyllabic animate are or consider the stem atimw-: atim atimwak is re- 'dog,' homophonous. Occasionally, this homophony with suffix atimotik of to the obviative 'dogs'; /etik/: 'you dogs!' solved by the addition another -wa In nouns in the identification of the form in -wa. in 49 we find ending /Vw/, ending Thus, T523p47, 48, is less clear-cut.46 Consider the noun ndpew maskwawa the obviative of /w/ both maskwa and as 'man' whose other inflected forms are and An -wa suffix also occurs in the ndpewak maskwa 'bear.' extra its stem could be set as either or obviative of certain other words most of which are ndpewa; up ndpe- ndpew-. The formation of the possessed theme fails 46 The following list of monosyllabic noun stems appears to be to throw light on the problem because ninapem might relatively complete: be formed from either stem; the contraction of Animate: /ew-e/ to /e/ occurs independently, e.g., in the -4w- 'wife' (dependentstem) inflection of the verb stem wicew- 'have him along': -Iskw- 'fellow wife' (dependentstem) niwicek 'he has me along,' kiwicetin 'I have you cts- 'Cisa' (name of trickster) along,' etc.; cf. appendix A: 4.2. The relevant mor- es- 'clam-shell' statements are all ihkw- 'louse' phophonological independently a kakw- 'porcupine' motivated and thus do not indicate solution. kon- 'snow' The existence of parallel stems in /Vy/, e.g., askiy maskw- 'bear' 'land,' apasoy 'tent-pole,' supports the analysis of the mdkw-,mwakw- 'loon' as part of the stem. But the evidence is mosw- 'moose' not and a full treatment of noun stems nisk- 'goose' conclusive, piskw- 'mosquito-hawk' ending in a vowel-semivowel sequence will have to siht- 'evergreen' await further synchronic and comparative studies.

Inanimate: 3.4. PARADIGM TABLES -ik- 'dwelling' (dependent stem) mey- 'dung' The tables show the possessive and number- mihkw- 'blood' obviation paradigms combined. For an explanation miht- 'firewood' of the blank in the first table see 3.2 and 2.21. 6s- positions 'canoe' in pihkw- 'ashes' The paradigms are given phonemic representa- wdt- 'hole' tion; cf. Appendix A, especially fn. 85. waw- 'egg' wlkw- 'kidney-fat' 46A parallel situation exists in Menomini where there are at win- 'marrow' least two derivational morphemes /w/ one of which "is homony- wls- 'belly-fat' mous with inflectional -3w and demands the same replacements wust- 'beaver-lodge' of precedingvowels" (Bloomfield, 1962: p. 242). VOL. 63, PT. 5, 1973] NOUN INFLECTION 31

TABLE 3 TABLE 4

ANIMATE NOUN, DEPENDENT STEM /-TEM-/ 'HORSE, DOG' INANIMATE NOUN, STEM /MASKISIN-/ 'MOCCASIN'

3 3p 3' O Op

nitem -ak -a 1 my horse, etc. nimaskisin -a 1 my shoe, etc. kitem -ak -a 2 your horse kimaskisin -a 2 your shoe otema 3 his horse omaskisin -a 3 his shoe ~- - otemiyiwa 3' his (3') horse omaskisiniyiw -a 3' his (3') shoe nitemindn -ak -a lp our horse nimaskisinindn -a lp our shoe kiteminaw -ak -a 21 our horse kimaskisininaw -a 21 our shoe kitemiwaw -ak -a 2p your horse kimaskisiniwdw -a 2p your shoe - otmiwdwa 3p their horse omaskisiniwdw -a 3p their shoe

portant, Hockett describes the "preterit" suffix as 3.5. THE /EPAN/ SUFFIX following the suffixes of the number-obviation The suffix /epan/ 'former, absent' indicates that paradigm. the denotatum of the noun no longer exists.47 It 3.6. LOCATIVE also occurs in the pronoun awwnipan 'nobody' (4.123) There are two and (in a slightly variant shape) with verbs (5.321). locatives, simple (3.61) and distribu- tive The locative Examples: kiseyiniw 'old man,' kiseyinipan 'old (3.62). suffixes are mutually ex- man no longer alive'; nimosom 'my grandfather,' clusive with the suffixes of the number-obviation nimosomipan 'my late grandfather.' paradigm. Thus, number and obviation are not /epan/ precedes the number-obviation suffixes as expressed in locative forms; obviation would also be excluded on semantic well as the possessive suffixes, e.g. nimosomipanak grounds. (But note that the 'my late grandfathers,' nohkomipanindnak 'our late obviation status of a possessor is not affected.) grandmothers.' Neither the present data nor La- Examples: niskdt, niskdta 'my leg, my legs': niskdtihk combe's discussion (1874b, pp. 18, 19) establish the 'on my leg(s)'; osklsikwa 'his (3) eyes': osklsikohk position of /epan/ relative to that of the thematic 'on his eye(s)'. obviative sign /eyi/. 3.61. Simple My informants reject /epan/ added to inanimate nouns while Lacombe (loc. cit.) gives an entire in- The simple locative suffix /ehk/ 'at, in, on, etc.' is animate paradigm without even mentioning the used with stems or possessed themes. problem. sdkahikan- 'lake' sakahikanihk 'at the lake' It is curious that no corresponding suffix is reported 'ashes' pihkohk 'in the ashes' for Fox, Ojibwa, Menomini, and Bloom- pihkw- Kickapoo by 'his boat' 'in his boat' field 1958, or Voorhis otosiyi- (3') otosiyihk (1924, 1962) (1967). Only kikinaw- 'our Potawatomi seems to have what Hockett considers (21) home' kZkindhk'at our home' wikiyi- 'his (3') home' 'at his home' a "preterital suffix" (1948: pp. 8, 73; 1958: p. 238). wikiyihk Delaware also has a noun but "preterite paradigm" 3.62. Distributive with different suffixes (see Voegelin, 1946: p. 144).48 The Potawatomi situation differs from that in Cree The distributive locative suffix /enahk/ is used in three ways: (1) The "preterit" suffix occurs with with nouns which denote humans or animals. The inanimate stems. (2) It occurs with possessed resulting forms mean 'in the land of such-and-such themes only; note that according to Lacombe Cree beings,' 'at the place of such-and-such people.' /epan/ "ordinairement" occurs with possessed themes, /enahk/ has not been recorded with possessed themes. and only secondarily with all nouns. (3) Most im- 'human hu- 47 ayisiyiniw- being' ayisiyinnadhk 'among The status of a nominal dubitative remains doubtful owing to mans, in this world' the lack of adequate data. Only one clear form has been re- 'Plains 'in corded (in Bloomfield's paskwdwiyiniw- paskwawiyinmndhk the syllabary texts, MS(a): p. 1138): Cree' ayisiyinitokenik 'what people, I wonder.' Cf. section 4.21, fn. Plains Cree country' 51, and especially 5.311. mostosw- 'buffalo' mostosondhk 'in the buf- 48 Nominal tense markers are not restricted to Algonquian; for falo country' an Athapaskan example consider Hupa: "By the use of suffixes sdsiw- 'Sarci Indian' sdstndhk 'at Sarci the time of the noun's existence may be indicated. This process Reserve' practically gives tenses to nouns. For the past, -neen is em- ployed: for example, xo2tneen 'his wife used to be' (she is now kihci-mohkomdn- 'Big kihci-mohkomdnindhk'in dead). The same form might mean only that the possession of Knife, American' the USA' her had ceased" (Goddard, 1911: p. 110). But while Cree shows /epan/ followed by other suffixes, this does not seem to be the The suffix /enahk/ is based on the locative suffix case in Athapaskan (where Hoijer speaks of 'enclitics'). /ehk/ preceded by a noun-final /enaw/; (for the 32 WOLFART: PLAINS CREE [TRANS. AMER. PHIL. SOC. contraction cf. appendix A: 4.2). Nouns in /enaw/ Vocative identical to proximate singular: are derived from inanimate intransitive (II) verbs in nipdpa 'my father' /ena/, with their most common inflected forms being nimamd 'my mother' homonymous; e.g., ispatinaw II 0 'it is a hill,' or the niscds 'my male cross-cousin' noun 'hill.' Since these nouns and verbs typically nitotem 'my fellow tribesman' denote topographical features, the semantic connection with the /enahk/ suffix is obvious. However, none 3.72. Plural of the nouns which take /enahk/ as a distributive locative suffix are actually paralleled by a verb in The vocative plural is always formed with /etik/: a form like mostosondw 'it is buffalo /ena/ (although nisim brother' nislmitik well be while the 'my younger country' might expected). Thus, nitoskinikim nitoskinikimitik forms must be as derived 'my young man, /enahk/ ultimately regarded crew member' nouns, for all practical purposes /enahk/ functions as a somewhat restricted inflectional suffix. dtayohkan 'spirit guardian' dtayohkanitik

3.7. VOCATIVE 3.8. "QUASI-NOUNS" The vocative is formed in a of singular variety Some otherwise non-paradigmatic forms, i.e. while is used for the ways /etik/ uniformly plural. particles, take the vocative plural (3.72) and simple locative suffixes in a few even the 3.71. Singular (3.61) and, cases, plural marker /k/ (3.31, 5.481). Syntactically, these no is used for the vocative Normally special ending are predicative particles but because they make use of 20 nitokimam 'oh singular, e.g., T7p5, my king.' nominal suffixes they are, for the nonce, called most of the kin terms and a few nouns of However, QUASI-NOUNS. intimate possession have vocative forms which are still largely used; the younger generation uses these 3.81. Locative in free variation with the simple proximate singular form (not the stem; but cf. Lacombe, 1874b: p. 6). The locative suffix /ehk/ occurs freely but not The vocative singular frequently ends in a long very frequently, e.g., dstam 'here, come here': dstamihk vowel; this fact may well be related to phenomena 'on this side.' The same meaning is usually expressed of rhetorical distortion. While some kin terms remain by the locative morpheme ita (which also occurs as a unaffected, others undergo apocope (the loss of final particle by itself), e.g., dstamita 'on this side, closer.' , vowel-consonant, sequences, or whole ) or add a suffix -e; vowels are often length- 3.82. Vocative ened or distorted. The following lists are by no means exhaustive; the glosses are incomplete. More typical and more frequent are particles which take the vocative plural suffix /etik/. Apocope: Extremely frequent are dstam 'come here,' awas 'go nohtawiy 'my father' nohtd away'; when more than one person is addressed, we nikdwiy 'my mother' nekd find dstamitik, awasitik. nimosom 'my grandfather' nimoso A less common example is the particle mdcikotitdn, nohkom 'my grandmother' nohko mdcikocicdn 'look, let me show you' whose internal nitihkwatim 'my cross-nephew, nitehkwd structure is obscure. Although a "plural" form son-in-law' mdcikotitak does occur in Bloomfield's texts, mdciko- nitawemaw 'my cross-sibling' nitawemd titan may be addressed to one or several persons. In nitdnis 'my daughter' nitan the narrow use, when only one person is addressed, niciwdm 'my male parallel cousin' niciwa our texts oppose mdcikocicdn to a "vocative plural" nikwemes 'my namesake, friend' nikweme mdcikocicdnitik. Suffix -e: A more problematic form is ekotik 'let's go' nimis 'my elder sister' nimise (T523p86) which seems to be based on the pronominal nisikos 'my father's sister, nisikose stem eyakw--e-kw- (4.41). mother-in-law' nisis 'my mother's brother, nisise 3.83. Plural father-in-law' nisim 'my younger brother' nisime A few particles have forms with and without final nistes 'my older brother' nistese /k/. Whatever the etymological origin of the /k/, nicdhkos 'my (f) sister-in-law' nicahkose the forms with /k/ seem to be interpretable as plural: Apocope and suffix -e: S326-3 "niya, nisim, sipwehte!" 'Go, little sister, nikosis 'my son' nikose depart! (AI 2)'. S69-23 "niyak! mdcik!" 'be off, nosisim 'my grandchild' nosise hunt! (AI 2p).' VOL. 63, PT. 5, 1973] PRONOUN INFLECTION 33

4. PRONOUN INFLECTION awa has the stem alternant aw- in the proximate and the alternant o- elsewhere. ana exhibits There are three of singular, major paradigms pronominal the stem an- in the same Like inflection as well as a few which are found with shape throughout. only the naha has the stem alternant one stem each. This inflectional classification of aw--o-, pronoun nah- in the and the alternant ne- coincides with classifications proximate singular, pronouns only partially elsewhere.0 based on syntactic or semantic criteria.481 Further- more, most of the pronominal stems not only select 3 awa ana naha inflectional paradigms but are also subject to a host 3p oki aniki neki of derivational processes (cf. 6.422). 3' ohi anihi nehi The various pronominal paradigms share a feature 0 oma anima nema which is of great interest to the understanding of the Op ohi anihi nehi grammatical categories of Cree (cf. 2.02). In spite of great differences in phonemic shape between the 4.111. The demonstrative pronouns awa, ana, and different paradigms, they all, without exception, naha clearly constitute a semantic field; its internal exhibit an identity in form of the animate obviative structure, however, can only tentatively be indicated. and the inanimate plural forms. They seem to be ordered in such a way that awa is One specifically pronominal paradigm (I) is found "nearest" to a point of reference and naha "farthest with the demonstratives awa, ana, and naha (4.11); away" from it. The exact nature of the point of with awina 'who' (4.12); and tdni 'which one' (4.13). reference remains to be discovered. A second pronominal paradigm (II) is found with The relation among the demonstratives does not taniwa 'where is he' and ewakwd 'there he is' (4.21) as seem to correlate, as one might expect in a three-term well as with oya (4.22). system, with proximity to speaker, addressee, or neither. the demonstratives be visual- The usual number-obviation paradigm of nouns Instead, may ized as in a linear The (in some cases with slight modifications) is found with arranged sequence. "distance" between awa and ana seems to be kikway 'what; what sort' (4.31) and kikway 'some- the same as that thing' (4.32); it also occurs with kotak 'another' between ana and naha. Informant responses indicate (4.33) and aya '. . . one; person; thing' (4.34). that awa refers to someone in reach, ana to someone not in and naha to someone far ewako 'the selfsame' (4.41) and awiyak 'someone' reach, quite away. There is evidence that naha is almost (4.42) each show isolated paradigms. strong always accompanied by pointing, both actually and figura- The personal pronouns (4.5) are not inflected for tively. number and obviation. As a set, however, they When of the relation between demonstra- closely parallel the possessive paradigm of nouns speaking we use a in the systematic use of personal prefixes and suffixes. tives, spatial metaphor ("distance," "near," "far"). Further investigation is expected to show 4.1. PRONOMINAL PARADIGM I that this relation is indeed essentially spatial. But until such evidence is forthcoming, the arbitrariness Pronominal paradigm I has the following basic set of the metaphor should not be lost sight of. of endings: 4.112. awa, ana, and naha are also very similar in 3 -a syntactic function. All three function as modifier or 3p -ki as verbal complement, e.g., T1-4 mistikwa ohi 'these 3' -hi trees'; T4p15 ana moniydw 'that White Man'; 0 -ma T104p8 nehi ondpemiyiwa 'those lovers of hers.' Op -hi T33-4 mdmdkwaht oma! 'chew this!'; T73pl7 kakes- kimdw anima 'he was told (indf-3) that'; T131-4 4.11. awa, ana, naha 'this, that, that yonder' kiwdpamaw ci naha? 'Do you see that one yonder?' All three frequently combine with ewako (4.41). The demonstrative pronouns awa 'this,' ana 'that,' However, only awa and ana seem to occur as and naha 'that yonder' inflect exactly alike.49 predicates, e.g., T10p25 tans oma mdna nipdpa 48a For example, while the following four pronouns are clearly k-etwet. 'How is it my father says'; T10p15 Saskat- correlated semantically and derivationally, they reflect three chewan an ohci. 'This one is from (ohci) Saskat- different inflectional classes: chewan'; T60pl0 ekwa anima k-J-nanapotokanecik oki awina 'who' kikway 'what' sihkihpak. 'And that (0) is why these hell-divers (3p) awiyak 'someone' kikway 'something' are crooked at their rump.' For further details see sections 4.12, 4.3, and 4.42. 49awa also shows a 3'/0p form ohoand a 3p form oko. But in a 56The relation between the two pairs of alternants remains sample of 112,000 words of running text, and leaving aside the obscure; while it is tempting to speculate about an underlying sandhi forms oh and ok, ohi is almost fifty times as frequent as form /aw-eki/ for oki, /nah-eki/ for neki, etc., little will be gained o5h, and oki ninty times more common than oko. by such an analysis until similar contractions are found elsewhere. 34 WOLFART: PLAINS CREE [TRANS. AMER. PHIL. SOC.

4.12. awina 'who' kikway, awtnipan oma mikiwdhpis. '. . . when he looked around, nothing, that wigwam (0) was gone!' awtna 'who' has only animate forms: 4.13. tdni 'which' 3 awmna 3p awiniki tdn- is a delimiting interrogative. 3' awinihi 3 tini 4.121. awina has two distinct but clearly related 3p tdniki uses. As an interrogative pronoun, awina shows 3' tdnihi concord with the other nominal or pronominal expres- 0 tdni, tdnima sions in a sentence. Having animate forms only, it is Opop tanihi in this function the inanimate complemented by The inanimate singular form is problematic; the interrogative kikway (4.31). variant one would expect in this paradigm is tdnima. awmnamay occur by itself, e.g. T10p89 eha, awmna? tdnima is homonymous with a particle meaning 'how I 'Yes, who?'; T10p7 awTn etokwe. "Who, wonder'; much, where' (which presumably developed from the 8 kanihk. 'Who are then?' T520p7, awiniki they, pronoun), and this may well be the reason for the P72-4 "kitdnisindw eh-witapimdtndpewa." "awinihi?" spreading of tdni. ' "Our is the side itwew kiseyiniw. daughter sitting by Just like the numerous particles based on the root of a man (3')." "Who is he," said the old man.' tan-, the delimiting interrogative tan- functions It may occur as part of an equational sentence, e.g., primarily as a conjunction, e.g., T28p3 mdnakisk T27p2, 3 awina naha nete. 'Who is that one yonder?' tdnihi ekoni e-wiyinoyit e-ati-pimikwepitdt . . . awina ana naha kd-pe-sdkewet. 'Who is that, that 'Then, which ever ones (3') were fat (3'), of those one yonder, coming into the open?' (3') he twisted the neck (3-(3')) . . . .' S280-8 Or it may function predicatively with a conjunct tdnimah kostahkih, otah k-esiwepinanaw. 'Whichever clause depending on it, e.g., T10pl2 awina ekosi he fears, there we shall throw him.' tdni also occurs e-itwet. 'Who says so?' T54p3 awina kd-nakatiht. with the pronoun ana as predication: P200-31 tdn ana 'Who was left behind (indf-3)?' mdka eh-okimdwit . 'But which is the one who 4.122. In combination with a following demonstra- is the chief . .'; P280-1 tdn dnim dyi okimdw tive, awina expresses surprise, e.g., T504p2 awana wkkih? 'Which is the chief's tent?' tdni is typically ana . . . 'Who (was it but) that one ... . counterbalanced by the delimiting demonstrative In this function, awina is not usually inflected (but ewako (4.41) or by a particle based on the same root, see P98-5); thus, we find it with any inflected form, e.g., ekota 'there': S103-24 ekwah tdnihi eh-wiyinoyit, of either gender, of the demonstratives; e.g., T73p15 ewakonih tehtapiw awa mistanask. 'Then whichever awmn es ohi (3') ekota owLkimdkana (3') ki-apiyiwa one was the fattest on this one Badger sat down'; (3'), . . . 'What was this (3'), his wife (3') sat (3') P222-32 tdn ewakw etoke mdka, ntdtem? 'But which there, . . .'; T103p4 aspin esa awina oma watihk. one is it, Fellow-tribesman?' S129-46 ekwah tdnihi 'Away (she went), lo and behold, into a hole.' eh-miywdsiniyikih wiyasah, ekotah pdnahikeyiwah Where it expresses surprise, awina exactly parallels tawdsima. 'Then by the best stores of meat, there his the particle pdti, e.g., T46p6 pot dhi (3') esa owikimd- children cleared away the snow.' kana (3') . . . 'What was that (3') but her husband (3') . . ' 4.2. PRONOMINALPARADIGM II 4.123. The stem awin- also occurs with the suffix The pronominal paradigm II has the following 'former, absent' which in this environment is /epan/ basic set of endings. mutually exclusive with the number-obviation end- ings; cf. 3.5, 5.321. Thus, T115p8 awtnipan ocahpih- 3 -a cisa 'gone was his tobacco-pouch (3').' 3p -ehkdk Much like its counterpart nama klkway (4.323), 3' -ehd awmnipanmost typically means 'not here any more'; 0 -e in this form, too, there is clearly an element of surprise Op (cf. 4.122 above). Thus, T18-8 wiyapaniyik e-kosko- The pronouns of this type are not common in texts; payit, awmnipan otema (3'). 'When he got up the only the tdniwd paradigm is fully exemplified in next morning, his horse (3') was gone'; T28p10 recently collected texts. e-apasdpit, awlnipan. 'When he looked back: nobody.' 4.21. tdniwd, ewakwd 'where is he, there he is' awinipan may even occur in collocation with an inanimate noun and, indeed, together with kikway: tdniwd 'where is he' and ewakwd 'there he is' are T125-6 awmnipan oma mikiwdhpis. 'That wigwam verb-substitutes which might well be called "ex- (0) was gone!' T125-8 ... -nandtawdpit, awinipan istential" pronouns. They function as predications VOL. 63, PT. 5, 19731 PRONOUN INFLECTION 35 and may be complemented by nominal expressions, for the insertion of y between long vowels see appendix e.g., S8-42 d, nitoydkan, tdniwehkdk? 'Come, my A: 3.1. dish, where are they (3p)?'; S89-42 ewakwd wdwds- 4.3. PRONOUNS WITH NOUN ENDINGS kesiw . . . 'there goes that elk .. .'; T18pll ..., taniwe omaskipayiwiniyiw? '. . ., where was his Several pronouns select the number-obviation para- (3') limping (0)?'; S179-39 nohkj, ewakwe kipisdkan- digm of nouns which is here given morphophonologic- dpiml 'Grandmother, here is your rawhide thong!' ally; except with monosyllabic stems, final vowels are T35pl tdnjwehd kotaka kitatdwewina? 'Where are deleted (cf. appendix A: 5.1). your other groceries (purchases, Op)?' taniwda5 and ewakwd are based on the stems 3 /a/ tan(iw)-(?) and ewakw- which recur in a host of 3p /ak/ 3' derivations as well as in the pronouns tani (4.13) and /ah/ 0 ewako (4.41). /i/ Op /ah/ 3 taniwa ewakwa kotak and aya (4.34) also show the locative 3p taniwehkdk (4.33) 3' taniwehd suffix /ehk/. and the in- 0 tdniwe ewakwe The interrogative pronoun kikway definite are related etymo- Op taniweha pronoun kZkway obviously logically.52 They differ primarily in their syntactic The distinction of 3 tdniwa and 3' tdniwehd is and semantic function (but also in their inflectional sometimes neutralized in favor of the former (cf. paradigms since the indefinite pronoun shows no 2.23); e.g. tdniwa etokwe omdamwdwa. 'Where is (3) 3 form).53 their mother I wonder.' (3p) (3'), 4.31. kikway 'what' 4.22. 6yd 'that no longer here' kikway 'what sort; what': The classification of oya with this paradigm is 3 kikwaya highly tentative. It is based on the final long a and 3p kikwayak the fact that Bloomfield gives the full set of forms in 3' kikwaya his lexicon (ms(b)). However, only the animate 0 kikway, kikways proximate form oyd is textually attested. Op kzkwaya In Bloomfield cites a form [sic] as Language, oya (Because of the neighboring y, the 0 form kikwayi is the third member of a set which also includes pronoun very difficult to distinguish from the Op kikwaya: in awa and ana. While the final short a remains puzzl- fact, they seem to be used interchangeably.) the identifies the textual 'that ing, gloss clearly oya: The animate forms of the interrogative kikway no longer here, that recently present but now out of mean 'what sort' whereas the inanimate forms in 1933: sight' (Bloomfield, p. 259). addition have the meaning 'what.' In this purely one shows the Of a dozen textual instances only interrogative function they are complemented, for one would on the basis predicative function expect the animate gender, by awina 'who' (4.12). of tdniwa and ewakwd: P296-10 "ehey," itwew nito- kesiw, "wisahkecdhk es oyhl" ' "Oho," cried that 62 While klkway is the standard form in Alberta, Bloomfield "so that was Wisahkecahk!"' In records both kikway and kekway;in his published texts, kekway woman, person is five times more than while the serves as modifier or as the actor almost frequent kikway opposite the others, J6y is true of the unpublished (syllabary) series: kikway outnumbers complement of a verb, e.g., S201-21 niwih-ntawi- kekwayby 4:1. -pakamahwdw 6yd nnah2hkim! 'I am going off to club Bloomfield regarded these as real doublets although he was son-in-law who has just now left!' S169-33 ta-wih- fully aware of the difficulty of distinguishing! and e. The prob- my lem has to be re-assessedin view of the fact that a mer- 'the fellow might be telling the truth' complete -ta-tdpwew oyda ger of i and e has been observed in the Saddle Lake area of north- (referent not present). eastern Alberta (not far from Sweet Grass Reserve) as well as in It is only very tentatively that we identify the certain areas of northern Saskatchewanand Manitoba. stem 6- with the demonstrative stem aw--o6- of 4.11; 63 In spite of insufficientevidence it should be mentioned that they appear to be differentiated phonologically by contrasting 61A pronominaldubitative tdniwitoke 'where is he, I wonder' patterns in the inanimate singular forms; the bisyllabic which occurs a few times in Bloomfield's texts requires further alternant of the interrogative seems to be stressed on the first investigation. Cf. section 3.5, fn. 47, and especially 5.311. , kikway, and the indefinite pronoun is stressed on the An isolated form in -hk is found in S319-44 taniwahknimoso? final syllable, kikwdy. (No distinctions have been observed 'Whereis my grandfather?' While it may simply be a slip of the among the trisyllabic forms of either pronoun.) However, these tongue or of the record,this form could also be interpretedas con- observations must be considered highly tentative since the data taining the nominal locative suffix /ehk/ (3.61); in that case, it are not without contradictions. A detailed study of Cree stress would run counter to the normal pattern of forming particles and pitch phenomenais requiredbefore a fuller statement can be frompronoun roots (cf. 6.422). attempted. 36 WOLFART: PLAINS CREE [TRANS. AMER. PHIL. SOC.

kikway 'what sort; what' always functions predica- thing I follow (TI 1) is . . .; nothing else'; T105p12 tively; it is never used as a modifier (nor is it modified; e-ki-wayesimiskik nanatohk kikwayak 'all sorts of contrast 4.32). It may occur by itself, as part of an things (3p) are tricking you (3p-2)'; T131-4 wiya equational clause, or as a dependent clause. T2-4 kahkiyaw kikway e-k~-wayesihtahk . . . 'For he used kikwaya kiya. 'What kind (3) are you?' P116-38 to trick (TI 3) everything . . .'; T15p74 pitos kikway ekosi mdka kikwayak oki. 'But now, what sort (3p) 'something different'; T13p29 peyak kikway 'one are they?' T58-4 wiya e-mosci-kitapamdt, moy kehci- thing.' nahowkzkwaya. 'But since he looked at him (3-(3')) Modified by participial clause: T15p74 kikway plainly (without telescope) he was not sure (3) what e-cimdsik kikway. 'Something short.' he (3') was'; T10p22 ahpo aya, kikway mdna kiyawdw In composition: T72p14 maci-kikway 'something kt-dpacihtdndwdw.'or, what do you all use?' P304-36 bad'; T73p24, T102p9 mayi-kikway 'something bad'; kikwayih te-miciyahk? 'What are we to eat?' S8-45 neo kakway kih-miyew . . . 'four things he had The interrogative stem kikw- also occurs in deriva- given them (3-(3'))'; T104p7 ndntaw newo-kikway es tion, e.g., (T1-4, T87-5, T108-5, 6) kikwdhtikowiw anima kd-miydt, . . . 'And four things is what he 'what kind of tree is he?' had given them (3-(3')) ...' 4.322. kikway also enters into phrasal combination 4.32. kikway 'something' with certain particles and these phrases, as units, are kikway 'something, a thing, an entity': then used as modifiers, e.g., T103p8 nandtohk kikway Jhi 'all sorts of these 3 ayisiyiniwa (3') people (3').' Thus, it is used with the marker ci in kikwayak question 3p which are either 3' kikwaya introducing questions completely neutral or kikway ci ohci 0 kikway, kikwayi negatively oriented; T4p16 maskocisihk ekota aydwak? 'Were there any Op()P kikwaya (3p) from Hobbema there?' T20p139 kikway ci wiyaso- Although there is some overlap, kikway is comple- wewin ihtakon? 'Is there (0) any law?' kikway ci mented, in the animate forms, by awiyak 'someone' kocawakanis kihaydn? 'Do you have (AI 2) a (4.42). For their semantic characterization see 4.422. match?' The syntactic function of kikway has been analyzed Most typically, however, kikway combines with the only in a highly tentative fashion; to facilitate the negators nama, namoya or eka (depending on the construction of alternative hypotheses, a large number order of the verb; cf. 5.3). nama kikway may, of of examples are given below. course, mean 'nothing' e.g., T58-2 . .. .-ayitdpit, 4.321. Like all pronouns which are not purely nama kikway. '. . . when he looked around, noth- predicative, kikway may function as a noun-substitute; ing'; T121-3 nitonikewak, nama kikway. 'They however, it assumes this role more completely than looked (for them), nothing'; T80plO nama kikway the other pronouns. Thus, it may not only be ohtahtam. 'He didn't get anything to eat from there.' modified by other pronouns, particles, and participial As a phrase, however, nama kikway or ekd kikway clauses but actually enter into composition (6.5) means 'not,' 'not at all,' 'not any,' etc.; note that the with pre-noun particles and numerals. In this substi- members of the phrasal unit do not have to be con- tute function it also has some animate forms, e.g., tiguous. nama kikway or eka kikway are used T509p47 kikwayak oki 'these things (3p)'; T105p12 primarily to modify verbal or nominal expressions e-kZ-wayesimiskik nandtohk kikwayak 'all sorts of (but nama kikway may also function as a predication; things (3p) are tricking you (2).' (In these various 4.323). uses as well as in those described in 4.322 below, Modifying a nominal expression: T120-5 nama kikway greatly resembles the English morpheme thing: kikway matokahp ahpo ta-nanokwaniyik, . . . 'not a thing, not a thing, something, anything, nothing, etc.) even a trace of the campsite (0) could be seen (0'), Unmodified: T58-10 moy konta oma ka-ki-miyin ...'; T10p78 namoya mina pakahkam aya kikway kskway 'not in vain did you give me (2-1) something'; osk-dya wdskahikana e-wi-osihtdhk semak. 'They T120-1 ta-nipahtamawacik kikway 'for you to kill (indf) are not going to build any new (Op) houses (Op) something for them (2-3p).' right away, I don't think (pakahkam).' Modified by other pronouns: T58-10 ekoni ohi Modifying a verbal expression: T58-8 nama kikway kikwaya etokwe, atayohkana 'these (3') things (3') pehow. 'he didn't wait at all'; T9p3 . . ., wiy ekdya there, well, dream spirits (3')'; T58-15 eyakw dnima kikway e-ospitonit. '. .., because he had no arms kikway 'this sort of thing (0)'; T55pll anohc moy left.' kikiskeyihtendnaw kikway eyakw dnima . . . 'today 4.323. Finally, nama kikway may function as a we don't know (TI 21) that (0) . . .'; T131-3 5hi verb-substitute with the meaning 'be gone, not be kikwaya mistikwa 'these different sticks (Op).' here any more'; (cf. also Bloomfield, 1934: p. 284). Modified by particles: T105p14 nayestaw kikwayi In this function nama kTkwayclosely parallels awinipan e-pimitisahamdn . . .; moy kotak kikway. 'The only (4.123), e.g., T56p3 mdk ekwa, nama kikway ekoni VOL. 63, PT. 5, 1973] PRONOUN INFLECTION 37 dcimowina. 'But now there aren't any such stories.' certain resemblances with other pronominal and non- In at least one instance recorded by Bloomfield, nama pronominal paradigms, e.g., /k/ in the animate kikway is actually inflected to show concord: S22-9 plural forms. When the full paradigms are con- nama kekwayah oh otospwakana . . .'Gone (3') was sidered, however, these two appear to be isolated. his pipe (3') ... 4.41. ewako 'the 4.33. kotak 'another' selfsame' kotak 'another, a second one': The stem of the delimiting demonstrative ewako occurs in several alternants, namely eyakw-, ewakw-, 3 kotak ekw-. Except perhaps for stylistic differences, these kotakak 3p alternants seem to be completely interchangeable. 3' kotaka This paradigm is the only non-verbal one in Plains 0 kotak Cree (cf. 5.74) to have an inanimate obviative form, Op kotaka ewakoyiw. (Note that ewakoyiw frequently occurs in The locative is kotakihk 'in another place, elsewhere.' a surface variant ewakwayiw, and cf. also Ellis's form (1962: p. 4-15) ewakweliw.) Examples: 4.34. aya '. . . one' TlOOpll eyakwayiw kd-pimi-kiskisit 'this (0') he remembered as he ran anima aya 'someone': (3)'; T16p77 eyakwayiw kd-nitaweyihtahk. 'That (0') is what he wants (3)'; 3 aya T105p3 maskoc kotak ana manitow eyakwayiw anima 3p ayak 'Rather it was the other 3' e-kZ-totahk. Spirit (i.e. aya devil, 3) who did (3) this (O').' 0 ayi Op aya 3 ewako ewakonik The status of aya as a pronoun (rather than a noun) 3p is doubtful since it never seems to occur as a substitute 3' ewakoni 0 ewako by itself. Although it is often glossed 'someone,' ewakoni its meaning seems to be more fully reflected such Op by 0' glosses as'. . . one; person, people; thing.' aya does ewakoyiw not seem to be definable the semantic by categories While ewako is clearly a kind of demonstrative, the of 4.422. details of its syntactic and semantic function are functions as the final member of aya very commonly yet to be investigated. For it is used to nominal- compounds (6.5). instance, ewako is used in combination with one of ize oski 'a typically particles, e.g. 'young': osk-dya young one,' the other awa and which in turn demonstratives, especially ana, osk-dyak '(the) young people,' may e.g., T20p37 ewakoni ohi dpacihcikana 'these utensils give rise to animate intransitive (AI) verbs, e.g., (Op)'; T91-8 ekonik neki mac-ayisiyiniwak 'those evil 'he is young.' The compound status of osk-dyiwiw people yonder (3p)'; T35p3 ewakw anima ohpihkasikan such words (as distinct from unit words) is evident 'that baking-powder (0).' However, it is by no from the fact that aya keeps its final vowel in the means restricted to this combination, e.g., T33-6 proximate singular form; cf. 3.31. The locative form ekosi nikts-acimon wiy eyako. 'Now I have finished occurs only where aya is part of a compound, e.g. telling (AI 1) this one (0)'; T35p5 . . . kayas eyako pimic-ayihk 'alongside,' kape-ayihk 'all the time,' etc. e-ki-ihkihk, 'long ago this (0) really happened (0)'; Further examples: mistahi 'big': T7p2, 5, T27p2 T35p3 ekwa e-miyosicik ekonik mostoswayanak. 'And mistah-aya 'the big one; bear.' kehte 'old': keht-ayak these (3p) buffalo-hides (3p) were good (3p).' 'the old folks,' nikehte-ayimak 'my old folks.' aya occurs also in compounds whose prior member is a verb: T72p24 peyakwahpicikew-ayak 'carts pulled by 4.42. awiyak 'someone' one horse'; cf. peyakwahpicikew 'he drives one horse.' awiyak 'someone' generally seems to be restricted As a hesitation signal, aya is of extremely frequent to animate forms; it is complemented, for the in- occurrence and may be inserted virtually anywhere animate gender, by kzkway in a sentence; e.g., T47pll 6k dyak notokwesiwak 'something' (4.32). 'these old women'; T10p2 pihtaw mana aya ka- 3 awiyak -nehiyawehk moya ci. 'But you (indf) can't talk Cree 3p awiyak for everything, can you?' 3' awiya 4.4. ISOLATEDPARADIGMS 4.421. The pronouns awiyak 'someone' and kikway The paradigms of the delimiting demonstrative 'something' (4.32) are very similar in syntax and ewako and of the indefinite pronoun awiyak show meaning. 38 WOLFART: PLAINS CREE [TRANS. AMER. PHIL. SOC.

There may be some overlap in the animate forms 4.5. PERSONALPRONOUNS While the data do not a more detailed state- permit Personal pronouns are not inflected for number or it is worth that is attested as a ment, noting kikway obviation. As a set, however, they closely parallel false start where was intended: awiya T97p2 kahkiyaw the possessive paradigm of nouns (3.22), albeit with - esa awa kzkway- -, awiya e-ki-osimimit, wzsahkecdhk. certain differences. There is no distinction of proxi- - he had for a little brother 'Everything- -, everybody mate and obviative in the personal pronoun paradigm; this Wisahkecahk.' (AI 3), the one third-person form is used for both categories. like also be modified Just kzkway, awiyak may by There is also no indefinite actor form. (Note also and pronouns particles, e.g., T71p16 kd-wdpamdcik the w- alternant of the third person prefix.) ohi saw this e-wdwdskaweyit awiya. 'They (3p-(3')) There are two sets of personal pronouns: nmya 'I' around P260-26 ekd (3') somebody (3') moving (3')'; represents the simple type, nmsta'I, too; I myself' the wlkim. 'Don't wiya pitos awiya marry (2-3) anyone affirmative type. else.' (Note that kikway is also found in a number of further combinations; cf. 4.321.) 4.51. Stems Similarly, awiyak may also function as a phrasal modifier, especially when combined with a particle The stem of the simple personal pronouns is -Iya-, such as namoya 'not'; e.g., S257-3 ekosi namoy awiya that of the affirmative pronouns -ista-. Even though aydwew iskwewa; 'He, then, didn't have (3-(3')) any the details of derivation remain to be discovered, (3') wife (3').' In this function, awiyak may even there can be little doubt that -4ya- and -Ista- are occur with an inanimate noun, e.g., P8-1 ekwah etymologically related. The Proto Algonquian forms kiydm peyak eh-aydt awiyak maskihkiy, kit-etohtatdw. which Haas (1967b) reconstructs for the simple set, 'And please let anyone (peyak) who has any (awiyak) *nilawa 'I' and *nilawendn 'we (excl.),' suggest that medicine (0) bring it (AI 3) there.' the current system has undergone a good deal of 4.422. In meaning, awiyak and kikway are both leveling. indefinite and SPECIFIC. also the (Note particle peyak 4.52. Paradigms 'one, a certain' which is both definite and specific.) Thus they are distinct from the "indefinite actor" The full sets follow. The 21 form of the simple set category of verbs and the "indefinite possessor" shows an alternant of the stem, -iyd-, which occurs category of nouns whose major characteristic is not in- only here; thus, klydnaw is also the only form of this definiteness but GENERALITY. (Whether "indefinite" set stressed on the penultimate syllable. (It may be functions at all in these categories remains to be of interest that the same state of affairs is described investigated; cf. also 2.12, 2.511, and 2.52.) by Howse (1844: p. 60) for an eastern dialect.) Consider as an example the following sentence: 1 nmya nmsta S63-5 kehcindh awiyak e-sihkimikoyek. 'Surely some- 2 kiya kista one (3) has put you up to this (3-2p).' If we compare 3 wzya wista this with the "indefinite (general) actor" sentence lp niyanan nsstandn which would roughly correspond to it, the semantic 21 kiydnaw kistanaw difference becomes fairly obvious: kehcind e-sihkimika- 2p kwyawdw klstawdw wiyek. 'Surely someone has urged you (indf-2p) . .., 3p wiyawaw wlstawdw there has been directed to surely urging ' you ..., surely you have been urged ... 5. VERB INFLECTION The difference becomes even more obvious when, 5.1. BASICVERB TYPES for instance, the corresponding negative sentences are considered. kehcind namoy awiyak e-sZhkimikoyek. There are four BASIC verb types which are distinct 'Surely nobody has urged you (3-2p) . . .'; contrast in derivation and inflection. There is also a small kehcind namoy e-sihkimikawiyek. 'Surely there has number of MARGINALparadigms which use slightly been no urging directed to you (indf-2p) . .., surely divergent sets of endings; see 5.8. you have not been urged ... ' The four basic classes are defined by the dimensions The difference between the general ("indefinite") of and gender. Thus, there are intransi- categories of nouns and verbs and the specific in- tive verbs with animate actor (AI), e.g., apiw 'he definite pronouns is not only semantic but is also sits,' and intransitive verbs with inanimate actor reflected in inflection and syntax. Thus, for instance, (II), e.g. kisikdw 'it is day.' Transitive verbs, on the there is only an "indefinite actor" form in verbs other hand, differ by the gender of the goal; thus, (action on a general goal is indicated derivationally; there are transitive verbs with animate goal (TA), cf. 6.436). Indefinite pronouns, by contrast, occur e.g. wapamew 'he sees him,' and transitive verbs with either as actor or as goal, e.g., T50-2 . .., ka-wdpamdt inanimate goal (TI), e.g. wdpahtam 'he sees it.' awiya e-osiskwepayihoyit, 'he saw (3-(3')) someone (3') Derivationally, transitive as well as intransitive sticking his head up (3').' verbs largely come in pairs which differ by the gender VOL. 63, PT. 5, 19731 VERB INFLECTION 39 of the goal or of the actor, respectively; e.g., TA e.g. dpacihtaw 'he uses it.' Bloomfield used the iteyim- 'think so of him,' TI iteyiht- 'think so of it'; terms "pseudo-transitive" or even "pseudo-transitive TA pakamahw- 'strike him,' TI pakamah- 'strike it.' inanimate" (1946: pp. 95, 112) for this type. Stems AI akohcin-, II akohte- 'be in water'; AI mihkosi-, II ending in -htd- (many of them paralleled by TA mihkwd- 'be red.' stems) are most characteristic of this type, e.g., The dimensions which are labeled "transitivity" wanihtaw 'he loses it' (cf. wanihew 'he loses him'); and "gender" define the focal type of each class; the but there are others as well, e.g. miciw 'he eats it': syntactic and semantic properties implied by the label T55p35 nama kikway e-mtciydhk. 'There was nothing "transitivity" are not always shared by the entire for us to eat'; for contrast consider mzcisjw 'he eats,' class. The morphological basis of the present classifi- as in T55p26 ekdy ekosi isi-mlcisokl 'Don't eat like cation therefore needs to be emphasized. While the this!' classification accounts for both the inflectional para- The other type is generally intransitive, e.g. apiw digms and the derivational structure of Cree verbs, 'he sits,' and constitutes the great majority (in terms there is a conflict of morphology and syntax in one of list frequency) of animate intransitive (AI) verbs. important point: animate intransitive (AI) verbs However, verbs of this type are occasionally also used syntactically fall into an intransitive and a transitive transitively even where regular transitive animate type, e.g., apiw 'he sits' and oszhtaw 'he makes it'; for (TA) and transitive inanimate (TI) parallel stems further details see 5.12. An expression such as "a exist also. Thus, itwew 'he says so' which is typically syntactically transitive animate intransitive verb," intransitive (and is paralleled by TA itew 'he says therefore, makes sense only when the essentially so to or of him' and TI itam 'he says so to or of it'), morphological nature of the classification, and the occurs transitively in T115p3 ". . ." itwew es oma necessarily limited scope of the labels, are kept osokan. '". . ." he said about his rear-end.' An- in mind. other example of this phenomenon is T60p9 sihkihpa Similarly, while the inflectional paradigms are an ekoni ohi esa kd-nawaswet awa wisahkechk, . . . integral part of the classification into four basic 'This mudhen (3') he chased (AI 3), this Wisahkecahk types, the labels again are not to be taken too literally (3), . . .'; the TA and TI parallel stems are nawas- with regard to all forms. Thus, not even the transi- wdtew, nawaswatam 'he pursues him, it.' tive animate (TA) paradigm shows both referents, 5.13. Transitive Inanimate actor and goal, expressed in all its forms (cf. 5.61, 5.64) and the transitive inanimate (TI) paradigm, As has been noted above (5.1), transitive inanimate while clearly transitive in derivation, syntax, and (TI) verbs show inflectional affixes for actor only. meaning, in Cree shows no suffixes for the goal at all; In meaning, syntax, and in the derivational parallelism cf. 5.13. with the transitive animate type, however, the transi- The classification of verbs into four basic types is a tive inanimate class as a whole is distinctly transitive. common trait of Algonquian languages. There can In many other Algonquian languages (but by no be no question about its appropriateness to the means in all) the goal of transitive inanimate (TI) description of Cree even though its terminology is not verbs is morphologically expressed. Ojibwa has a immediately obvious from the Cree situation. double paradigm in the independent indicative where one paradigm indicates a goal, e.g., otepwettdn 'he 5.11. Transitive Animate believes it,' and the other does not, e.g., tepwettam 'he believes.' Goddard (1967) has made a convincing Among transitive animate (TA) verbs, there is a case for the existence of a double paradigm in Proto distinction between a "two-place" and a "three- Algonquian; cf. also 5.614. In Cree this distinction place" type. (For these terms, cf. Lyons, 1968: has been obliterated. The goal is not indicated by a p. 350.) The two-place type involves an actor and a suffix but the meaning and syntax of most transitive goal, both animate, e.g., wdpamew 'he sees him.' inanimate verbs is nevertheless clearly transitive. The three-place type further involves a second goal There are some transitive inanimate (TI) verbs which may be of either gender or number and which which "refer to no identifiable object, but have a is not morphologically expressed, e.g., miyew 'he gives merely (Bloomfield, 1946: p. 95; 1962: it or him to him'; kimotamawew 'he steals it or him formal goal" p. 46), e.g., mdham 'he canoes downriver.' That is, from him'; this has traditionally been called the such verbs never occur with an expressed goal in "double goal" type (cf. 6.446). texts; whether they cannot, under any circumstances, take a whether the deletion" is 5.12. Animate Intransitive goal (i.e., "object obligatory) has not been established. Bloomfield Animate intransitive (AI) verbs fall into two types also referred to such verbs as "pseudo-intransitive" which are syntactically and semantically distinct (1958: p. 34) and Goddard (1967: p. 67) even uses from one another. One is always transitive even "pseudo-AI" (which further obscures the contrast though only the actor is morphologically expressed, of morphological and syntactic-semantic criteria in 40 WOLFART: PLAINS CREE [TRANS. AMER. PHIL. SOC. classification). In Cree, however, the evidence for as lists rather than two-dimensionally. They follow this subtype is inconclusive, and further informant the remainder of the basic paradigms. work is required before a fuller statement can be attempted. 5.23. Level of Representation The are in 5.14. Inanimate Intransitive paradigms given phonemic represen- tation; cf. appendix A, especially footnote 85. How- Inanimate intransitive (II) verbs are always ever, to facilitate use of these tables for quick reference intransitive. (Cf. also 6.43.) purposes, suffix-initial i which reflects the palatalizing /i/ (rather than /e/) is underlined (cf. appendix A: 2). 5.2. PARADIGM TABLES The codes for the person-number-gender-obviation are defined in table 1 of section 2.01. The basic are here to summa- categories paradigms presented For the alternant forms of the see rize the data which are in this personal prefixes, analyzed chapter. A: 6. Direct and inverse are discussed in Without this aid to the sections appendix orientation, analytic and the terms and be difficult to follow. The 2.5, mixed, third-person, you-and-me might paradigms may are introduced in 5.611 and 5.612. also serve as a practical reference guide in the identifi- cation of newly encountered forms. 5.24. Versions of the TA Independent 5.21. Sources Indicative Paradigm With the exception noted below, the tables only The TA independent indicative paradigm is pre- include forms which are textually attested, either in sented in two versions; version B is omitted for the my own or in Bloomfield's texts. TA conjunct. The rationale for this double presen- A few forms which are not attested in the texts tation is discussed in detail in sections 5.61-5.63. are cited in parentheses. These forms are considered Version A reflects the semantic structure of the to be of fairly high authenticity since they are uni- paradigm as a whole, as well as the morphemes formly given by the four major missionary sources actually present in each form; thus, there is no on Plains Cree (Lacombe, Hunter, Lessard, Edwards). obviative (3') morpheme present in wapamew 'he sees They are also at least partially confirmed by the him.' remainder of their respective paradigms.64 Version B reflects the syntagmatic relations entered into by each individual form; for example, refer- 5.22. Empty Positions entially and with regard to concord, wapamew 'he sees him' has a proximate (3) actor as well as an A paradigmatic position which is left empty in obviative (3') goal. principle presents a choice of two interpretations: either no such form exists, or it has not been recorded. The non-existence of a form for a paradigmatic TABLE 5 position may reflect the neutralization of a contrast; SUMMARY OF VERBAL PARADIGMS this is indicated by the layout and the choice of parameters. For example, contrast the direct and Basic paradigms: inverse third-person sets of the TA independent TA independent indicative indicative paradigm. Other cases may involve se- (A: meaning and morphemes present) Table 6 TA indicative Table 7 mantic restrictions (e.g., TA imperative 21-1) or independent of (B: reference) they may correlate with the development suppletive TA conjunct simple and changed forms (e.g., the indefinite form of the AI independent (indicative) Table 8 indicative paradigm). In such cases, the fact that TA conjunct subjunctive and iterative Table 9 no form exists is indicated a dash: - . TA imperative Table 10 by AI Table 11 Where a lacuna is considered accidental, it is TI Table 12 indicated by empty brackets: [ ]. In the present II Table 13 this is restricted to the non- tables, type conjunct Incompletelyattested paradigms: indicative (subjunctive and iterative) paradigms Preterit (TA, AI, TI, II) Table 14 which are readily derived from the conjunct indicative Dubitative (TA, AI, TI) paradigms (cf. 5.33). Marginaland suppletiveparadigms: Since the various and dubitative preterit paradigms Relational (AI, TI) 5.813 are very incompletely attested, they are presented Diminutive (TA, AI, TI) 5.82 4 TA inanimate actor 5.83 No attempt is made to incorporate the incredible diversity TA indefinite actor 5.84 of paradigms found in missionary sources; while composite pre- TI indefinite actor 5.85 sentation would obviously open interesting perspectives, practical AI, AI inanimate actor 5.86 difficulties and problems of interpreting these sources preclude such an approach in the present context. VOL. 63, PT. 5, 1973] VERB INFLECTION 41 TABLE 6

TA INDEPENDENT INDICATIVE A: Meaning and MorphemesPresent

DIRECT INVERSE

[MIXED] -3 -3p -3' 3- 3p- 3'- indf- -aw -awak -imawa 1- ni- -aw ni- -awak ni- -imawa -1 ni- -ik ni- -ikwak (ni- -ikoyiwa) 2- ki- -aw ki- -awak ki- -imawa -2 ki- -ik ki- -ikwak (ki- -ikoyiwa) lp- ni- -anan ni- -ananak ni- -imanana -lp ni- -ikonan ni- -ikonanak ni- -ikonana 21- ki- -anaw ki- -anawak ki- -imanawa -21 ki- -ikonaw ki- -ikonawak ki- -ikonawa 2p- ki- -awaw ki- -awawak ki- -imawawa -2p ki- -ikowaw ki- -ikowawak ki- -ikowawa

[THIRD-PERSON] -3' 3- -ew -imew -3 -ik 3p- -ewak (-imewak) -3p -ikwak 3'- -eyiwa -3' -ikoyiwa

[YOU-AND-ME] -1 -lp 1- lp- 2- ki- -in ki- -inan -2 ki- -itin ki- -itinan 2p- ki- -inawaw ki- -inan -2p ki- -itinawaw ki- -itinan

TABLE 7 TA INDEPENDENT INDICATIVE B: Reference

DIRECT INVERSE

-3 -3p -3' -3" 3- 3p- 3'- 3"- indf- -aw -awak -imawa 1- ni- -aw ni- -awak ni- -imawa -1 ni- -ik ni- -ikwak (ni- -ikoyiwa) 2- ki- -aw ki- -awak ki- -imawa -2 ki- -ik ki- -ikwak (ki- -ikoyiwa) ip- ni- -anan ni- -ananak ni- -imanana -lp ni- -ikonan ni- -ikonanak ni- -ikonana 21- ki- -anaw ki- -anawak ki- -imanawa -21 ki- -ikonaw ki- -ikonawak ki- -ikonawa 2p- ki- -awaw ki- -awawak ki- -imawawa -2p ki- -ikowaw ki- -ikowawak ki- -ikowawa 3- -ew -imew -3 -ik 3p- -ewak (-imewak) -3p -ikwak 3'- -eyiwa -3' -ikoyiwa

5.25. Summary of Verbal Paradigms The distribution of the negators crosscuts this classifi- cation: nama, namoya, and others based on nam- Table 5 a of verbal presents summary paradigms, occur with independent clauses while ekd is found including the marginal and suppletive paradigms with imperative and with conjunct clauses. treated in 8.5.

5.3. MODAL CATEGORIES 5.31. Subcategories

Plains Cree verbs are inflected in three ORDERS: The missionary sources indicate a wide variety of INDEPENDENT, CONJUNCT, and IMPERATIVE. The subcategories. Using data from the James Bay orders use different sets of affixes (although some dialect, Ellis (1961, 1971) posits three dimensions subsets recur in more than one order, cf. 5.67). of contrast below the order level. Ellis's scheme is The orders also differ in their syntactic function. given in full (table 16) to serve as a point of reference Briefly, independent and imperative order forms occur and of contrast. as whole sentences, e.g., T33-3 nipimipahtdn. 'I was However, the three modes which Ellis posits for running'; P264-38 dhkameyimokl 'Do your best!' the conjunct order are not coordinate. It is clear Conjunct forms occur in dependent clauses, e.g., from the morphology of these categories, and their S58-5 tdpweh e-kih-micisot kawisimow eh-nipdt. syntactic functions seem to pattern correspondingly, 'Truly, when he had eaten, he lay down to sleep.' that the indicative and subjunctive together are 42 WOLFART: PLAINS CREE [TRANS. AMER. PHIL. SOC. TABLE 8

TA CONJUNCT SIMPLE AND CHANGED (INDICATIVE)

DIRECT INVERSE

[MIXED] -3 -3p -3' 3- 3p- 3'- indf- -iht -ihcik -imiht 1- -ak -akik -imak -1 -it -icik (-iyit) 2- -at -acik -imat -2 -isk -iskik (-iyisk) ip- -ayahk -ayahkik -imayahk -Ip -ikoyahk -ikoyahkik (-ikowayahk) 21- -ayahk -ayahkok (-imayahk) -21 -ikoyahk -ikoyahkok (-ikowayahk) 2p- -ayek -ayekok (-imayek) -2p -ikoyek -ikoyekok (-ikowayek) [THIRD-PERSON] -3' 3- -at -imat -3 -ikot 3p- -acik (-imacik) -3p -ikocik 3'- -ayit -3' -ikoyit

[YOU-AND-ME] -1 -lp 1- lp- 2- -iyan -iyahk -2 -itan -itahk 2p- -iyek -iyahk -2p -itakok -itahk

TABLE 9

TA CONJUNCT SUBJUNCTIVE AND ITERATIVE

DIRECT INVERSE

[MIXED] -3 -3p -3' 3- 3p- 3'- indf- -ihci -ihtwawi -imihci 1- -aki -akwawi -imaki -1 -ici -itwawi [ ] 2- -aci -atwawi [ ] -2 -iski [ ] [ ] lp- -ayahki -ayahkwawi [ ] -lp [ ] [ ] [ ] 21- -ayahko -ayahkwawi [ ] -21 -ikoyahko -ikoyahkwiwi [ ] 2p- -ayeko -ayekwawi [ ] -2p -ikoyeko -ikoyikwawi [ ] [THIRD-PERSON] -3' 3- -aci [ ] -3 -ikoci 3p- -fitwawi 3 -3p -ikotwiwi 3'- -ayici -3' -ikoyici

[YOU-AND-ME] -1 -lp 1- lp- 2- -iyani -iyahki -2 -itani [ ] 2p- -iyeko -iyahki -2p -itakwawi [ ]

opposed to the dubitative. Thus, the forms of the indicative and subjunctive, e.g. (Ellis, 1961: p. 122) subjunctive differ from those of the indicative simply ke-wdpamdwakwe '(I wonder whether) I'll be seeing by the suffixation of a closing morpheme /ih/ which him' versus ke-wapamak 'I'll be seeing him'; consider is accompanied by the automatic selection of the also /ikwe/ (3-1) and /eskwe/ (3-2) versus non- plural marker /waw/ (instead of /k/; cf. 5.48), e.g., dubitative /it/ and /esk/, respectively; etc. apiydn 'that I sit' versus apiyani 'if I sit,' or wiydpa- Consequently, a further opposition, of dubitative mdcik 'when they see him' versus wiydpamdtwdwi versus non-dubitative, seems to be called for. The 'whenever they see him.' The formation of dubi- corresponding level of the classificatory hierarchy is tative forms, on the other hand, involves a fairly tentatively termed "sub-order." Under this revised complex set of suffixes differing considerably from the scheme there would be no contrast, in the independent VOL.63, PT. 5, 19731 VERB INFLECTION 43 order, in two of the four dimensions. There may TABLE 12 even be grounds for questioning the construction of a TRANSITIVEINANIMATE (TI) unified scheme for all three orders. The present data do not permit verification of the Conjunct Conjunct Ellis scheme, especially with regard to the less common Independent Simple and Subjunctive and Indicative dimensions.55 Thus, only non-dubitative forms are Changed Iterative considered for and no data are available any order, indf -amihk -amihki for the tense contrast within the conjunct order. 1 ni- -en -aman -amani Therefore, we follow Bloomfield's practice of using 2 ki- -en -aman -amani one generic term for all subcategories (except that lp ni- -enan -amahk -amahki for the dimension "sub-mode" is 21 ki- -enaw, -enanaw -amahk -amahko changed-unchanged 2p ki- -enawaw -amek -ameko 3 -am -ahk -ahki TABLE 10 3p -amwak -ahkik -ahkwawi 3' -amiyiwa -amiyit -amiyici TA IMPERATIVE Imperative Immediate Immediate Delayed

[YOU-AND-ME] [MIXED] 2 -a -am6hkan 21 -etan -am6hkahk -1 -lp -3 -3' -3p 2p -amok -am6hkek 2- -in -inan -i, -0 -i, -0, -ik (-im) 21 - - -atan -atanik (-imatan) 2p- -ik -inan -ihk -ihkok (-imihk) TABLE 13 INANIMATEINTRANSITIVE (II) Delayed Conjunct Conjunct [YOU-AND-ME] [MIXED] Simple Subjunctive -1 -3 -3' -lp -3p Independent and and 2- -ihkan -ihkahk -ahkan -ahkanik -imahkan Indicative Changed Iterative 21- - - -hkahk -ahkahkik (-imahkahk) 2p- -ikhek -ihkahk -ahkek -ahkekok (-imahkek) 0 -w, -0" -k -ki Op -wa -ki [ ] O' -yiw .yik -yiki O'p -yiwa -yiki [ ] TABLE 11 ANIMATEINTRANSITIVE (AI) The first ending occurs with vowel stems, the second with n-stems; cf. 5.53. Conjunct Conjunct TABLE 14 Independent Simple and Subjunctive and Indicative Changed Iterative PRETERIT(INDEPENDENT)

indf -hk -hki h-preterit ht-preterit p-preterit 1 ni- -n -yan -yani TA mixed 1-3 ni- -ah ni- -ahtay 2 ki- -n -yan -yani 1-3p ni- -ahtayak ni- -nan 1-3' ni- -imah lp -yahk -yahki 2-3 ki- -ah 21 ki- -naw, -nanaw -yahk -yahko 3-1 ni- -ikoh 2p ki- -nawaw -yek -yeko 3p-1 ni- -ikohtayak, -ikohtayik 3 -w, -0' -t, -k' -ci, -kis 3-2 ki- -ikoh 3p -wak -cik, -kiks -twawi, -kwawia 3-21 ki- -ikohtanaw third-person 3- (3) o -ah 3' -yiwa -yit -yici 3p-(3') o- -ahtawRw (3')-3 o- -ikoh you-and-me 2-1 ki- -ih Imperative 1-2 ki- -itih Immediate indf actor indf-1 ni- -ikawih Delayed indf-2 ki- -ikawih AI 1 ni- -h 2 -0, -i -hkan 2 ki--h 21 -tan -hkahk 21 ki- -htanaw -k -hkek 2p ki- -htRwaw 2p 3 o- -h o- -htay 3p o- -htawaw 3' o- -yih a The first ending occurs with vowel stems, the second with relational 1 ni- -wah n-stems; cf. 5.52. TI 1 ni- -Eh 2 ki- -eh 5 There is a contrast between the 3 o- -eh striking relatively simple relational 1 ni- -amwah verb paradigms attested in Plains Cree texts and the immense proliferation of forms which Ellis (1971) reports for the Swampy II 0 -h -pan and Moose dialects of eastern James Bay. 44 WOLFART: PLAINS CREE [TRANS. AMER. PHIL. SOC.

TABLE 15 dubitative occurs with an interrogative pronoun or DUBITATIVE conjunction: T91p20 tdnisi e-itinikekwe. 'I wonder how he is faring'; T109plO . . ., tnte mTna wehti- Independent Conjunct ndhkwe askihkwa. 'I don't know where he got a pail'; P200-35 awlna wekimdwikwe ekw ana . . TA 3-1 ni- -okwe 'whoever is chief, he will . . .'; T99p8 tdntahto 3- (3') -ahkwe tdntahto. indf-1 ni- -ikawinatoke nepahdhkwe, ahpo eyako namnoynikiskeyihten 'How many he may have killed, that I don't know, AI 2 ki- -natoke just how many.' 21 -yahkakwe (?) There are also isolated cases of a dubitative ending 3 -toke -kwe with a nominal or stem 3.5 -tokenik -wakwe (?) occurring pronominal (cf. 3p and 4.21, footnotes 47 and 51). TI 3 -am6toke -amokwe Most striking, perhaps, is the use of the dubitative 3p -am6tokenik particle etokwe 'I wonder' (Bloomfield consistently 3' -amiyitokeni records etoke). etokwe occurs very freely both with independent and conjunct order verbs, as well as in non-verbal sentences. Statistical studies (cf. Wolfart TABLE 16 and Pardo, 1972) show that etokwe is exactly ten SUBCATEGORIES OF THE VERB ACCORDING times more frequent in the Alberta texts than in TO ELLIS (1961, 1971) Bloomfield's texts. Stylistics, dialect differences, the time difference, etc. undoubtedly must be taken into Order Mode Tense Submode account in this matter. But the most interesting which has to remain at independent indicative neutral question, completely open preterit the present time, is whether there is an observable dubitative neutral decline in the frequency of dubitative verb forms preterit which would correlate with the increasing frequency of the non-verbal dubitative marker etokwu. conjunct indicative neutral unchanged changed preterit unchanged 5.32. Independent Order changed The to the subjunctive neutral unchanged non-preterit paradigm corresponds changed INDICATIVE mode of Bloomfield and Ellis, and the dubitative neutral unchanged latter term will be used for the sake of convenience. changed This mode is used in simple statements. preterit unchanged There are three different PRETERIT which changed paradigms are much less fully attested. In spite of some imperative immediate syncretism, they seem to be distinct (but note that delayed there is a much greater degree of syncretism evident in James Bay Cree; cf. Ellis, 1971). Following Hunter and Bloomfield, the preterits are identified sometimes convenient). In choosing MODE (Bloom- by reference to morphological characteristics: H-, field, 1946; in 1928 he had used the less neutral HT- and P-PRETERIT. "tense"), the implications of the Ellis scheme are 5.321. The p-preterit is distinct from the others: expressly disavowed for Plains Cree until further it does not take the third-person prefix o- (cf. 2.11); data permit deeper analysis. We interpret Bloom- its marker, which is tentatively set up as /Lpan/, field's usage in the same way: certainly not as implying occupies a different suffix position (6, 7? cf. 5.46) a linear structure; but to avoid commitment in a from that of the h- and ht-preterits (4); finally, its situation which remains insufficiently clear. meaning seems to affect the meaning of the stem, 5.311. Although the data do not permit a consistent much as a derivational suffix would, rather than that morphological analysis of the dubitative in Cree, the of the construction as a whole. In Bloomfield's available verb forms are recorded in table 15. words (1928: p. 429), the p-preterit "is used of past The range of meaning of the dubitative is exempli- occurrence no longer true in the present: pimdtisipan fied in these sentences: S300-15 . .., nipehikawinatoke, 'he was alive (and is now dead)'; it seems to be con- . . 'I wonder if they are waiting for me'; S130-7 fined to archaizing language." The p-preterit is rare; kaskeyihtamotokenik aniki nitehtawaw. 'Surely these textual examples seem to be restricted to II stems or co-parents-in-law of mine must be lonely'; S84-11 to AI/TI indefinite actor forms which pattern like II mah-mominewak, itah misawdc kiw-oh-pimdtisiwakwe. stems (5.85): S182-32 miywdsinopan eh-otihtikoyahk! 'They are eating berries from the bush, they will be 'It seemed so good (just a moment ago) that he came living on that, no doubt.' Quite typically, the to us'; S13-7 misiwe esah kiy-iskipepan. 'The flood VOL. 63, PT. 5, 19731 VERB INFLECTION 45 had covered all'; T520p23 aspin nmmihitonaniwipan. ki-waskawiyin, semak ka-kih-nipahikoh awa niteskan! 'There had been dancing there.' 'Certainly, if you had budged, at once this horn of The same suffix also occurs, in a slightly variant mine would have killed you'; S73-12 kont itah misk- form, with nouns (3.5) and with at least one pronoun wamiyah kih-pwkwdskawdt kohtdwiy, '. . .' ka-k~h- (4.123). -itwdhtdnaw! 'If your father had broken through the 5.322. Both h- and ht-preterit use the third-person ice, no matter where, then we could have cried '. ..'!'; prefix o-; this is the only occurrence of the o- prefix P250-14 oka-kih-miyweyihteh ninahdhkisim, itah ta- in the entire verb system. -wdpamdt okdwiya . . . 'How glad my son-in-law Bloomfield (1928) considered the ht-preterit archaic would have been to see his father here ... . but the present data support no such differentiation. The full meanings of the h- and ht-preterits have not 5.33. Conjunct Order been established, and the present evidence neither The four modes of the conjunct order are defined supports nor contradicts Bloomfield's claim (1928) by two morphological criteria which intersect one that there is no difference in meaning between them. another. One of these is INITIAL CHANGE which is A few observations be worth in may noting spite described in detail in appendix A: 7. It is the basis of their highly tentative nature. The ht-preterit for Ellis's submode dimension. seems to be used of events which are mostly completed, The other criterion is the presence or absence of e.g., T122p4 niwanihahtayak 'I lost them'; S9-46 the which is not followed other ewakw awa 'That morpheme /ih/ by any otawdsisiwihtay wisahkecdhk. boy suffix the environment of the third was in his (5.49). (In /ih/, Wisahkecahk childhood'; cf. also S12-41, is in northern S245-41. person pluralized by /waw/ (5.48); Alberta, /waw/ is used instead of /k/ throughout with the the of the Compared ht-preterit, meaning the conjunct order, not only in the subjunctive and is much more different from that of h-preterit clearly iterative modes.) the since the denote occur- p-preterit h-preterit may The diagram displays the two oppositions and the rences which mdninakisk mihta persist, e.g., T110p5 modes they define. otawatd. "Then he kept hauling wood" (informant's translation). The continuing, imperfective meaning initialchange of the h-preterit is fairly obvious in T108-7 mdninakisk + ekwa ekoni opahkwatind. piyisk kaskhhtdw oma ka- -pahkwatindt, . . . 'He then kept taking these off. + ITERATIVE SUBJUNCTIVE in /ih/ Finally he succeeded taking them off, . . .' The - CHANGED SIMPLE h-preterit very often occurs with the particle mdnina- kisk, dninakisk whose meaning is not quite clear; The terms which in the are those of glosses include 'then, right away, vigorously, entirely, appear diagram Bloomfield to Ellis's terms just like that,' etc. Thus, T27p4 pikohtitdw otatosisa, (1928); they correspond as follows: mdninakisk opikohtitd. 'He broke his arrows (on he broke them.' him), just simple: indicative neutral unchanged Both the h- and ht- preterit commonly express a changed: indicative neutral changed past expectation which has not been fulfilled or subjunctive: subjunctive neutral unchanged describe a situation which no longer exists: S281-39 iterative: subjunctive neutral changed yah, kipj-wzwil 'Indeed, you came here to take a wife!' (sc. but have not done so)'; S300-5 ekwah Note that Bloomfield's terms avoid ranking the micisoyahk! knohtehkatdhtdnaw, nisimisl 'Now we oppositions with respect to each other. (Neverthe- shall eat! We surely were hungry, little sister!' less, it may sometimes be convenient in discussions of (Note that the same function is also exemplified for morphology to let "conjunct indicative" refer to the the p-preterit in S182-32, above.) In the following simple and changed modes collectively.) The mean- example, the preterit verb form includes the preverb ing of the modes may be hinted at in glosses like these: wi 'will, intend to': S257-11 kiwz-sikahotih oma! 'I simple: 'that it is . . meant to comb your hair!' changed: 'it . . The "irreal" function of the is being preterit particularly subjunctive: 'if it be . . .' striking in conditional sentences; the condition is iterative: 'whenever it is . . .' usually (but not necessarily) expressed by a verb in the simple conjunct mode: P192-16 ekosi kiweh- There follow brief indications of the meaning and -nipahitih, ki-kisiwdhitdn. 'Thus I was going to kill use of the conjunct order modes. you, if I had incurred your anger.' Most examples 5.331. By far the commonest use of the simple of the "irreal" preterit show the future preverb ka conjunct is with the future marker kita, ta, e.g., combined with the perfective (?) preverb k1l, resulting S12-8 nama kaskihtdw kita-mosowit. 'He didn't in an explicitly past irreal conditional: S50-8 misawdc succeed in turning into a moose.' 46 WOLFART: PLAINS CREE [TRANS. AMER. PHIL. SOC.

But even without this marker, simple conjunct 'Presently, he came to a house'; P262-10 tanehki k-oh- clauses may express subsequence or purpose; they are picicik? 'Why have they moved camp?' usually introduced by a conjunction like nawac 'it is 5.333. The subjunctive mode expresses a condition, better,' e.g., S238-39 . . ., nawac kakwe-wdpamacik. in a very wide sense. For instance, S62-2 kispin 'you ought to try to see them.' nipahikawiydni, . . . 'if I am slain (indf-1), .. .'; The simple conjunct is also governed by certain P8-7 nika-pimitdcimon, pihtokeydnih. 'I shall be specific conjunctions such as maywes, pamoyes 'before' crawling along the ground when I enter the lodge.' (which may, however, also take the changed con- The predication on which a subjunctive clause junct with preverb ka); tdnika, pitane 'would that!' depends typically (but not obligatorily) contains an etc. Examples: T55p71 maywes pe-sipwehteydhk 'be- indication of subsequence (futurity), e.g., P8-10 fore we left there'; S8-24 maywes askiy ihtakohk ekwa ki-nipahiyeko, ka-pakocenindwaw. 'And when 'before the earth existed'; tdnika kipa takohtet can. you have slain me, you will (ka) cut me open'; 'I wish John would get home soon'; P78-1 pitaneh S238-3 takwdmoyani ekotj, omisi itwehkan: 'When pimdtisit! 'May he live!' you get there, speak then (delayed imperative) thus.' 5.332. The changed conjunct indicates subordina- Finally, subjunctive clauses are often used for tion in an entirely neutral way. It is the most expressing the time of day (cf. 5.334), e.g., P6-34 versatile, in its syntactic use, of the conjunct modes, hdw, tipiskdkih isko nika-pehon. 'Well, I shall wait and consequently the most widely used as well. until dark.' But note that besides, say, wdpahki 'in the tomorrow' we also find the Initial change may operate on the first syllable of morning; changed form with the same the verb stem, e.g., S11-31 tekohtetawa kiseyiniw 'when conjunct e-wdpahk largely the old man arrived.' More typically, however, it meaning. 5.334. The iterative combines not the affects one of a small set of (6.521), such as only morpho- characteristics of the and ki which is changed to kd, e.g., S246-25 eh-pzhtoket, logical changed subjunctive initial and the suffix but kd-wdpamdtohi ka-kih-nipahdt napesisah. 'When she modes, namely change /ih/, it would some of their more salient came in, she saw (3-(3')) the boy (3') whom she had also, seem, syn- tactic features: the mode's use in killed (3-(3')).' The most frequent preverb, e, seems changed participial and narrative and the feature of to be nothing but a "vehicle" for initial change; its clauses, subjunctive's underlying, unchanged form does not occur in Cree. conditionality. S244-19 mistahi mdna Impressionistically speaking, the use of e is gaining Examples: miyweyihtamwak oki minisah 'Those at the expense of forms where the stem itself undergoes oskinikiwak, miyZcitwawi. young men were when had berries to initial change. Examples: S41-41 papdmitdcimow, always very glad they S8-2 mdna owikimdkana e-nitonawat 'he crawled about, looking for him'; eat'; kitahtawe sepwehteci, 'Then whenever he went S41-29 kd-pehtamdn e-nikohteyin 'I heard you chop- wawesiyiwa. presently, his wife dressed ping'; etc. away, up.' of season often show the iterative The uses of conjunct forms may be grouped into Expressions (cf. S253-16 'in summer-time'; four basic types. While most of these are found 5.333), e.g., niyipiniyikih S254-2 'in winter-time, winter.' in all conjunct modes, they are most clearly seen in peponiyikih every the changed conjunct. 5.34. Order Narrative: where main and subordinate clause Imperative show no agreement of referents, e.g., P264-32 mwehci There are two subcategories, IMMEDIATE and eh-dpihtd-piponiyik, eh-ati-tipiskdyik, wsadmew wiwa. DELAYED. Imperative forms are used for commands, 'Exactly in the middle of winter (0'), at nightfall exhortations, etc. (0'), he asked his wife (3') to go with him (3-(3')).' The delayed mode indicates that the command, Participial: where there is some agreement, between exhortation, etc. is to be obeyed not immediately but clauses, of overt or covert referents, e.g., S237-5 at a later point in time. Most typically, it is found k-dyztawihkwakanetwihtikow 'the two-faced Windigo'; together with a conditional clause, e.g., S254-13 P262-27 kitimakisindwdw kd-wi-kakwe-nipahiyek. . . ." itdhkan, mayaw wdpamacih; '". . ." do you 'You are pitiable (2p) who mean to try to kill me then say to him, as soon as you see him.' But it is (2p-l).' by no means restricted to such a context; consider Substantive: where a clause functions as the adjunct S247-33 haw, awdsis, oma nawaci; micihkahk oma of a verb, e.g., P2-14 kitahtawe peyak kih-pawdtam (3) otakisih. 'Come, child, roast this (immediate); let e-wih-kapayit (3') moniyaw-iyiniwa (3') wapiski- us later eat (delayed) this tripe.' -wiyasah (3'). 'Then at one time a certain man The marker of the delayed imperative is /Lhk/; it dreamt (3) that the Canadian (3'), the White Man is followed by the same person markers as are found (3') would land here (3').' in the conjunct order. For Proto Algonquian, Focal: where the predication is a particle expression, Bloomfield (1946: p. 100) sets up a special "prohibi- e.g., P256-19 namoya eh-ahkosiydn. 'It is not that tive" order to which the delayed imperative paradigm I am ill'; S40-4 kZtahtawewdskahikan kah-otihtahk. of Cree would correspond. VOL. 63, PT. 5, 1973] VERB INFLECTION 47 5.4. AFFIX POSITIONCLASSES order it occurs in the 21-form of the immediate mode and the mode The position classes of verbal affixes correspond (5.671) throughout delayed (5.672). closely, as far as applicable, to those of the nominal 5.422. /ekw/-/eko/-0 marks inverse action ex- in affixes; cf. 3.1. The present section may also serve as cept the you-and-me set. /ekw/ occurs in all an index of morphemes. inverse forms of the independent order (5.621, 5.622). In the conjunct order, the zero alternant occurs in Morphophonemic rules are given in appendix A. those forms of the mixed set whose non-third referent is singular (5.662). /ekw/-/eko/ occur in all other The personal prefixes are described in detail in inverse forms of the conjunct order (5.661); for the section 2.11. They occur in the independent order extended form /ekow/ see 5.663. only. The TA inanimate actor paradigm (5.83) is clearly The suffix position classes and their order are based on /ekw/-/eko/; the relation between /ekw/ in summarized below. The brief labels used this and the suffix of the TA indefinite actor paradigm list are intended as approximations only. (5.84), /ekawi/, remains to be explored. 1 thematic obviative sign /em/ Even in nominal and verbal derivation, "inverse 2 theme signs direction" (2.5) is typically expressed by suffixes 3 thematic obviative sign /eyi/ which include /ek/ (6.418), /ekw/ and /ek6w/ 4 mode signs: h- and ht- preterit, delayed imperative (6.431), etc. This link with derivation might eventu- 5 non-third person suffixes ally complement semantic evidence for the apparent tendency of the direction category to develop from an 6modemode signs:p-prterit dubitativedubitive 7 signs: p-preterit, inflectional to a derivational phenomenon; cf. also 5.664. 8 third person suffixes in 9 third person plural and obviative suffixes 5.423. /i/ marks direct action the you-and-me 10 mode signs: subjunctive and iterative set (5.623, 5.64). 5.424. /et/-/eti/ marks inverse action in the 2 One of the theme signs of position (5.422) and the you-and-me set (5.623, 5.64). /et/ occurs in the obviative suffix of 3 differ from position (5.43) the conjunct, and /eti/ in the independent order. (Note of the suffixes their remainder by nearly universal that an alternant /eti/ is indicated by the TA 1-2 occurrence in different orders, paradigms, word classes h-preterit form -itih where the theme sign is followed Positions 1 to 3 are (5.2, 3.22, 6.431). conveniently directly by the preterit marker -h; /eti/ cannot be termed "thematic." interpreted as /et/ followed by connective /i/ because of the palatalizing effect of the latter.) 5.41. Suffix Position 1: Thematic obviative Sign /em/ The thematic obviative sign /em/ occurs in the direct 5.43. Suffix Position 3: Thematic Obviative Sign /eyi/ sets of the TA see 5.633 and 5.663. paradigm; (Note The thematic obviativesign /eyi/ occurs in obviative the homonymy of this morpheme with the possessive verb forms which do not also involve a non-third theme marker of nouns; cf. 3.21.) person. (An exception to this statement is the TA 3'-1, 2 ending -ikoyiwa which is discussed in 5.65.) 5.42. Suffix Position 2: Theme Signs Apart from its wide occurrence in different orders, Only the theme signs of the TA paradigm are listed paradigms, and word classes (cf. 5.4), the thematic here; those of the TI paradigm are described in 5.71. nature of /eyi/ is further indicated by its occurrence (For the terms mixed, third-person, and you-and-me in the h-preterit which has no personal endings. see 5.61.) 5.421. /a/-/e/-0 mark direct action except in 5.44. Suffix Position 4: Mode Signs: the you-and-me set (5.621, 5.622). h- and ht-preterit,Delayed Imperative occurs in the set of the /e/ third-person independent The mode of the order signs delayed imperative, /Lhk/ (5.61). the and the ht- Zero occurs in those forms of the mixed set of the (5.672); h-preterit, /h/ (5.651); preterit, /htay/ (5.652). conjunct order which involve a non-third singular referent (5.662) and in the mixed 2- and 2p-forms of 5.45. Suffix Position 5: Non-Third Person Suffixes the immediate imperative (5.671). /a/ occurs elsewhere, namely in the mixed set of 5.451. The suffixes /enan/, /enaw/, and /ewaw/ the independent indicative (5.65) and throughout the may be viewed in two different ways. On the one independent preterit (5.651, 5.652). In the conjunct hand, /enan/, /enaw/, and /ewaw/ are non-third order, it occurs in the third-person forms (5.61) and person markers, alternating with extended alternants in those forms of the mixed set which involve a /enanaw/ and /enawaw/ and, in the case of /enan/, non-third plural referent (5.661). In the imperative occurring also in the immediate imperative which has 48 WOLFART: PLAINS CREE [TRANS. AMER. PHIL. SOC. no personal prefixes. The alternant forms are clearly an AI stem in i, e, d, o (5.52) combines with /an/, the result of paradigmatic leveling. /ahk/, or /ekw/. Moreover, the insertion of /y/ On the other hand, the use of /enan/, /enaw/, and prevents the merging of vowels elsewhere in the /ewaw/ parallels that of the same suffixes when they paradigm, or with short-vowel stems. The closely occur in the possessive paradigm of nouns, i.e. in the parallel TI paradigm with its consonantal theme sign inner layer of nominal affixation (cf. 3.22). In this may also be mentioned. Outside of the AI paradigm, function they are mere pluralizers which pluralize the /y/-alternants occur in the conjunct order of the TA personal prefixes. paradigm (5.661, 5.663); the TA indefinite and in- The purely pluralizing function of /ewaw/, and the animate actor paradigms (5.83, 5.84) actually use hierarchical structuring of affixation it indicates, be- the AI endings (as is seen by the independent order comes obvious only where /ewaw/ pluralizes the third forms). person prefix o-. (Since the use of this prefix in Cree /eht/-/hk/ 'indf.' /eht/ occurs in the TA verb inflection is restricted to a few archaic and rare (5.662), /hk/ in the AI and TI paradigms (5.72). forms, /ewaw/ has come to be associated almost /an/,/ak/-,/it/ '1.' /ak/ and /it/ occur in the exclusively with the second person. Except for direct and inverse sets of the TA paradigm (5.662), these rare forms, the third person is expressed in the /an/ occurs elsewhere. outer layer of affixation and is pluralized by /k/; cf. /an/-/at/ /esk/ '2.' /at/ and /esk/ occur in 5.48.) In verb inflection, the third person prefix o- the direct and inverse sets of the TA paradigm (5.662); occurs only in the h- and ht-preterits, and the 3p /an/ occurs elsewhere. /an/ also occurs in the form of the h-preterit is not attested. Thus, only the delayed imperative (5.672). ht-preterit remains, and there we actually find the /ahk/ 'ip.' See also 5.454. /ahk/ also occurs in affix combination /o- -ewaw/; for examples see 5.652 the delayed imperative (5.672). and 5.7. /ahkw/ '21.' See also 5.454. /ahkw/ also occurs 5.452. The non-third person suffixes of the in- in the delayed imperative (5.672). dependent order (5.65). /ekw/-/akw(k)/ '2p.' /akw(k)/ occurs in the '0 'indf.' 1-2p form of the TA conjunct (5.642); /ekw/ occurs /n/'0 '1, 2.' The zero alternant occurs il the elsewhere. /ekw/ also occurs in the delayed impera- direct and inverse sets of the TA independent, /n/ tive (5.672). elsewhere.56 /n/ also occurs in the 2-1 form of the 5.454. The markers of lp and 21 in the conjunct TA immediate imperative (5.643). order, /ahk/ and /ahkw/, are subject to some partial /enan/ 'ip.' /enan/ occurs also in the TA im- syncretism; both occur with and without final /w/. mediate imperative (5.642). Historically they have the shapes cited above; see /enaw/-/enanaw/ '21.' /enanaw/ occurs besides Bloomfield, 1946: pp. 101, 102. However, the James /enaw/ in the AI and TI paradigms (5.72), /enaw/ Bay dialect as described by Ellis is the only one to occurs elsewhere. consistently reflect this situation. In Plains Cree, /ewaw/-/enawaw/ '2p.' /enawaw/ occurs in the the above forms seem to be most frequent but the AI and TI paradigms (5.72) and in the you-and-me other two possibilities, /ahkw/ and /ahk/, also occur. set of the TA paradigm (5.64); /ewaw/ occurs else- (On a rather weak statistical basis it would seem that where. (Cf. also 5.451 above; for the near-homony- /ahkw/ is particularly typical of the Saddle Lake- mous plural marker of some conjunct modes cf. 5.48.) Sweet Grass area of northeastern Alberta and north- 5.453. The non-third person suffixes of the conjunct western Saskatchewan. A converse statement about order. /ahk/ cannot be made since it occurs freely in the Most of the morphemes of this class which begin Hobbema dialect of central Alberta.) in a vowel also have an alternant beginning in /y/. As in other instances, the missionary sources show /y/-alternants are attested for /an, an, ahk, ahkw, some variety. Edwards consistently gives /ahk/; on ekw/. (For the endings /iyit/ and /iyesk/ see the other hand, while she reports only /ahkw/ for the 5.663.) The /y/-less forms, which occur in the TI conjunct indicative (1954: p. 41-3), she gives /ahk/ conjunct, in the delayed imperative, and in the besides the more "general" /ahkw/ for the conjunct (inverse) you-and-me forms of the TA paradigm are subjunctive (p. 57-2). Hunter exhibits a fairly taken as basic. A model for the emergence of the complete lack of predictability of the final /w/. /y/-forms by paradigmatic leveling is easily found: Lacombe, by contrast, stands out by the remarkable /y/ is regularly inserted when long vowels follow each symmetry of his description. Lacombe gives all other; cf. appendix A: 3.1. Such is the case when four possibilities for the AI and TI, and only /ahk/ and /ahk/ in the TA paradigm. Lessard gives only 66 While a suffix /en/ would be supported by the parallelism of the /w/-less forms throughout. the other non-third person suffixes, this argument is considered 5.455. The non-third suffixes of the insufficient. Moreover, the plural suffixes show a much wider person impera- range of occurrence than /n/, e.g., in the mixed forms of the TA tive order. paradigm and in the possessive paradigm of nouns (3.2). In the you-and-me set of the TA immediate impera- VOL. 63, PT. 5, 19731 VERB INFLECTION 49 tive we find the markers of the independent order, and restricted; cf. 5.451. Instead, in verbs, it seems on throughout the delayed imperative of all paradigms the way to becoming a person marker (rather than there appear the conjunct person markers of the a mere pluralizer) for the second person plural exclu- non-third persons; see 5.67. sive; cf. 5.451. In either of these verbal functions /i/-/h/ '2.' /i/ occurs in the mixed set of the TA it occurs in the independent order. paradigm, /h/ in the AI and TI paradigms. /waw/ alternates with the position 9 plural marker /tan/ '21.' /k/ and /ih/ in some modes of the conjunct order; /ehkw/-/k/ '2p.' /ehkw/ occurs in the mixed for details see below. set of the TA paradigm, /k/ elsewhere (including the 5.481. The plural marker of animate third persons 2p-l form of the TA). is /k/-/waw/;57 note that in non-independent forms /k/ is usually preceded by connective /i/.58 /waw/ 5.46. Suffix Positions 6 and 7: Mode Signs: occurs in the subjunctive and iterative modes of the p-preterit, Dubitative conjunct order (5.33); /k/ occurs elsewhere. (In some dialects, notably in that of northern Alberta, of the The mode signs of the p-preterit, /Lpan/, and the distribution of /waw/ is much wider, at the dubitative to occur between the non-third and appear expense of /k/; cf. 5.33.) third suffixes. the data do person Although present 5.482. The of inanimate third persons is not a more detailed statement there plural support (cf. 5.31), marked by /ih/-/waw/ in the conjunct order; the is some evidence in Ellis's that the paradigms (1971) third person endings of the independent order have dubitative marker or indeed surrounds (dis- precedes not been (5.473). /waw/ occurs in the continuous the marker. segmented morpheme) preterit subjunctive and iterative modes, where it is followed by the position 10 subjunctive and iterative morpheme 5.47. Position 8: Third Person Suffixes Suffix /ih/; cf. 5.74. The /ih/-alternant of the plural 5.471. In the independent order the marker of the marker occurs elsewhere; in a purely synchronic animate third person occurs as /wa/-/a/. framework, /ih/ cannot be further segmented. In the AI and TI paradigms only /wa/ occurs, even 5.483. The obviative marker /h/ occurs in the after stem final /n/ or theme final /m/. In the TA independent order only (but cf. 5.663). paradigm, however, /wa/ occurs after vowels, and /a/ after consonants. (Note that those third person 5.49. Suffix Position 10: Mode Signs: forms of the ht-preterit which show the prefix o- do Subjunctive, Iterative not have a person marker of position 8.) The mode of the and iterative of the 5.472. In the conjunct and imperative orders the sign subjunctive order, cf. 5.33. animate third person is marked by /t/-/k//0. conjunct /ih/; occurs in the third forms of the TA In a synchronic study it is not possible to segment /t/ person this for instance into an actual and in the AI and TI paradigms except morpheme further, paradigm and iterative where it is by a nasal (5.73); subjunctive morpheme /h/ preceded immediately preceded an connective in these environments /k/ is selected instead. The by empty morpheme, namely /i/. zero alternant occurs in the mixed forms of the In the environment of /ih/, the third person is 5.48 and conjunct and throughout the imperative order. pluralized by /waw/ (cf. 5.74). 5.473. The inanimate third person (5.74) is marked 5.5. STEM SHAPES by /k/ in the conjunct order. The endings of the independent order cannot be segmented (cf. 3.31); The morphology of stem final elements is sketched third thus /wi/ marks the singular, /wah/ the plural in 6.43 and 6.44. The present section is concerned person. with their phonemic shapes. Statistical information is considered important in 5.48. Position 9: Third Person Plural Suffix placing the various stem-types into proper perspec- and Obviative Suffixes tive.59 For example, while AI n-stems are quite The historical relation of the near-homonymous prominent grammatically and in terms of text fre- morphemes /ewaw/ and /waw/ remains to be fully to Proto 57 But see also 5.642. investigated. /ewaw/ corresponds Algon- 68 52 and have 2.13 and 1946: Jones (1971: p. ff.) Piggott (1971a: p. 32) pre- quian *-wdw- (cf. Bloomfield, p. 96) sented arguments for a very different interpretation of this /i/ whereas /waw/ seems to correspond to Proto Algon- in the Odawa dialect of Ojibwa. quian *-wd- (Bloomfield, 1946: p. 101). 59The frequency information included in this section is to be the second and third considered highly tentative; for example, the lexicon at present /ewaw/ pluralizes person includes some entries which cause distortions. 3.22. In where it duplicate slight possessors of nouns; cf. verbs, It is based on a preliminary inverse stem lexicon of about 11,000 occupies suffix position 5, its occurrence as pluralizer entries which is among the early results of a computational pro- of a third person expressed by a prefix is highly ject described in Wolfart and Pardo, 1972. 50 WOLFART: PLAINS CREE [TRANS. AMER. PHIL. SOC.

quency, they constitute only 4 per cent of all AI In addition, the stems ending in clusters are as follows: stems. 60 Approximate percentage 5.51. Transitive Stems of TI stems hk 0.4 Transitive verb stems end in a non-syllabic or a sk 7.2 cluster of non-syllabics, e.g., ht 21 TA wdpam- 'see' st 2.1 pakamahw- 'strike' nipah- 'kill,' etc.; 5.52. Animate Intransitive Stems TI wapaht- 'see' Animate intransitive (AI) verb stems end in a pakamah- 'strike' vowel or n, e.g. sakin- 'seize,' etc. api- 'sit' 'see oneself' TA and TI stems are followed by theme signs which wdpamiso- are described in sections 5.42 and 5.71. tapasi- 'flee' 'walk Transitive animate (TA) verbs end in the following pimohte- along' single non-syllabics: nipa- 'sleep' pasiko- 'get up' pimisin- 'lie.' Approximate percentage of TA stems The AI stem types follow: h 18.3 m 17.2 Approximate percentage n 12.5 of AI stems s 0.07 i 29.2 t 11.9 0 25.4 w 24.6 3.2 y 0.35 e 27.4 a 10.9 In addition, the stems ending in clusters are as o 0.03 follows: n 3.9

Approximate percentage e-stems have stem alternants in e of TA stems and a. The alternant in a occurs in the non-third forms of the ht 0.76 independent indicative and throughout the inde- st 0.07 pendent preterit; the alternant in e occurs elsewhere. hw 11.1 (However, the distribution of the alternants is not mw 0.28 entirely stable. Thus we find S14-3 ayamihdhkan pw 0.07 (AI 2 delayed imperative), and T101-6 atoskdsiw spiv 0.07 (diminutive) beside T102pl acoskesiw.) swSW 2.7 n-stems have alternants in n and ni. The alternant in n occurs in the 3 and 3p forms of the independent verbs end in the Transitive inanimate (TI) following and conjunct orders; the alternant with final ni is single non-syllabics: found in the first and second person forms of the conjunct order. All other instances of n followed by i Approximate percentage are indeterminate since the i reflect the suffix- of TI stems may initial /e/ or, in the case of imperative or derivational h 20.1 suffixes with an initial consonant, connective /i/. n 22.7 That the first and second person forms of the conjunct s 5.1 order actually have the ni-alternant rather than n t 21.4 plus connective /i/ follows from two observations: (1) These forms have the y-alternant of the suffixes 60Consider also Ellis's recent statement (1971: p. 84) that (5.453) which normally occurs after vocalic stems or "most II stems end in /n/ but a few end in one of the vowels suffixes. (2) Lacombe cites sets of forms with A:hile there of course be dialect a parallel /i, e, a/." may differences, n and ni 'that I lie and survey of over 700 II stems shows that vowel stems (in i, o, e, a) (pimisindn, pimisiniydn down') constitute fully two-thirds of all II stems, and that only one-third Ellis (1971: p. 83) actually states the n-form to be are n-stems. more common in James Bay Cree. Thus, the histori- VOL. 63, PT. 5, 1973] VERB INFLECTION 51 cal development seems clear: the basic stem ended interest would be small compared to the effort of in n. But with i (representing suffix-initial /e/ or constructing it. connective /i/) occurring in most of the forms, and The transitive animate paradigm is subject to a few with all other AI verbs ending in a vowel and thus general constraints which do not derive from any showing the y-alternants of the first- and second- properties of the system itself but are descriptively person conjunct suffixes, the pattern pressure was prior to it. considerable and resulted in a new form in ni. Thus, no may occur as part of both actor and goal; i.e., forms such as "1-1" etc. 5.53. Inanimate Intransitive Stems are ungrammatical. Reflexivity and reciprocality are expressed derivationally by certain animate Inanimate intransitive (II) verb stems end in a intransitive (AI) stems where these relations are part vowel or n, e.g., of the meaning of the stem (6.437, 6.438). For third persons a situation like the above does ponipayi- 'stop running' not arise since only one of the referents of a third kikamo- 'be attached' person form (cf. 5.612) is morphologically and signifi- 'be timz- deep' catively specific (5.61, 5.63). In reference, however, 'be wet' sapope- the various third persons differ among themselves; misd- 'be big' cf. 2.01, 2.2. Only one non-obviative referent occurs otdkosin- 'be evening' in any one span, except in parataxis. Of obviative referents, on the other hand, there be The II stem types follow: may any number, at least theoretically. Thus 3'-3' is perfectly Approximate percentage regular in reference although not fully expressed in of II stems morphology and signification; see 5.63. i 17.6 There is only an indefinite actor, no indefinite goal. 2.1 Further, there are forms only for an indefinite actor 1.7 acting on third persons; a suppletive paradigm is used e 19 for the indefinite actor acting on non-third persons; 23.8 see 5.84. n 35.8 5.61. Morphological Structure and Syntactic Function 5.6. THE TRANSITIVE ANIMATE (TA) PARADIGM In syntactic and referential function, all transitive animate forms are alike: anaphoric reference is made In the of our we em- considering scope analysis, to both an actor and a goal. (For double goal verbs with Gleason's remark phatically disagree (1961: pp. see 5.11.) Except for the indefinite actor, these that the TA "fail to show 117, 119) paradigms clearly referents may be expressed by nominal or verbal recurrent of structure The any regularities .... adjuncts. A different situation emerges, however, if as stand are No paradigms they unanalyzable." morphological structure and meaning (rather than attempt is made to carry recurrent partial analysis reference) are considered. to its ultimate limits (such as trying to combine part 5.611. Let us first direct our attention to those of the first person plural morphemes of independent forms whose reference involves both a third and a and But most of and conjunct, /en-an/ /dh-k/). non-third person, e.g. niwapamdw 'I see him.' Here the thematic and affixes can be ex- "pronominal" we find both referents expressed morphologically, tracted and be consistent may assigned meanings namely by the personal prefix ni- 'I' and by the with the structure of the as a whole. paradigms suffix /wa/ 'he (proximate).' The direction of the Neither do we to construct a attempt complete action (2.5) is indicated by the theme sign /a/. (In structural account of the which paradigmatic system conjunct order forms the third person morpheme is would the of all forms. This permit prediction realized as zero but is regularly pluralized, by /k/ or restriction is due to the of only partly inadequacy /waw/.) Alluding to the presence of both third and the data and the intricacies of such a com- handling non-third persons in the morphological structure (and At least as is the considera- plex system. important in the meaning), we call this the MIXEDset. tion that the current Cree system is well known to be 5.612. If we now consider forms whose reference the result of extensive paradigmatic leveling.61 Thus, covers third persons exclusively, we find that only to far the within a go beyond present analysis purely one of the referents is expressed morphologically. framework a synchronic might yield mechanically Thus, in wapamew 'he sees him' only the actor is feasible account whose structural and historical expressed, by /wa/; /e/ indicates the direction of the action. refer to these as THIRD-PERSON 61 Cf. Michelson, 1912; Bloomfield, 1946; Goddard, 1967; the We forms. dialect data further indicate the extent and sometimes even the The remaining forms of the transitive animate direction of these developments; cf. 5.664. paradigm are referred to as the YOU-AND-ME set; they 52 WOLFART: PLAINS CREE [TRANS. AMER. PHIL. SOC. involve non-third persons exclusively. The morpho- linguistics in general but also provide an explanation logical structure of the you-and-me forms is discussed for the difference of the mixed and third-person sets in 5.64. of the Cree TA paradigm. 5.613. Consider the third-person forms: Thus, synchronic and diachronic evidence alike support the conclusion that in the third-person forms independent conjunct of the Cree TA paradigm, only one of the referents is direct -ew, -imew -at, -imdt expressed morphologically. -ewak, -imewak -dcik, -imdcik 5.615. The existence of a set with one expressed -eyiwa -dyit referent (namely the third-person set) alongside the more amply represented set having two expressed inverse -ik -ikot referents (namely the mixed set) is by no means -ikwak -ikocik universally recognized among those who have studied -ikoyiwa -ikoyit these paradigms. Since some analyses have been a few critical remarks to be in order. alone will the published, appear Inspection suggest primary analysis Most of is the of into theme 5.42 and and into one set widely known, course, analysis signs (cf. 5.62) Gleason He tried to find both of markers for each form. (1961: pp. 116-122). person-number-obviation referents in all TA forms but since he omits the The latter are, for the independent order: /wa/, prefixes altogether, his efforts are misguided from the /wa-k/, /eyi-wa-h/; for the conjunct order: /t/, outset, and not only with regard to the third-person t-Ei]-k/, /eyi-t/. forms. These same sets of person-number-obviation Much more striking yet is the analysis of Pittman markers are found in the animate intransitive (AI) (1965) which overwhelmingly illustrates the principal paradigm which throughout expresses only one danger of segmental analysis of paradigmatic forms. referent. The independent endings also parallel the Once a certain pattern has been discovered in part of forms of the possessive paradigm of nouns; cf. 3.22. the paradigm, it is extrapolated to the entire paradigm Thus, the forms themselves (although they might and no amount of empirical data may stem the conceivably permit it) clearly do not require analysis analyst's progress. Consider one example: for 3-3' into two person morphemes. ew there is contrived "a reconstructed 'ideal' (mor- 5.614. To Goddard (1967) we owe the comparative pheme-by-morpheme) form" iko-ima-awa which is background and historical perspective which corrobo- translated as "3 subject-3' object-3'." A "morpho- rate the present analysis of the third-person set. phonemic (contraction) rule" lets this iko-ima-dwa Goddard has out the existence, in some parts pointed become ew, and there are no fewer than twenty-five of the Ojibwa verb system, of two contrasting para- such rules. Surely Pittman cannot mean what he digms which differ as to whether or not the goal of seems to be saying, but equally surely his analysis the action is specified morphologically and signifi- not only fails to enlighten the reader but further catively. obfuscates a problem which is sufficiently complex in An ABSOLUTE form only specifies the actor whereas its own right. an OBJECTIVE form specifies both actor and goal. For in the 'he believes it' example, objective otepwettdn 5.62. Direction the actor is expressed by the personal prefix o- and the goal by a suffix /i/ (which does not appear in the Among the person categories, non-third persons phonemic representation, but whose presence is outrank third persons; within the former class, those confirmed by the prefix). In the absolute tepwettam including the addressee take precedence over those 'he believes,' by contrast, there is no prefix; the actor which do not include it. This principle of ordering is expressed by the suffix /wa/ (which again does not (precedence) is found not only among the personal appear in the phonemic representation). Bloom- prefixes which are members of one position class field's examples for this contrast are objective nent- (thus, paradigmatically) but also in the linear sequence indpentdn 'I see it thus' and absolute nentindpentam of verbal and nominal affixes (syntagmatically). 'I have such a vision' (1958: p. 34). (For further details see 2.11 and 2.5.) A "living opposition" (Goddard, 1967: p. 71) of The relative order of the various sets of affixes absolute and objective paradigms is found only in a which manifest the person categories in verbal in- small part of the Ojibwa verb system. Traces of flection is fixed. The one-third persons are repre- this opposition, however, are found so widely among sented by markers in two positions, namely the prefix Algonquian languages that Goddard was led to position and suffix position 5. With one exception,62 postulate a double system for the independent indicative of the TA and TI of Proto 62 Where the third person is marked by the prefix o- (i.e. in the paradigms forms of the h- and it is Such a double of absolute and third-person ht-preterits; cf. 5.652), Algonquian. system pluralized by the position 5 morpheme /ewaw/; but this use of objective forms would not only throw light on a /ewaw/ as a verbal third person morpheme is extremely re- number of recalcitrant problems in Algonquian stricted; cf. 5.451. VOL. 63, PT. 5, 1973] VERB INFLECTION 53 the third person is represented by suffixes of position 8 third-person forms do not specify as to the categories (cf. 5.47); it takes the plural and obviative markers of of number and obviation, that referent which is not position 9. expressed morphologically. (Its gender, on the other Since the order of the affix position classes is fixed hand, is animate by virtue of the meaning of the and cannot be reversed, the direction of the action has paradigm as a whole.) to be indicated independently. The themes of the This conclusion leads to a closer investigation of TA paradigm function as direction markers and this the goals (in direct forms) and actors (in inverse is, indeed, the only way in which the function of the forms) which have traditionally been assigned to themes was found to be synchronically relevant. (It third-person forms. is noteworthy that no meanings have been established 5.631. Most studies of Cree grammar assume the for the TI theme signs, where there is no contrast of existence of two obviative categories, a "nearer" and a direction; cf. 5.71.) We use the term direct for an "farther" (Bloomfield) or "further" (Hockett, Ellis) action which corresponds to the left-to-right sequence obviative. Apart from comparative considerations, of the classes, that is, to the actual sequence of the this distinction seems to have been predicated, above affixes in time when uttered. Inverse is used for all, on the assumption that each TA verb form must action in the opposite direction. See also 2.5. have a specific actor and a specific goal, and that these 5.621. For all mixed forms, i.e., those involving belong to different categories. Thus, if the actor of third as well as non-third person referents, the direct- -eyiwa, for example, is obviative, its goal may not also action marker is /a/ which alternates with zero in the be just obviative. We have seen that one of the indf, 1, and 2 forms of the conjunct order (5.662) and premises of this argument is false. in the 2 and 2p (but not the 21) forms of the im- A second possible source for the assumption of two mediate imperative (5.671). The corresponding in- obviative categories is the presence of two obviative verse marker is /ekw/ which also alternates with zero morphemes in -eyiwa and -ikoyiwa, /eyi/ and final /h/. in the 1 and 2 forms of the conjunct order. In the possessive inflection of nouns (3.22), these 5.622. The dimension of direction is of prime morphemes actually refer to different persons: /eyi/ importance in third-person forms. The one referent marks the obviative status of the possessor, and /h/ which is morphologically expressed and significatively that of the noun possessed. However, that a similar specific in direct forms functions as actor, and as goal analysis does not hold for the verb forms under in inverse forms. consideration is clear from the fact that the AI and TI Thus, direct third-person forms express all the paradigms which are both one-place verbs have the usual distinctions (proximate singular, proximate same complex ending for the obviative actor, /eyi- plural, obviative) for the actor but leave the goal -wa-h/. Furthermore, only one obviative morpheme completely unspecified: -ew, -ewak, -eyiwa, etc. The occurs in the corresponding conjunct forms, /a-eyi-t/ direction markers are /e/ for the independent and /a/ and /ekw-eyi-t/. for the conjunct order. Inverse forms, on the other Yet another basis for the distinction of two obvia- these same distinctions for their hand, express goal tive categories may be thought to lie in the existence, but leave the actor unspecific: -ik, -ikwak, -ikoyiwa, parallel to each other, of such third-person forms as etc. The direction marker for both independent and -ew and -imew. Since -ew is glossed 'proximate acting is See also 5.613. conjunct /ekw/. on obviative,' -imew would seem to have the proximate 5.623. The intrinsic of the cate- ordering person acting on yet a further category. An attempt is made gories and the same distinction of direct and inverse in 5.633 to account for the function of -im- and for is also found in the set. you-and-me Action by a the difference in meaning between the two sets in a person which includes the addressee is direct, action different manner. such a is inverse. Direct action is marked upon person If the three lines of reasoning suggested above by /i/, the inverse marker is /et/ /eti/ (5.424). underlie the distinction of two obviative The structure of the is actually you-and-me subparadigm categories, and if the Cree evidence adduced in considered in detail in 5.64. refutation is accepted, then the distinction may safely be abolished for Cree; cf. also 2.24. 5.63. Meaning and Reference in the Third- 5.632. In the dimension of obviation, "obviative" is Person Subparadigm clearly the marked member and "proximate" the It was argued in 5.61 that third-person forms unmarked (cf. 2.23). This is evident not only from express only one referent. This analysis was based on the inflectional paradigms of nouns and one-place their morphological structure and on the patterning verbs (all types but TA) but also from the TA para- of third-person forms in the other Cree paradigms, and digm itself. The obviative forms of the independent corroborated by comparative evidence. It was order add the morpheme /h/ and, more importantly, further seen in 5.62 that the feature of direction marks the direct forms of the mixed set of all orders obliga- this one expressed referent as either actor or goal. torily mark the occurrence of an obviative referent From these premises it would seem to follow that by /em/. 54 WOLFART: PLAINS CREE [TRANS. AMER. PHIL. SOC.

Syntactically and semantically the unmarked, more inflected for an obviative possessor and thus very general nature of the non-obviative category is specifically obviative, e.g., obvious from the fact that it occurs in contexts of neutralization. Thus a verb which has both a P126-19 ketahtawe ekwah ocihciyiw then hand(0) proximate and an obviative adjunct is itself inflected presently his(3') for proximate plural, e.g., otinam; ocahcanisiyiwa otinimew. take(TI 3) his(3') ring(3') take(TA 3-3') T 46p 2 . . . mdnokwewak, owikimdkana camp(AI 3p) his(3) wife(3') 'Presently she took his hand; she took his ring'; e-wTcihikot. S185-10 . . .; nehi kd-ks-mescihimdt help (TA (3')-3) that (3') annihilate (TA 3-3') . . . set his wife they up camp, helping him'; otawdsimisiyiwa notokesiwa, eyakoni her(3') children(3') old woman(3') that one(3') T 7p 4 ekosi tdpwe e-kS-isMhcikecik thus really settle things thus(AI 3p) eh-totdkot. do something(TA (3')-3) ohi okimdwa, ekwa sipwehtew. this(3') king(3') and leave(AI 3) . .; it was yon old woman whose children he had wiped out, that one was doing this to him.' 'So they had really made a deal, he and that king, and then he left.' But even where the referent is highly marked as obviative, the /em/-form is not obligatory. The The opposition of obviative and non-obviative is following example shows the /em/-less form used discussed in some detail in 2.23; we use the term with a noun inflected for obviative possessor: "proximate" for the narrow meaning of the non- obviative category, and "third person" for its wide T104p8 ekwa etokwe, asawdpiw meaning. then I guess look around (AI 3) 5.633. In direct third-person forms, a constraint ekwa awa ndpew on the range of the morphologically unexpressed then this (3) man (3) referent is sometimes the (the goal) provided by e-kZ-nitawi-mescihdt nehi morpheme /em/ which specifically marks it as obviative.63 go to annihilate(TA 3-(3')) those (3') In the mixed set, /em/ occurs in all direct forms ondpemiyiwa, Jhi wiwa, . . . involving an obviative goal. While the obviative her (3') lovers (3') this(3') his(3) wife(3') category is doubly expressed in the independent it this man looked around and order, namely by /em/ and /h/, the occurrence of 'Then, seems, then went to annihilate those lovers of his . /em/ is its only sign in the conjunct order. wife's, .. In the third-person set, by contrast, /em/-less The hypothesis that /em/ occurs only with specifi- endings like -ew freely occur with obviatives as their cally obviative goals finds strong support in the analysis second, morphologically unexpressed referent. /em/- of obviative and proximate/third person as marked forms like -imew occur only where a referent is highly and unmarked members of an opposition (5.632, marked as obviative,64 i.e., typically where a noun is 2.23). Using just two endings as an example the, situation can be diagrammed as follows: 63 This also seems to be Bloomfield's view of the function of /em/ although it is neither made fully explicit nor well exemplified (1946: pp. 98, 102). Curiously enough, I cannot find forms in- volving both non-third and obviative referents (let alone third- person /em/-forms) in Bloomfield's descriptions of Menomini, Fox, and Eastern Ojibwa; cf. also Goddard, 1967: fn. 13. Some non-third-and-obviative forms have been recorded for Potawa- tomi (Hockett, 1948: p. 142, for independent order direct set only); Northern Ojibwa (Rogers, 1963: p. 123); Fox (Voorhis, 1971: p. 70) and Kickapoo (Voorhis, 1967: pp. 99, 108, 114, 125). 64 The obviation inflection of inanimate nouns in the Mistassini dialect of Montagnais-Naskapi (Rogers, 1960: p. 110) provides an interesting parallel. A general third-person form, which may be used for both proximate and obviative, is in free variation with a distinctly obviative form. Thus, niclwdpahten t6 'I saw the canoe (0)'; czwdpahtam 5t or otiyiw 'he saw the canoe (0 or 0').' But if an inanimate noun occurs in collocation with an obviative, e.g. as the goal of a verb with obviative actor, then only the ex- plicitly obviative form occurs, e.g., okosisa ciwdpahtam otiyiw 'his (3) son (3') saw the canoe (0')'. (-iw) VOL. 63, PT. 5, 1973] VERB INFLECTION 55

Thus, the third-person forms without /em/, namely It is obvious on inspection that most of the person- -Fw, -ewak, -eyiwa and -at, -acik, -ayit, each have two number suffixes of the you-and-me set are identical meanings. The wide meaning covers the entire with the first and second person suffixes of the AI range of third-person goals, including the obviative; and TI paradigms; that is, the suffixes express only examples are plentiful. The narrow meaning, which one of the referents. derives from the contrast with the /em/-forms, would The independent order also provides for the expres- only cover those goals which are not obviative and sion of the second person by the prefix, ki-. But thus conflicts with the referential constraints on the even here, the suffixes do not necessarily complement co-occurrence of two non-obviatives65; there are no the prefix by marking first persons; rather, the 2p- examples. and-I forms with some redundancy express the second At the same time, the marked and unmarked person both in the prefix and in the suffix. In short, members of this particular opposition share at least the principle by which one referent rather than the one environment: unmarked /em/-less forms and other is chosen to be expressed has not been discovered. marked /em/-forms alike occur with nouns which are 5.641. All you-and-me forms conform to the same inflected for obviative possessor (see above), and we pattern of neutralizing the number distinction of the cannot yet state the factors which determine the second person in the environment of the first person selection of the marked or unmarked forms. plural. This pattern which is diagrammed below is However, if we re-consider the entire direct sub- also found in Proto Algonquian.68 a striking contrast emerges between the paradigm, 1 lp mixed set and the third-person set. In the mixed set, the normal obviative forms are highly marked; 2 they include two obviative signs. The question of an unmarked counterpart has never been raised with 2p respect to these forms.66 In the third-person set, on the other hand, the marked forms are extremely 5.642. If we first consider the 2p-and-1 and the rare67and seem to have given way to the unmarked, 2(p)-and-lp forms in the four modal categories under /em/-less forms. consideration (independent indicative, conjunct in- In summary, then, the morphological pattern of dicative, immediate and delayed imperative), it is the /em/-forms, their syntactic use, and the unmarked easily seen that they also pattern alike in their choice status of the non-obviative combine to make possible of expressing the "2p" referent in the 2p-and-1 a consistent account of the entire problem. forms and the "lp" referent in the 2(p)-and-lp forms. Part of our diagram may now be filled in: 5.64. You-and-Me Forms 1 lp The same theme signs occur in all three orders; 2 function as direction markers. /i/ marks direct they "lp" action, which in this set means action from a second 2p "2p" person on a first; note that the /i/ is lengthened when it is followed by the delayed imperative marker /Lhk/. The 2p-and-1 morphemes of the conjunct and of the /et/-/eti/ marks inverse action (5.424). immediate imperative need to be taken up individu- the of the direct 66 An apparent contradiction with the constraints of 5.6 would ally. In the conjunct order, /ekw/ rest in the difference of the frameworks: the constraints are stated set is as expected; the /akok/ of the inverse set, in terms of reference whereas here we are dealing with meaning however, is somewhat problematic. It clearly con- alone. tains the suffix connective 66 plural /k/ preceded by /i/ A highly suggestive example, which shows an obviative noun as is seen by the corresponding ending of the subjunc- with a non-obviative verb form, emphasizes the need for further research: tive mode, /akw-waw-/; note that /k/-/waw/ the third The . . kosi mdskoc namoya normally pluralize person (5.48). P 4_4nl , niyandn A -r rv of has not been without thus we perhaps not meaning /akw/ ascertained; further evidence it cannot be grouped with the 1-3 nikiskeyimiko nman mdci-manitowah. conjunct morpheme /ak/. Edwards (1954: p. 43-2) know (TA 3-1 p) evil spirit (3') 68 100. The same is attested '.., so perhaps we others are not known to the evil spirit.' Cf. Bloomfield, 1946: p. pattern in the works of Lacombe, Hunter, Lessard, and Edwards. 67 Their low frequency in current texts provides a curious con- Ellis's (1962, 1971) show an internal inconsistency trast with the fact that Hunter, Lessard, and Edwards give the paradigms which deserves further historical study: while the pattern of the full set of three third-person /em/-forms per order, e.g., -im~w, agrees with the one given here, he reports -imewak, the latter is omitted by Lacombe (and by imperative paradigm -imeyiwa; a different kind of number neutralization for the other orders, e.g., Ellis). Ellis's dialect is the only one to show the range of /em/ further -1 -lp expanded: it has inverse forms with /em/ in the mixed set of the 2- -in -inan conjunct order, e.g., the 3'-1 form -imici /em-it-Lij-h/; cf. 5.663. 2p- -inawaw -inawaw 56 WOLFART: PLAINS CREE [TRANS. AMER. PHIL. SOC. implicitly identifies this /akw/ with /ahkw/, the 21 The direction sign is followed by the suffixes of marker of the conjunct order. Not only does this the non-third referents, /enan/, /enaw/, /ewaw/, and analysis fail to take the meaning of the morphemes zero; and then by that of the third person referent, into account; it furthermore runs counter to the other /wa/-/a/, with its plural and obviative markers, published sources (which show no pre-aspiration) as /k/ and /h/. The third person obviative forms also well as to specific informant tests. include the thematic obviative morpheme /eyi/. The 2p-1 ending of the immediate imperative, /k/, The endings for 3'-1 and 3'-2 do not fit this general corresponds to the imperative ending of the AI and TI pattern. The forms one might expect would consist paradigms (5.72). (An alternative analysis might be of the inverse marker followed by the non-third thought of: to interpret /k/ as the plural marker marker zero, the third-person marker /a/, and the (5.481) rather than as a morpheme obviative marker /h/; thus, /ekw-0-a-h/. Its regular representing the immediate imperative 2p category. phonemic shape would be -ikwa alternating with -iko However, not only is /k/ primarily a third person in surface variation, and it is here that we may look pluralizer; comparative evidence also militates against for the reason for its replacement. For not only such an interpretation since the immediate imperative would it be homonymous (since final h is non-distinc- 2p morphemes of Proto Algonquian are reconstructed tive) with the 3'-1, 2 forms of the h-preterit, but it (Bloomfield, 1946: p. 100) as ending in *. . . ko.) would also fail to clearly exhibit its obviative nature. It may be noted that the pattern described here for The substituted form -ikoyiwa /ekw-eyi-wa-h/ is you-and-me forms in general again closely matches clearly marked as obviative by the thematic morpheme that of Proto Algonquian; the one set which in Proto /eyi/ (but differentiated from the 3'-goal form by the Algonquian differs from the rest, namely the im- prefixes). 70 mediate imperative, has obviously approached the 5.651. The independent h-preterit shows no personal general pattern in Cree. endings. The usual theme signs are directly followed 5.643. The 2-and-1 forms do not fit a common by the preterit marker /h/. In the the inverse set pattern. conjunct order, 5.652. The independent ht-preterit is marked by shows the first The direct person expressed, by /an/. the suffix /htay/ which follows the theme signs /a/ set the second and the expresses person, by /an/, or /ekw/. same is true of the delayed imperative (which through- Where /htay/ is followed by a suffix beginning in out uses the same person suffixes as the conjunct contraction takes A: 4.2. Since this is the clear case of differ- /e/, place; cf. appendix order). only Thus, the 3p-(3') form /o- -a-htay-ewaw/: o- -ahtawdw. entiation in the entire you-and-me paradigm, and Where is followed the suffix since the Cree situation with the /htay/ by third-person agrees general /a/ and the position 9 plural marker /k/, there is a Algonquian one as reflected in Proto Algonquian, we surface variant the in -ik. here see a to be to the semantic (conditioned by y) ending may clue, yet exploited, Thus, the form for structure of the 3p-1 /ni- -ekw-[i]-htay-a-k/, you-and-me paradigm. instance, appears either as ni- -ikohtayak or as ni- The /n/ of the independent order and the immediate the insertion of connective is as to first or second -ikohtayik; (note /i/ imperative ambiguous person, before /htay/). just as it is indeterminate in the AI and TI paradigms. In third-person forms, the third person is expressed Without doubt, the AI and TI paradigms may be by the prefix o-. Where it is expressed by this prefix, taken as the model of the emergence of /n/ in the the third person is pluralized by the suffix /ewaw/ of you-and-me set.69 On the basis of the argument of the position 5; cf. 5.451. The third person being ex- preceding paragraph, /n/ might conceivably be pressed by a prefix, no personal suffix occurs.7' assigned specific semantic values. 5.66. Order 5.65. Independent OrderEndings Conjunct Endings As with the the mixed The third-person forms are described in 5.61 to independent order, only set and its relation to the other sets remain to be 5.63, and the you-and-me forms in 5.64; the mixed the forms are described in set and its relation to the third-person set remain to be discussed; third-person discussed. 5.61 to 5.63, and the you-and-me set in 5.64. 5.661. Within the mixed there is a clear The direct theme sign has two alternants, /e/ and set, distinc- /a/. Their distribution coincides with the distinction tion between forms which involve a plural non-third of the and mixed it third-person sets; historically, 70According to Hunter, /eyi/ is also inserted in the h-preterit seems to correlate further with the absolute-objective where he gives -ikoyi /ekw-eyi-h/ as the form for 3'-1, 2. dichotomy (cf. 5.614). The inverse marker is /ekw/ 71 For the conjunct, however,where there are no prefixes,Hunter throughout. cites a p-preterit form with the 3 suffix /t/, atipan. For mixed forms involving the obviative, Hunter gives direct 69Note that a zero ending has been reconstructed for the Proto forms with /em/ but without the obviative marker /h/, e.g. Algonquian independent form whose reflex in Cree would merge 1-3' -imdhtay. The inverse forms he cites have the obviative with the reflex of the reconstructed imperative ending, *-ilo-. theme sign /eyi/ throughout, e.g., 3'-1 -ikoyihtay. VOL. 63, PT. 5, 19731 VERB INFLECTION 57 person referent, and those which do not. The forms to distinguish it from the direct theme sign /i/ of which involve a plural non-third referent are readily the you-and-me set (5.64).72 segmentable: the same direction signs as in the In some of the older sources, this theme /iy/ independent order, /a/ and /ekw/, as well as a occurs in all the inverse obviative forms of the conjunct variant form of the latter, /eko/, are followed by the order. Thus, Hunter gives only the /iy/-set whereas /y/-alternants of the conjunct non-third markers, Lacombe (1874b: p. 128) cites the /iy/-set and the /yahk/, /yahkw/, and /yekw/; cf. 5.453. The third more widely attested one side by side. person is marked by zero and pluralized by /k/ The forms for 3'-lp, 21, 2p are not fully understood. -/waw/; cf. 5.33. They seem to contain an extended form /ekow/ of There can be no doubt about the composite nature the inverse marker /ekw/ as well as the direct marker of these forms. The reason for the emergence of the /a/. If we also consider the form which Edwards /eko/- variant which results in forms like -ikoyahk gives for 3'-2, namely -ikowat, then there can be little rather than a hypothetical form -ikwahk, may be doubt that the usual direct endings are here added to sought in the pervasive occurrence of -iko- where it a derived stem which includes the inverse marker reflects /ekw/ plus suffix-initial /e/ or connective /i/. /ekw/ as part of the stem; cf. also 5.422 and 5.664. Such forms occur in the same paradigm, e.g. -ikot A specific paradigmatic motivation (as it were) for /ekw-[i]-t/; they occur in the corresponding in- the emergence of these remarkable forms has not dependent order paradigm, e.g., -ikondn /ekw- been discovered. Thus we may only suspect a rather -enan-a/; and they are also found in the TA inanimate general tendency towards more "transparent" forms actor paradigm (5.83). (cf. 5.664). It should be noted here that Lacombe, Finally, the remarkable structural similarity with Lessard, and Edwards report some corresponding the conjunct direct set deserves to be noted. forms for the independent inverse. The James Bay data of Ellis this view since 5.662. The forms involving singular non-third strikingly support there we find the obviative marker of the direct person referents, namely indf, 1, or 2, cannot be subpara- extended to the inverse. In further segmented once the third person plural marker digms fact, a fully and the obviative themes have been identified. Thus, "regular" system has evolved in the James Bay with both and the the remaining morphemes /eht/, /ak/, /at/, /it/, dialect, /em/ /h/ marking obviative the -imici and /esk/ might be viewed as portmanteau realiza- throughout conjunct order, e.g. /em-it- see 1962: 1971: tions of theme sign, non-third marker, and perhaps -[i]-h/; Ellis, appendix C; Ellis, p. 85; tables 18. even third person marker (solution I). 17, 5.664. There can be little doubt that the transitive An alternative analysis (solution II) would be to animate is in a state of consider- regard these morphemes as non-third markers only, paradigm currently able fluctuation. As had been noted and set up zero direction markers as well as zero already by third markers. The zero third marker Michelson (1912), and been clearly shown by Goddard person person the Cree TA in or is then regularly pluralized by /k/. In solution I, (1967), paradigm spite perhaps because of its is not of the on the other hand, the would seem to have apparent regularity typical pluralizer as a whole. the entire portmanteau as its domain. Algonquian languages morpheme Even such a lies outside the of It is mainly the matter of pluralization and the though study scope this the structure of the and their resulting general pattern which let solution II appear paper, paradigms seem to indicate that the direction of preferable; for further evidence in support of solution divergences II see 5.671. development in the Cree TA paradigm is from "fusional" to forms. 5.663. The obviative forms of the direct sub- "agglutinative" There also seems to take place a concomitant paradigm present no further problems; even though development from inflectional to derivational expres- no obviative marker corresponding to the /h/ of the order and of noun inflection is to be 72 Ellis reports n and I in this position for the Albany and Moose independent dialects and therefore identifies this suffix with the thematic found in the the theme conjunct order, sign /em/ obviative sign /eyi/ (1971: p. 85). But even in James Bay Cree, is sufficient to mark the obviative (but see below for the suffix under discussion palatalizes a preceding stem-final /0/. the James Bay situation). To get around this obstacle, Ellis appeals to "the analogy of its occurrence [sc. of palatalization] in the parallel forms with 3 as The obviative forms of the inverse subparadigm are actor." of a more complex structure. In the forms involving Given the low frequency of these forms, a hypothesis based on 1 or 2, namely -iyit and -iyisk, we find a theme sign essentially one other form, /-it/, is not very convincing. More which is attested in these two forms importantly, palatalization does not seem a plausible subject for /iy/ widely analogy in these rare forms. If analogy must be invoked, it (Hunter, Lacombe, Lessard; Edwards cites a form seems more likely that Swampy Cree n and Moose Cree I have -ikowat instead of -iyisk). Whether this morpheme displaced y than that /i/ should have crowded out /e/. Cree is actually or rather followed with speakers are almost universally aware of the y-n-th-l-r correspon- /iy/ /i/ by endings ence, and intrusions of one of these are not at all un- initial cannot be determined on the basis unexpected /y/ (5.453) common, e.g., Swampy Cree alikwacas 'squirrel' (cf. Proto of the present data; we tentatively set it up as /iy/ Algonquian *n). 58 WOLFART: PLAINS CREE [TRANS. AMER. PHIL. SOC. sion of certain semantic relations. Consider, for person markers or portmanteau realizations of both instance, the emergence of forms based on the "inverse person and direction remains to be answered elements" /ek, ekawi, ekw, eko, ekow/ etc. (cf. 5.422) adequately.) most of which are followed by the usual animate 5.672. The delayed forms show a remarkably intransitive (AI) endings. It does not seem unlikely regular structure: the direct theme sign /a/ and the that many of these are, in fact, derivatives. Es- delayed imperative suffix /Lhk/ are followed by the pecially in view of the extensive symmetry of the TA non-third person suffixes of the conjunct order; the paradigm (2.5), these developments indicate a third person is marked by zero and pluralized by /k/. fascinating area for further research. 5.7. THE AI, TI, AND II PARADIGMS 5.67. Imperative OrderEndings In the animate intransitive (AI), transitive in- The set is described in 5.64. The animate (TI), and inanimate intransitive (II) para- you-and-me one referent is direct set is best treated separately for the immediate digms, only expressed morphologically and modes. Both modes have (cf. 5.1). delayed perfectly TI it is convenient obviative forms with Within the AI and paradigms regular /em/. to between and non-third 5.671. In the immediate the 21- distinguish third-person imperative only forms. This distinction to the distri- forms show a non-zero theme The corresponds sign, namely /a/. bution of the stem-final alternants of some AI verbs structure of the other forms exactly parallels that (5.52), and to the distribution of the theme signs in described for the conjunct in 5.662; see also below. the TI (5.71). The 2-3 suffix is to paradigm /i/ subject apocope, e.g., The preterit endings (5.2) of the AI and TI para- plkiskwas to him' With 'speak (stem pikiskwat-). digms require little comment beyond the more stems which are and have a short monosyllabic general remarks of 5.322 and 5.652. Note that the vowel, the word-final /i/ usually remains, e.g., third forms of the which show the isi 'tell him A: 5.1. person ht-preterit T16p84 so'; cf. appendix prefix o- have no person suffix; cf. 5.451. Examples: two one with and one For 2-3p, there are endings, T20p43 opimohtahtawaw 'they walked'; T100pl2 without the third suffix added to person plural /k/ otoh-takopayihtawaw 'they would arrive from there.' 2-3 suffix. own texts show the form the My only In the h-preterit, the mode marker /h/ is preceded without thus nds them' or /k/, T20p67 'get T55p46 only by the TI theme signs and by the obviative kitimakinaw 'look with on them.' In Bloom- pity marker /eyi/. field's texts the /k/-form seems to be more typical although both are found73: S246-22 ntaw-asamik oki 5.71. TI Theme Signs dpakosisi-notokewak. 'Go feed these Old Mouse Unlike the situation in the TA paradigm (5.62, Women'; S247-20 ntaw-asam kohkominawak. 'Go 5.64), the function of the TI theme signs has not been feed our grandmothers.' established. The historical development of their The suffix of the 2p-forms, /ehkw/, occurs only distribution, however, has been sketched by Goddard here. The suffix of the 21-forms, /tan/, on the other (1967: p. 74) who sees their original function as hand, occurs here as well as in the AI and TI para- distinguishing the absolute and objective paradigms digms. Thus, /tan/ clearly marks the 21 category (5.614). specifically rather than being a portmanteau realiza- Synchronically, /e/ occurs in the non-third forms tion of 21 and 3. This fact lends support to at least of the independent order and in the 21 form of the part of our earlier analysis (5.662): that there is in immediate imperative. fact a distinct third person marker (here realized as /a/ is found in the 2 form of the immediate impera- zero) which is pluralized by /k/. (Unfortunately, it tive; it also underlies the diminutive paradigm (5.82). throws no further light on the other part of our /amw/ occurs in the 2p form of the immediate analysis, since /tan/ is the only person marker of imperative and in all forms of the delayed imperative. this paradigm to occur with a non-zero direction The etymological relation of /amw/ and /am/, if marker; thus the question if /i/ and /ehkw/ are just any, is not clear. has the widest distribution of the TI theme the sources show a of /am/ 73 Note that missionary variety endings it occurs in the forms of the for 2p-3p and especially 21-3p: markers; third-person independent order and throughout the conjunct 21-3p 2p-3p order. is further used as the basis for a host Hunter -dhkahkwawik -ihkekok (/am/ -ahkahkwdnik of derivatives; see, for instance, 6.413 and 6.446; cf. Lacombe -dhkwaniiwik -dhkekwdwik also 5.812). -ahkwdwik -ahkwdwinik 5.72. AI and TI Non-Third and Imperative Endings -ahkwdwiniwik Lessard -dhkwiwik -dhkckwdwik The non-third endings of the independent and Edwards -dhkwdnik -dhkekwdw conjunct orders and the endings of the imperative VOL. 63, PT. 5, 1973] VERB INFLECTION 59 order are identical for the AI and TI paradigms: /k/ is used instead of /t/ and the nasal is replaced by /h/ (cf. 5.74 and appendix A: 1.3). Thus, AI Independent Conjunct /hk/, /hkik/ /hkwaw/ and TI /ahk/, /ahkik/ indf /hk/ -/ahkwaw/. In the obviative form, where the 1 /n/ /an/ third person marker is preceded by /eyi/, /t/ is found 2 /n/ /an/ throughout. lp /enan/ /ahk/ 21 /enaw/, /enanaw/ /ahkw/ 5.74. II Endings 2p /enawaw/ /ekw/ In the independent order, the II endings correspond Imperative exactly to those of inanimate nouns (3.31). The Immediate Delayed third person endings cannot be segmented. The is set as that the is 2 /h/ /an/ singular ending up /wi/; (note /i/ on the basis of the nominal 21 /tan/ /ahkw/ posited corresponding and of The 2p /k/ /ekw/ ending (3.31) comparative evidence). plural ending is /wah/. The thematic sign /eyi/ The 21 ending in the independent order has two marks the obviative. alternants, /enaw/ and /enanaw/. These seem to be In the conjunct order, the third person is marked in free variation, unless some subtle differences of by /k/ before which the /n/ of n-stems is replaced style have gone unnoticed. The evidence for a by /h/ (cf. 5.73); /eyi/ marks the obviative. In the correlation of the choice of alternant with age group simple and changed modes of the conjunct order, the or geographical criteria is inconclusive. Moreover, plural marker is /ih/, e.g. T45p6 e-ndkwahki 'when both forms are used by some speakers, as in T55p49 they are seen.' In the subjunctive and iterative where kika-sdkewdnaw and kika-sakewndnnaw 'we'll modes, however, we find /waw/ instead, which is then come into open view' occur in successive sentences. followed by the subjunctive and iterative marker (A different situation involving an epenthetic element /ih/; e.g. ohpikihkwawi 'if they (Op) are growing.' /na/ is described in 5.852.) There is some syncretism between the n-stem It may be interesting to review briefly the situation forms and the forms used with vocalic stems. Instead as described elsewhere, especially since Goddard of the normal -hk, some n-stems end in -k. (This regards the insertion of /na/ as pivotal in the develop- is not a case of weak pre-aspiration but of indis- ment of the AI paradigm (1967: p. 76). Edwards, putable contrast.) Thus, besides miywdsin 'it is Lessard, and Ellis report only the longer forms, with good,' T35pl e-miywasiniyik 'it (0') is good,' etc. we /na/, while Hunter and Lacombe give both forms. find T16p79, T113p2 e-miywdsik 'it is good'; besides Lacombe even cites TA forms with the /na/ option- yotin 'it is windy,' T10p48 e-nipahiyotik 'it is really ally inserted; (but I can find no evidence of 2p forms windy,' etc. without /na/ in Lacombe, as Goddard has it in 1967: fn 36). 5.8. MARGINAL AND SUPPLETIVE PARADIGMS In the AI paradigm, the non-third-person forms of A MARGINAL the conjunct order regularly show the /y/-alternant paradigm is characterized as one of the suffixes (5.453). which formally diverges, however slightly, from one There is no indefinite actor form in the independent of the basic paradigms, and which is used with certain order74; for a discussion of derived forms which secondarily derived stems only. function in this capacity see 5.85. SUPPLETIVE paradigms similarly occur with stems which are secondarily derived by certain specific 5.73. AI and TI Third-Person Endings suffixes. Although their forms do not diverge from those of the basic paradigms, the suppletive paradigms The third-person forms of the independent order generally serve to fill gaps, of whatever origin, in the show the usual third person endings /wa/, /wa-k/, basic paradigms; they are often themselves incomplete. and /eyi-wa-h/. The boundaries delimiting marginal and suppletive In the conjunct order, there are two alternative paradigms from each other and from certain (deriva- sets of endings. AI stems ending in a vowel take the tionally) late derivatives are not easy to draw. The third person marker /t/ which is pluralized by RELATIONAL (5.81) and DIMINUTIVE(5.82) paradigms /k/-/waw/ and which may be preceded by the are marginal; they show formal differences from the thematic obviative marker /eyi/. Where the third basic paradigms they are modeled on. The TA IN- person ending immediately follows a nasal (that is, in ANIMATE ACTOR paradigm (5.83) differs from the the non-obviative forms), the third person marker basic AI type only in one minute point; and the difference is leveled out in some dialect areas. 74 For the Odawa dialect of Ojibwa, Piggott (1971b: pp. 92, 101) already The TA ACTOR reports an independent ending m which might point towards a INDEFINITE paradigm (5.84) shows the segmentation of the Cree conjunct ending -hk. suppletive type most clearly: it has some AI forms 60 WOLFART: PLAINS CREE [TRANS. AMER. PHIL. SOC. with secondarily derived stems which complement I guess, a lean-to was built for her'; S10-22 tdpwe some perfectly regular TA forms (5.2). Finally, the matotistnihkwdn, . . . 'Accordingly a sweat-lodge AI and TI INDEFINITE (5.85) and INANIMATE (5.86) was built for him, . . .' ACTOR forms are morphologically indistinguishable 5.812. Relational forms are formed from AI and TI from the II basic type; syntactically and semantically, stems with the suffix /w/. With AI verbs, this however, they fill obvious gaps within their respective marker is added to the stem; no instances of a rela- basic paradigms. Of course, the gradual scale does tional form derived from an n-stem have been re- not abruptly end here: there is a vast number of corded.75 With TI stems, the choice of theme is formations which pose the familiar problem of the ambiguous: either /amw/ or /am/ may be posited distinction of inflection and derivation (cf. also since /w/ followed by /w/ yields /w/. 5.664). The existence of an identifiable paradigmatic Relational forms are not part of inflection because gap may well provide an operational criterion, how- they show in linear order two theme signs, /am(w)/ ever tentative, for distinguishing suppletive paradigms and /a/'/e/, which are mutually exclusive in from derivation. inflection. Their derivation is very close to the surface (i.e., late) since they do not underly further 5.81. The Relational Paradigm derivation; cf. 5.814. 5.813. The forms which are attested in texts are RELATIONAL forms indicate that the action of the verb relates to a other than the actor in a listed below; in spite of their small number, they person way characterize the structure of the which is not specified; there is no concord, nor is the clearly paradigm. A full is Ellis 14-10; "related" person specified by the verb form as to paradigm given by (1962: p. p. 23-13; C-3; 1971: gender, person, or number. E.g., T54p6 . . . app. p. 94). e-pimitisahiket tdntahto-tipiskdw e-pimohtewiht naway Independent Conjunct .. . . .he followed and as e-nipdt behind, many Indicative h-preterit nights the party (indf) traveled (with relation to him), indf he slept behind . . .'; the example indicates that the /eht/76 1 "related" person is not necessarily a beneficiary of /ak/ the action. 2 /at/ Bloomfield seems to have coined the term for this 3 /ew/ /at/ formation which he states to be to Cree peculiar The conjunct forms are obviously identical with to the relational (1928: p. 429). By failing distinguish the endings of the TA paradigm. The independent from the Hunter and Lacombe further obviative, forms may be interpreted in at least two ways. confused an Edwards follows already complex matter; Ellis (1962: p. 14-9) regards the vowel following the them in at least in but see this, part (1954: p. 47-2; /w/ (/a/ in non-third, /e/ in third person forms) as This confusion is not if we p. XII-10). surprising the stem vowel of the AI e-stem. This interpretation consider phrases like the following: Sll-32 tdnehki would not account for the indefinite actor form which mina kitihkwatim k-o-mestihkasamwat otayowinisah? is absent in the basic AI paradigm.77 More im- now did burn clothes 'Why you your nephew's (in portant, however, would be the resulting lack of the relation to ?' him) parallelism between the independent and conjunct 5.811. A of this simpler example fairly frequent orders; it seems more likely that the set of personal is oskisikwa. 'I see his phrase type niwdpahtamwdn endings has been modified in one point than that the eyes (in relation to him).' However, this is only orders should have been modeled on different basic of the of relational forms. Consider the part range paradigms. sentences: ana following T10p94 eha, niwdpahtamwan Thus, we assume instead that the independent order 'Yes, I saw it (sc. her leg) in relation to this one (ana, namely the person being talked about).' T54p9 ekwa 76 Ellis reports that n-stems show i before the w (1971: p. 87); kiskisamwdn oma acosis ekos Tsi as he notes himself, his interpretationof this i as connective /i/ napate e-ohcipitamoht. violates the usual restrictionson the latter. It seems much more 'Then someone cut this arrow off on one side (in likely that n-stems exhibit their ni-alternant in this environ- relation to the patient), thus pulling it out (in rela- ment, exemplifying further their development towards an even- tion to him).' T103p4 .... ekd kd-nitaweyihtahk tual vowelstem; cf. 5.52. 76 That the -wiht t-apiwiht Jma ohtapiwinihk. '. . . that he doesn't like conjunct indefinite ending fails to show the contraction A: after AI vowel stems is it for someone to sit on this seat relation to him, expected (cf. appendix 4) (with most plausiblyattributed to the structureof this paradigm:under i.e. on his seat).' These fairly complex examples contraction, the characteristicw would disappear. It might also should not obscure the fact that many instances of mean, however, that contraction does not take place if there are relational forms occur without any adjuncts; this is two morpheme-boundaries-a restriction that requires further so with a form which seems to be the study. particularly 77But note that this is not an most relational the oversight on Ellis's part since he frequent single form, namely gives a different indefinite actor form. His view would seem to indefinite actor form (most typically of AI verbs); be less defensible with regard to the postulation of a freely mov- e.g., T71-10 ekwa etokwe e-napatestawikewiht. 'Then, able stem vowel just to account for this partial paradigm. VOL. 63, PT. 5, 1973] VERB INFLECTION 61

paradigm, too, is modeled on the TA paradigm. Since is almost entirely lacking, we will briefly summarize there is only one series of forms involving non-third the data given by Lacombe (1874b: pp. 124-126). persons (and none with plural and obviation markers The one instance found in our texts corresponds to added), the presence or absence of the third-person Lacombe's form: T10p107 ewako mana kd-miyweyi- marker /wa/ /a/ cannot be determined. Thus, masit ohi. 'That's the one (3) who likes this one these forms are phonemically identical to the corre- (3')'; the 3 referent is a baby. sponding AI set; just like AI forms, they make As presented by Lacombe, the diminutive paradigm anaphoric reference to an actor only; and finally, shows one very curious feature. If we leave the the distribution of the TA theme signs /e/ and /a,/ in you-and-me forms aside, the direct forms have the the independent order corresponds exactly to that of diminutive morpheme /esi/ inserted after the theme the AI stem vowels /e/ and /a/ (5.52). It is not signs, and the inverse forms, before. Thus, /paka- surprising, then, that the singular non-third forms mahw-e-esi-wa/ pakamahwesiw 'he (3) hits him (3')' have been remodeled after the AI paradigm. but /pakamahw-esi-ekw-[i]-t/ pakamahosikot 'he (3') 5.814. Superficially similar but structurally quite hits him (3)' (conjunct form).' different from the relational forms is a class of TA In the you-and-me paradigm, the order of theme verbs whose stem ends in -amaw-. These verbs belong sign and diminutive marker is reversed. In the TA to the double-goal type of verbs, e.g. nitotamawew inverse subset, /esi/ follows the theme sign rather 'he asks him for him or it' (5.11); although such verbs than preceding it as in the inverse forms described are typically derived from TI stems, e.g. nitotam 'he above; /ki-pakamahw-et-esi-en/ kipakamahotisin 'I is asks for it,' the second goal not determined in any hityou.' (In the direct set, a form like pakamahosiyan way as to gender, number, or person. Thus, P34-16 'you hit me' (conjunct) might be interpreted either . . .-wih-kakweh-kimotamawdcik otemiyiwah, . . . way, as /-i-esi-yan/ or /-esi-i-yan/.) '. . to try to steal (3p-(3')) their horses from them . (3') . .'; the second goal here is otemiyiwa 'their 5.83. The TA Inanimate Actor Paradigm horses,' an animate noun with obviative possessor. The exact difference in meaning between relational The transitive animate (TA) inanimate actor forms and TA verbs in -amaw- has not been estab- paradigm is based on the theme sign /ekw/-/eko/; lished; cf. also 6.446. They are obviously different for the variant form /eko/ see 5.661. The following in syntax, but individual examples may be difficult forms are attested: to assess. Morphology provides the simplest cri- terion: TA verbs in -amaw- are subject to further Independent Conjunct derivation, e.g., nipahtamdsdw 'he kills him or it for -1 ni- -ikon -ikoydn himself,' from nipahtamaw- 'kill him or it for him.' -2 ki- -ikon -ikoyan Relational forms, by contrast, do not undergo further -lp -ikoydhk derivation; that is, they close the construction. -21 ki- -ikonanaw -ikoyahk -2p -ikoyek 5.82. The Diminutive Paradigm -3 -ikow, -ik -ikot The individual TA DIMINUTIVEforms differ from the -3p -ikwak basic TA the diminutive marker paradigm only by The theme sign is generally followed by the usual the indefinite actor form of the (except possibly AI endings. The third-person forms of the in- The of the total conjunct order). divergence para- dependent order show some deviation from the rest is more and makes it digms, however, complex appro- of the paradigm. In the analysis of the TA inverse to consider the diminutive a priate marginal paradigm. paradigm (5.613, 5.622, 5.65) it was seen that a also A: 2.3. Cf. appendix sequence /ekw-a/ results in -ik, as in the independent 5.821. From AI and TI diminutive be stems, may order forms for 3-1, 2 or for (3')-3. The emergence of formed the suffix It is to the by /esi/. joined directly the longer form -ikow,78 however, is attributed pri- stem of AI verbs instance of an n-stem is re- (no marily to the leveling influence of the paradigm (and TI verbs exhibit the theme corded); sign /a/ (cf. at best secondarily to homonym-avoidance). All 5.71). The usual AI endings are used with the forms but the two under consideration resemble AI stem. . . . oma resulting Examples: T34p6 szpzsis forms by having -iko- (partly from /ekw/ plus /e/ or . .. 'I crossed this little nicasowahasin stream'; cf. /i/, partly the alternant /eko/; cf. 5.661), and it is dsowaham 'he crosses it.' dskaw niki- T72p22 easy to see how the pattern pressure might work. In -acoskesindn, . . . 'sometimes we used to work a fact, this interpretation is supported by the occurrence, .. atoskew 'he works.' T72-31 little, .'; cf. ekospi side by side, of -ikow on the one hand, and -ik and . 'Then I used dsay e-nihtd-cehcapisiydn, already -ikwak on the other, e.g., kikway mdkwahikow. 'Some- to be a .. 'he rides on good rider, .'; cf. thtapiw thing bothers him'; T125-3 nipahikwak kikway. horseback.' 5.822. Since documentation for the TA diminutive 78 For the lengtheningof o before w cf. appendix A: 3.5. 62 WOLFART: PLAINS CREE [TRANS. AMER. PHIL. SOC.

'Something has killed them'; ninohtehkatewinnipahik. 5.851. Another form which also seems to belong 'My hunger killed him.' here ends in -ni-, e.g., P4-28 mestinikdniwa 'there is no Lacombe (1874b: p. 119) cites -ikow and -ikowak wasting (Op)' (from mestinikew AI 'he uses things up'; for this paradigm, versus -ik and -ikwak in the TA cf. the conjunct form in P4-28 ey-isih-mestinikehk inverse set. This is also the situation described by 'one uses things up'). -ni- occurs particularly fre- Edwards (1954: p. 31-3) except that she reports quently with the AI stem itwe- 'say so,' e.g., T79-2 the plural form with a different plural marker: itwdniw 'it is said.' All glosses are tentative since -ikowdw. Most interesting in this connection is the full meaning of this form remains to be ascertained. Ellis's paradigm where the rebuilt forms -ikow, -ikowak 5.852. Unlike the situation described in 5.72, the are found not only in the inanimate actor set but also distribution of /na/ in indefinite actor forms is in the TA inverse paradigm! (See Ellis, 1962: statable in terms of stem shape. AI stems ending in appendix A, C-l; 1971: table 6; note that -ikow /e/ or /a/ take the alternant without /na/, while all occurs only for 3'-3, not for the 3-1, 2 forms.) others select the alternant with /na/. In the environ- ment of the suffixes -niwiw, -niwan, -niwin, e-stems 5.84. The TA Indefinite Actor Paradigm show the /a/-alternant; cf. 5.52. In effect, then, all these forms contain an /a/, of whatever structural The transitive animate (TA) indefinite actor identity, before the -niwiw, -niwan, -niwin suffixes, paradigm well exemplifies the suppletive type: it only which makes for considerable superficial similarity: consists of forms involving a non-third goal while the T122-1 nitopayindniwiw 'there was going on the war- indf-3 forms belong to the basic TA paradigm. path (-ndniwiw)' and tipiskihtdniwiw 'nightfall was The suffix of the TA indefinite actor paradigm is encountered (-niwiw).' /ekawi/. It is followed by the usual AI endings of the non-third persons. 5.86. The AI Inanimate Actor Paradigm The meaning and use of this is paradigm exactly The animate intransitive inanimate actor parallel to that of the other indefinite actor forms.79 (Al) kikwdsihikawindanw 'we are forms are based on the suffix /Lmakan/ and then Examples: T20p73 being inflected as II T17-3 kidnapped'; T55p75 nipThtokwehikawindn'they took verbs, e.g., pe-nipimakanwa die nipiw 'he us inside'; T72p5 ". . ." e-itikawiydn '". . ." I am 'they (Op) out," cf. dies'; T46p10 e-tocikemakahk 'it does called'; T55p68 asamikawiydhki 'if we are fed.' (sc. drinking) things (to people),' cf. tocikew 'he does things.' forms based on TI stems 5.85. The AI and TI Indefinite Actor Paradigm Although parallel (with the theme /amw/) have been reported, no examples Both animate intransitive and transitive inanimate were found in our texts. Instead, the forms of the verbs have regular indefinite actor forms in the basic TI paradigm occur, e.g., T10p128 ekwa wiya onma conjunct order. In the independent order, however, micihciy e-pe-otinahk 'and then this hand took it.' this function is taken over by secondarily derived II The AI inanimate actor forms are included here verbs. But in spite of their morphological inanimate- because they parallel the TA inanimate actor paradigm ness, these forms are exactly parallel in syntax and which is truly marginal. But the point is reached meaning to the indefinite actor forms; their meaning where inclusion among the marginal or suppletive is perhaps best indicated by such glosses as 'one does paradigms rests on slender criteria indeed. x' or 'there is x-ing being done (by animates)'; cf. also 4.422. Thus, S27-14 kipah mana picinaniwiw, 6. WORD FORMATION kiksepd e-wih-picihk. 'We (i.e., "one") always move The basic distinction Cree words is between camp (indf, early, when we move among independent) camp those which select inflectional and those (indf, conjunct) in the morning.' paradigms There is considerable variation in the formation of which do not, namely particles. A second dichotomy divides the former set into verbs on the one hand and these stems. Thus, we find the suffixes -(na)niwan-, nouns and on the other. -(na)niwin-, and, most frequently, -(nd)niwi-. (The pronouns resemblance between these suffixes and some of the In the more general aspects of derivation the various word classes are much alike. The brief most II finals 6.431, very productive (cf. 6.433) might sketch of Cree word formation which is here indicate their analysis into an element -niw- plus one presented of the II finals.)80 follows closely the model of Bloomfield's Menomini grammar.81 79 Note the historical implicationsof Ellis's statement that the 6.1. DERIVATION "prefixlessforms [e.g., wapamdwindf-3] are rejected at Albany and Moose as being 'incomplete';instead the AI verbs of under- A stem which has only bound constituents is 'he is seen'] are used" going [e.g., wapamakaniwiw (1971: p. 85). PRIMARY. Where one of the constituents is a free 80In the James Bay dialect, this secondary formation has spread to the conjunct orderwhere it has crowdedout the regular 81In addition to the data describedin 1.42, this sketch also in- ending /hk/. Thus, conjunct -(nd)niwahk besides independent cludes words cited directly from Bloomfield'smanuscript lexicon -(nd)niwan; cf. Ellis, 1962: appendix C; 1971: table 2. (MS(b)), which is based on his texts. VOL. 63, PT. 5, 1973] WORD FORMATION 63 form (i.e., a stem), we are dealing with SECONDARY occur where neither meaning nor an intervening final derivation. For composition see 6.5. show two medials to be hierarchically distinct, e.g. (T505p17) kiskacaydskwahwew 'he (tree) cuts his 6.11. Primary Formation belly by tool'; the root kZsk- 'cut' is followed by the The immediate constituents of primary stems are medials -acay- 'belly' (diminutive, cf. -atay-) and -dskw- 'wood' and the final INITIALS (root or extended root; cf. 6.111), MEDIALS -ahw- 'by tool.' Such a form is considered comical. which are optional, and FINALS; e.g. root paw-, final highly -ahw-: pawahw- TA 'brush him by tool'; with medial 6.122. Some suffixes are specifically secondary, e.g., -apisk-: pawapiskahw- TA 'brush him (pipe, stove) the reflexive -iso- in wdpamiso- AI 'see oneself,' from with or as metal or stone.' the stem wdpam- TA 'see him.' Others occur in 6.111. All three immediate constituents may exhibit primary as well as secondary word formation, e.g., a (lexomorphemic) alternation of shorter and longer the transitive animate abstract final -h-: primary forms. We follow Bloomfield in regarding the shorter kisihew 'he completes him,' secondary tapasihew 'he alternants as basic and calling the longer alternants flees from him'; cf. tapasiw 'he flees.' EXTENDED; the morphemic status of these post- Since most suffixes occur in secondary derivation radical, pre- and post-medial, and prefinal ACCRETIVE as well as in primary stem-formation, it is not at all elements is purposely left open. In fact, the entire surprising "that no clear line can be drawn between problem of lexomorphemic alternation goes beyond these two types of construction; many stems could the scope of the present sketch. It is mentioned be described in either way" (Bloomfield, 1962: p. 66). here primarily for practical reasons (cf. the brief discussion of extended medials and finals in 6.33 and 6.13. Derived (Deverbal) Suffixes 6.4, respectively) and to contrast it with the hierarchi- cal structure of derived (deverbal) suffixes (6.13). The mechanism of deriving stems from stems is 6.112. The root or the final suffix may occasionally complemented by another, and more characteristically be set up as zero. Thus, for instance, we posit a zero Algonquian one, namely the formation of derived root in a verb like oskastimwew 'he has a young horse'; medials and finals. the remaining elements are the derived medial Medials and finals are of two types. SIMPLE -oskastimw- and the animate intransitive final -e-. medials and finals show no internal structure. They The medial in turn consists of the root osk- 'young' consist of only one morpheme (but may sometimes and the final -astimw- 'dog, horse.' have "extended" alternants; cf. 6.111). They are A zero final is conveniently posited with an un- not paralleled by independent stems. Examples: analyzable noun root such as sissp- 'duck.' medial -dpisk- 'stone, metal'; transitive animate and No root is present in dependent noun stems (cf. inanimate finals -ahw-, -ah-, 'by tool.' 3.2), and in many the final is zero as well; thus, the DERIVED (deverbal) medials and finals are paralleled medial -sit- 'foot,' as in kisipekisitew 'he washes his by independent stems from which they are said to be (own) feet,' occurs with zero root and zero final in the derived. The derivative status may not always be dependent stem nisit 'my foot.' obvious, as in okimdskwew 'queen' where the final -iskwew is derived from the stem iskwew- 'woman' 6.12. Secondary Derivation (which in turn consists of the root and a zero suffix). In secondary derivation further suffixes are added It is indisputable where the derived medial differs in to stems to form derived stems. Thus, from a root phonemic shape from the parallel independent stem, nito- 'seek' (cf. nitonam 'he seeks it') and an animate e.g., atimw-: -astimw- 'dog, horse'; or kTsasinah- TI intransitive final -payi- 'move' we get a primary stem 'finish writing it' where the final -asinah- is derived nitopayi- AI 'go on the war-path'; from this a second- from the stem masinah- TI 'mark it by tool.' No ary noun stem may be derived with the abstract matter what the internal make-up or derivational suffix -win; finally, another intransitive suffix -ihke- history of the derived medial or final may be, in its completes the verb nitopayiwinihkew 'he leads a medial or final function it is treated as a unit. Bloom- war-party.' field called this type of formation DEVERBAL; note that 6.121. Secondary derivation shows the same ternary this term does not imply the form-class of the base structure as primary formation: a stem (rather than a as do the term "deverbative" and "denominative." root) is followed optionally by a medial and obliga- One may visualize the formation of derived (de- torily by a final.82 For examples see 6.133. verbal) suffixes as "vertical" derivation in contrast Two medials or finals occurring in succession belong to the "horizontal" or "left-to-right" pattern of to different layers of derivation. A nonce-form may deriving stems from stems. Together these two of derivation account for much 82 In interlocking types some of his earlier writings Bloomfield appears to have of the and of Cree word regarded all secondary suffixes as finals; according to this view great productivity complexity the ternary structure would be restricted to the domain of pri- formation. mary stem formation. For a more detailed discussion see 6.333. 6.131. There is a clear distinction between INITIAL 64 WOLFART: PLAINS CREE ITRANS. AMER. PHIL. SOC.

and NON-INITIALalternants of single morphemes as tected. Thus, besides the stem minis- 'berry' (the well as of stems. -is- is a diminutive suffix) we find the suffix -min- Thus, for example, the root atimw- 'dog, horse' is which is very frequent indeed; e.g., misdskwatomin paralleled by a non-initial alternant -astimw- which 'saskatoon berry,' takwahiminew 'he crushes berries.' occurs in wdpastim 'white horse or dog,' manastimwew Its meaning is wider than that of the initial stem, e.g. 'he goes on a horse-raid,' or even atimwastim 'dog mahtdmin 'grain of maize,' oskahtdmin '(young) of a dog.' kernel or stone (of fruit),' wapimin 'white bead.' A stem may similarly have a non-initial alternant. Another typical example is -sip-, initial stem sisip- Thus, besides the initial alternant pdhpih- TA 'laugh 'duck'; (cf. the Proto Algonquian forms *-eqsip- and at him' (which consists of the root pahpi- and the *?sqsvp). Examples: kaskitesip 'black duck,' iyinisip transitive animate final -h-) we find the non-initial 'mallard,' oskacanisip 'mudhen,' etc. alternant -dhpih- which is used as a derived (deverbal) (c) Several of the non-initial elements which under- final in itdhpihew 'he thus laughs at him.' (For lie dependent nouns (3.2) have an alternant form further examples see 6.133.) when occurring in other combinations. Most typical With many forms, however, the initial and non- are the pairs of alternants with or without suffix- initial alternants have the same phonemic shape even initial s, e.g., niskat 'my leg': -kdt- in peyakokdtew though they are distinct theoretically; (cf. Bloomfield, 'he is one-legged' or wepikdtew 'he flings his legs'; 1930: p. 72 fn.). Consider the morpheme wep- nispiton 'my arm': -piton- in sakipitonenew 'he seizes -.-wep- in wepinew 'he throws him away by hand' him by the arm'; niskan 'my bone': -kan- in pahko- and in kwdskwewepahwew 'he knocks him aloft by kanehwew 'he crushes his bone by tool.' tool'; cf. kwdskwepayiw 'he leaps.' Consider also (d) The non-initial alternant often lacks part of the stem akim- TA 'count him' (which consists of the the initial from which it may be said to be derived. root ak- and the transitive animate final -m-); the Thus, from maskwamiy 'ice' we get the medial non-initial alternant has the same phonemic shape; -askwam- in manaskwamew 'he gets ice for use.' This as in itakimew 'he counts him so.' type of alternation where the non-initial alternant Consequently, the sequence of derivation is often shows the loss of stem-initial w, m, or n, appears to indeterminate. Consider the verb itacimostawew 'he occur most frequently, e.g., -dpew-, from ndpew 'man': narrates so to him'; it is either secondarily derived mosapew 'spouseless man, bachelor, widower,' mistdpew from itdcimo- AI 'narrate so' (where -dcimo- is a 'giant,' etc. Consider also -dposw- in ndtdposwew 'he derived (deverbal) final), or it is a primary verb hunts rabbits,' cf. the stem wdposw- 'rabbit'; -ahtdmin- consisting of the root it- and the derived final -dcimo- in oskahtamin '(young) kernel or stone (of fruit),' cf. staw-. Since both underlying forms, itdcimo- AI mahtamin 'grain of maize'; -askisin- in pahkekin- 'narrate so' and dcimostaw- TA 'narrate to him,' waskisin 'leather moccasin,' ketaskisinew 'he takes actually occur, no decision seems possible. his (own) shoes off,' miywaskisinew 'he has good 6.132. Only the more obvious types of derived shoes,' sdpopeskisinew 'he has wet moccasins,' etc; cf. medials and finals can be indicated here. maskisin 'shoe.' Note also the alternant -eskisin in (a) Non-initial forms are rarely longer than the nisweskisinew 'he has double moccasins,' and mihcetos- initial forms with which they alternate; (the pre- and kisin (recorded at The Pas, Manitoba) 'many post-suffixal extensions mentioned in 6.111 are not moccasins.' included in the present discussion). A very clear 6.133. A few more complex examples of derived example is the initial atimw- and the medial -astimw- suffixes follow. From the simple stem masinah-: 'dog, horse.' masinaham 'he marks it by tool' the final -asinah- Sometimes there are variations in the length of may be derived, as in kisasinaham 'he finishes writing vowels; consider the initial awds-, as in awdsis 'child,' it.' From masinah- we can also form a secondary nicawdsimisak 'my children,' etc.; and the non-initial animate intransitive stem masinahike- 'write things'; -dwas(o)- which seems to occur primarily as a second- the secondary suffix -ke- indicates a general goal. ary suffix, e.g., wicewdwasow 'he has his children From this stem, a final -asinahike may be derived, as along,' cf. wicewew 'he has him along'; wdpamdwasow in petasinahikew 'he writes hither.' 'she sees her child, gives birth to it,' cf. wdpamew 'he The primary stem postayjwinise- 'dress, put on sees him'; etc. Another example is the initial kon-, clothes' consists of a root post-, a medial -ayowinis- as in kona 'snow,' koniwiw 'it is snowy,' and the 'clothes,' and an animate intransitive final -e-. The medial -dkon- which occurs in ndtahdkonew'he fetches medial -ayowinis- is paralleled by a diminutive (-is-) snow,' sipdydkonakihew 'he makes him go under the noun which is derived by the suffix -win from a verb snow,' moskdkonepitew 'he pulls him out of snow'; stem which is not attested in Cree; but cf. Fox concerning the -d- cf. 6.332. ayowini 'thing used' and ayowa 'he uses it' (Bloom- (b) Many non-initial morphemes which are clearly field, 1927: p. 401). Consider also postayowinisahew related to initial stems have to be listed individually 'he dresses him,' miskotayowinisew 'he changes since no general pattern of alternation can be de- clothes,' etc. VOL. 63, PT. 5, 1973] WORD FORMATION 65

Further complex medials are -atdawkan- in notata- Many roots appear in shorter and longer, extended, wdkanew 'he hunts for furs' and -dcimowin- in mandci- forms; cf. 6.111. Thus, the root wdp- whose focal mowinew 'he gets a story.' atawdkan 'fur' is derived meaning is 'light, see' occurs in the meaning 'white' in from atdwdkew 'he sells' and literally means 'what is wapastim 'white horse or dog,' wdpimin 'white bead,' sold'; atdwdkew in turn is derived from atdwew 'he wapahkesiw 'white fox,' etc. The extended root sells, trades.' dcimowin 'story, text' shows the suffix wapisk-, with no apparent difference in meaning, -win used to derive nouns from verbs; the underlying occurs in wdpiskdpakonis 'white flower,' wdpiskiwiyds verb dcimo- 'he tells (of himself)' is a middle reflexive 'white man,' etc. derived from the transitive animate stem acim- 'tell of him,' which in turn consist of a root dt- and a 6.21. Root Classes transitive animate final -m- 'by speech.' A classification of Cree roots has to take account Derived suffixes are of course not restricted to of the freedom with which roots occur in in stems but of great may occurring primary only examples different stem note that this statement suffixes are much rarer than either classes; applies complex secondary to stem formation alone and does not refer or suffixes. As primary complex primary simple secondary to the classes to which derived stems an consider 'he washes secondarily may example kistpekinayjwinisew Since each stem class has at least a few his clothes the first of derivation belong. (own) by hand'; layer members whose roots do not seem to recur in another consists of the underlying stem kisipekin-, the medial stem class also, we may tentatively set up classes of -ayowinis- 'clothes' (cf. above), and the animate noun, verb, particle, and pronoun (?) roots. Such a intransitive final -e-. The underlying stem in turn classification would imply that a very great proportion consists of a root kisi- 'agitate,' an extended medial of roots from each class is subject to class-cleavage. -ipek- 'liquid,' and a transitive final -in- 'by hand,' 'he washes it.' In Bloomfield's view, there are only two distinct cf. kisipekinam root classes: nominal roots occur in stems consider another where "vertical" specifically Finally, example with zero roots are of derivation combines with "horizontal" derivation. suffix; general capable occurring in nouns "concrete" From a root kanaw- can be further (with suffixes), verbs, particles, (which probably and This classification does the medial mental pronouns. obviously analyzed), -eyi- indicating activity, not that each root in the whole and the transitive animate final -m- we form a stem imply general appears of it is TA him.' this as range possible functions; conversely, however, kanaweyim- 'tend, keep Using without detailed in each case we can then add the medial -iskwew- impossible investigation underlying stem, to state a certain root to be restricted to one stem 'woman' and the animate intransitive final -e- to get class. a verb kanaweyimiskwewew 'he watches his wife.' In a further round of derivation an abstract noun The two classifications outlined above may be viewed as on different levels of can be formed, as in operating generaliza- kanaweyimiskwewewin T46p8 tion and therefore do not exclude each kz-poyow anima okanaweyimiskwewewin. 'He quit necessarily this of his.' Or another verbal other. For the practical purposes of the present wife-watching yet sketch we shall the more suffix, -iski-, be added, to form the verb kana- adopt specific classification, may in mind that it well be weyimiskweweskiw 'he watches his wife or keeping may overly specific, constantly' and theoretical for a fuller 'he is a constant (habitual) wife-watcher.' reserving argument investigation. 6.2. ROOTS Thus there is an enormous amount of class-cleavage. For example wdp- occurs as a verb root in wapiw 'he Roots occur as the initial constituents of stems. sees' or wdpamew 'he sees him'; as a noun root in For some formations a zero root is posited; in oskisk- wdpastim 'white horse or dog' or wdpimin 'white wewew, for instance, the medial -oskiskwew- means bead'; and as a particle root in wapi and wdpiski 'new woman,' and the meaning of the zero root (or 'white.' mihkw- 'red' occurs in both nouns and of the construction) is 'he has _'; thus, the meaning verbs, e.g. mihko 'blood,' mihkwekin 'red cloth'; mihkosiw 'he is 'it is red of the stem is 'he has a recent wife,' 'he is newly red,' mihkwdpiskdw (as note that this verb is not based on a noun married.' No clear meaning is evident for the zero metal)'; stem (6.32). kis- 'complete' occurs in both verbs root of dependent nouns; cf. 6.112 and 3.2. and particles, e.g. kisihew 'he completes him,' kTsisiw Since roots often occur in one or a few only just 'he is mature,' kWseyihtam'he completes his plan of their and semantic stems, morphological analysis it'; kisi, ki 'completely.' kikw- 'what' occurs in may remain incomplete. In some cases, as Bloom- both nouns and pronouns; consider the stem kik- field put it (1962: p. 425), "it would be an idle exercise wdhtikw- 'what tree' from which a secondary verb of ingenuity to seek a formula that would cover the kikwdhtikowiw 'what kind of tree is he' is derived; meanings of stems that begin with the same sequence kikway 'what' (4.3). eyakw- 'this selfsame' occurs of sounds." in both pronouns and particles, e.g., eyako 'this 66 WOLFART: PLAINS CREE [TRANS. AMER. PHIL. SOC. selfsame,' eyakwdc 'just then,' ekota 'there,' (ya)kosi and one without word boundary, remains to be fully 'thus,' etc. investigated; the writing of hyphen is not always The following roots seem to occur in only one type consistent. of stem each. Noun: atimw- 'horse or dog'; verb: post- 'put on (clothing)' in postastotinew 'he puts on 6.23. Relative Roots his headgear'; particle: mastaw 'later'; pronoun: RELATIVE roots require an antecedent; the ante- awiyak 'someone.' cedent may be a clause, a particle expression, directly quoted speech, etc. Thus, from the root it- /eO-/ is 6.22. Reduplication formed the verb itew 'he says so to him,' e.g., S8-13 Verb and particle roots are freely reduplicated. omis 5tew wiwa. 'So he spoke to his wife'; note that Reduplication adds the meaning of continuity, repeti- the relative root it- is often balanced by another tion, intensity, etc. Some verbs customarily appear occurrence of the same root, here in omisi /6m-eO-i/ with reduplicated root, e.g., mdmitoneyihtam 'he 'this way.' Another example of two relative roots ponders it.' balancing each other is T55p51 ekosi esi-wapahtamdn. 'Thus I this 6.221. One type of reduplication involves a change (it was) what thus saw'; but balancing itwew of the root which undergoes reduplication; e.g., pim- is by no means obligatory, e.g., S8-18 "eha," aw ' this For details 'along': papdm- 'about,' as in pimihaw 'he flies along': iskwew. "Yes," said woman.' the of relative roots papdmihdw 'he flies about.' nitaw- 'go to do .. .' concerning syntactic properties 1946: 1958: 1962: nanataw- 'go intensively to do . . .,' as in nitawdpamew cf. Bloomfield, p. 120; pp. 36, 130; 443-447. 'he goes to see him': nandtawdpamew 'he looks out pp. constitute a small the most for him.' Consider also ohpikihew 'he raises him': Relative roots set; common relative roots in Cree are it- (S243-6) oyohpikihew 'he brings him up over a long 'thither, thus,' oht- and tahto- 'so time' (in this case neither the record nor the interpre- 'from there, therefore,' many'; e.g. 'he sits isi ohtinam 'he takes him tation are entirely beyond doubt). This type of itapiw thus,' 'thus'; thence or 'he is so reduplication is relatively rare and does not seem to therefore'; tahtopiponwew many winters old' tahtwdw 'so be productive: Bloomfield (1930: p. 6) considered it (secondary verb), many archaic. times'; etc. 6.222. The productive type of reduplication does 6.3. MEDIALS not affect the root itself. The root is instead preceded by the reduplication syllable. Medials in verbs appear between root and final suffix. In nouns and With roots beginning in a consonant, the redupli- particles, "they mostly appear at the end of the stem as we in fixed cation syllable usually consists of the first consonant or, may say, association with a final of the zero" (also of a cluster) plus d, e.g., kakipa 'over and over,' shape (Bloomfield, mdmescihtdszw 'he carried on his work of extermi- 1962: p. 69). Medial suffixes are characterized nation,' cacimatdw 'he plants it upright (everywhere),' primarily by etc. At least one instance has been encountered, their freedom of occurrence (especially in contrast to to be similar however, where both the initial consonant and the concrete noun finals which otherwise tend in function and vowel of the first syllable are repeated: mizmciw 'he meaning; cf. 6.412). Furthermore, medials are distinct from the various and eats it' (T91-9, in an emphatic denial). post-radical elements their which in Where a root begins in a vowel, the reduplication pre-final by meanings general are concrete. Wolfart, is marked ay- dy-?), e.g., ayohpikiw fairly (Cf. 1971a.) normally by (or Derived medials are illustrated 'he also wayohikiw 'he (deverbal) amply grows up' (cf. (T73p4) keeps in 6.133. medials fall into 'he thinks so'; aydtotam 'he 6.13, especially Simple growing'); ayiteyihtam two distributional classes which coincide with tells it over and over,' etc. A deviant type of re- largely semantic duplication has been recorded for the stem itwe-AI rough groupings. 'say so.' Besides the usual reduplicated form ayitwew 6.31. "Body-Part" Medials we find the form ititwew (T80pll, T91-13, Tl10p9); similarly with the transitive animate stem it- 'say A large group of simple medials occurs in dependent so to him,' we get a reduplicated form (T115p7) nouns (6.112, 3.2); they denote kinsmen, body-parts, ititew 'he says so to him all the time.' and a few intimate possessions. In a great number of cases, the vowel of the redupli- 6.311. The medials which denote kinsmen and cation syllable shows gradual devoicing which is personal possessions by and large seem to occur only symbolized by h in Bloomfield's texts, e.g., kdh-kinwes in primary dependent nouns (or in verbs secondarily 'for a very long time.' Elsewhere, such devoicing derived from these), e.g., nisim 'my younger sibling,' is interpreted as indicating word boundary, and is osimisiw 'he has a younger sibling'; nitem 'my dog then symbolized by hyphen (6.5). The question or horse,' otemiw 'he has a dog or horse,' otemihew 'he whether there are in fact two distinct types, one with makes him have a dog or horse,' etc. VOL. 63, PT. 5, 1973] WORD FORMATION 67

That these medials do not seem to occur in primary in nouns, e.g. piwapiskw- 'piece of metal' or ospwak- verb stems may well be due to semantic and contextual andpiskw- 'pipestone,' and in particles, e.g., paskwdpisk restrictions rather than to an inherent morphological 'bare mountain' or peyakwdpisk 'one dollar.' feature. They are certainly not restricted to occurring Semantically, many of these medials denote not a with a zero root, as is evident from the complex specific object but a class of objects or, indeed, the nouns which follow. Thus, the medial -kosis-, as in characteristic features of this class; see the examples nikosis 'my son,' occurs in dpihtdwikosisan 'halfbreed'; below. The superficial resemblance to a system of also -sokan-, as in nisokan 'my buttocks,' recurs in classificatory markers deserves detailed investigation. wdpiskisokan 'white hind-quarters' (kind of horse; Probably most frequent is -dskw- 'wood or solid of cf. Mandelbaum, 1940: p. 197). The medial -tas- in similar consistency.' It occurs very commonly in nitas 'my trousers' recurs in an extended form -etds- in verbs, e.g., akotdskohwew 'he hangs him on a tree mihkwekinwetds 'red-cloth breeches.' The medial of (by tool)'; mdkwdskohwew 'he (tree) oppresses him nmwas 'my pack' (plural nmwata) recurs as -iwas- in (as tool)'; mihcdskosiw 'he is a big tree'; mandskwew mTnisiwas 'berry bag,' pimihkdniwas 'pemmican bag,' 'he takes up (wooden) weapons,' etc. Nouns with etc. -dskw- are rare; it occurs as a secondary suffix in 6.312. Medials denoting body parts are more apwdnaskw- 'roasting spit,' for example. One might frequent than those discussed above, and more suspect that the apparent rarity of -dskw- in tree typically occur in verbs rather than in dependent names has to do with competition from the noun final stems alone; cf. also 6.132 (c). Where these medials -dhtikw- 'tree, stick.' (Note also that Siebert 1967a: are followed by a transitive final, post-medial -e- p. 27 glosses the Proto Algonquian equivalent of appears to be universal; cf. 6.333. -adhtikw-as 'evergreen or needle tree' and that of Examples: -ihcikwdn-: nihcikwdn 'my knee,' kaskih- -dskw- as 'wood; hard wood or deciduous tree.') cikwdnehwew 'he breaks his knee by shot.' -tokan-: Examples are akimaskw- 'black ash' (cited by Siebert, nitdkan 'my hip,' kaskitJkaneskawew 'he breaks his 1967a: p. 27 after Faries, 1938: p. 234) and ahcdpdskw- hip by foot.' -hkw-: nihkwdkan 'my face' (expanded 'oak' (recorded by Bloomfield at The Pas, Manitoba); by final -dkan), tomihkwew 'he greases his (own) face,' cf. ahcdpiy 'bow' and contrast ahcdpdhtikw- 'stick kaszhkwenew 'he washes his face by hand.' -stikwdn: for bow.' nistikwan 'my head,' sakistikwdnenew 'he seizes his -apisk- 'stone or solid of similar consistency' has head by hand.' -pwdm-: nipwam 'my thigh,' ohpip- already been exemplified above. Note that here we wdmeyiw 'he lifts his (own) thigh.' -cihc-: nicihciy 'my find a non-zero noun final -w-, as in mohkomdndpiskw- hand,' sakicihcenew 'he seizes his hand (by hand),' 'knife-blade,' wdpamondpiskw- 'glass, window,' cf. nistocihc 'three inches.' -atay-: natay 'my belly,' wdpamon 'mirror'; etc.; contrast the distinctly verbal pdskatayeskawew 'he opens his belly by kicking.' form mihkwdpiskiswew 'he reddens him (stone) by 6.313. However, there are a number of medials heat.' If we compare -apisk- to the extended root which denote body parts but are not attested to occur wdpisk- 'white, light' (6.2), it is tempting to regard in dependent nouns. (They are listed here, rather -apisk- as derived from wapisk- (cf. 6.132(d)) and to than in 6.32, primarily for semantic reasons.) In find a semantic connection in the brightness of stones many cases these medials semantically parallel those and metals, etc. We certainly cannot exclude the of the other set (6.312). possibility that such a connection actually exists in Examples: -nisk- 'arm': sakiniskenew 'he seizes his the "sprachgefiihl" of the Cree; historically, however, arm (by hand).' -ikw- 'neck': sakikwenew 'he seizes the two are distinct: wdpisk- corresponds to Proto his neck (by hand),' kiskikweswew 'he severs his neck Algonquian *-Sk-, Menomini -sk-, Ojibwa -sk- (cf. (by cutting edge),' kiskikwewepahwew 'he severs his Bloomfield, 1946: p. 121), whereas -apisk- is matched neck by throwing a missile.' -iskw- 'head': ohpisk- by Proto Algonquian *-Ok-,Fox -hk-, Menomini -hk-, weyiw 'he lifts his (own) head,' kweskiskwew 'he turns Ojibwa -kk-, Northern Ojibwa -hk- (cf. also Bloom- his (own) head.' -tihp- '(top of) head': paskwatihpew field, 1927: p. 398; 1946: p. 118). 'he is bald-headed,' pahkwdtihpepitew 'he scalps him' Further examples: -ekin- 'cloth or clothlike, ex- (literally 'he peels his head and pulls'). panded object': askekinw- 'raw leather'; mdswekinw- 'moose hide'; masinahikanekinw- 'paper'; taswekinam 6.32. Medials "Classificatory" 'he spreads it out (by hand)'; misekan 'it is an expanse The medials of another group neither function in of ice.' dependent nouns nor are they paralleled by initial -apek- 'rope, elongated object' (cf. Bloomfield, 1946: elements. On the other hand, they freely occur in p. 118): itapekinew 'he holds him thus on a rope (by verbs, nouns, and particles and may thus be considered hand)'; askihtakwdpekan'it is a green string'; niswapek the most narrowly characteristic of the medials. 'in two strings.' Consider the example of -dpisk- 'stone or solid of -ipe- 'liquid' (cf. Bloomfield, 1946: p. 118 and also similar consistency' which occurs in verbs, e.g., 6.332): natipew 'he fetches a liquid'; sikipesin 'he kispdpiskaham 'he closes it with or as metal or stone,' spills a liquid in failing'; monahipepitam 'he drills a 68 WOLFART: PLAINS CREE [TRANS. AMER. PHIL. SOC. well (literally: he digs by tool for a liquid and pulls)'; recurs in the dependent noun ndwikan 'my spine.' iskopew 'he is so deep in water'; moskipew 'he emerges For contrast consider also the verb pakamdpahwew from water.' -ipe- shows a specialized meaning in 'he strikes his eye by tool' where the medial -dp- kiskwepew 'he is drunk' (cf. kiskwe- 'crazy') or kawipew 'eye, vision' is not followed by a post-medial extension 'he falls from drunkenness.' but directly by the final -ahw-. The -e- is clearly an animate intransitive final, 6.33. Extended Medials rather than a post-medial extension, in pakamdskwew Many medials occur in shorter and longer, extended 'he hits wood'; the root is pakam- 'hit,' the medial forms; cf. 6.111. -dskw- 'wood.' Consider such parallel forms as 6.331. Pre-medial extension is rare; one very clear pakitaskwahwew 'he (tree) lets him go (by tool)' with case is that of -kam-.-dkam- 'liquid; body of water.' the transitive animate final -ahw-; or asdskonew 'he The short alternant occurs in kihcikamzhk 'in the sea,' piles him up (as wood) by hand' with the transitive misikamdw 'it is a big expanse of water,' isikamdw 'it animate final -in-. The -e- is also unambiguously is thus shaped as water'; the extended alternant identified as final where it is followed by a suffix occurs in dsowdkamew 'he crosses a body of water,' which is distinctly secondary, e.g., -iski- 'constantly, osdwdkamiw 'it is yellow liquid,' etc. habitually' in notiskweweskiw 'he is a woman chaser'; 6.332. One type of post-medial extension is exempli- the underlying form notiskwewew 'he courts a woman' fied by -ak-. Consider the medial -dkon- 'snow' in also occurs. pihtakonew 'he gets snow into his shoes (literally: he In the vast majority of instances, however, the loads snow),' sohkdkonew 'he is strong enough for identity of the -e- is subject to different interpretations. snow,' moskdkonepitew'he pulls him out of the snow,' Consider the intransitive verb ketaskisinew 'he takes etc; it shows the post-medial -ak- in sipaydkonakihew his (own) shoes off' and the transitive ketaskisinenew 'he makes him go under the snow' or in the particle 'he takes his shoes off by hand, undresses him as to his atdmdkonak 'under the snow' (cf. atam- 'beneath'). shoes.' Both have the derived medial -askisin- (The initial stem corresponding to -dkon- is kon- (cf. 'shoe,' and in the first example it is followed by the 6.132); the -d- of -dkon- might be considered a pre- animate intransitive final -e-. For the second medial extension if there were any instances of just example, however, where the -e- is followed by the -kon-; cf. also Bloomfield, 1962: p. 416). transitive animate final -in- 'by hand,' two interpre- Consider another example: besides the medial tations are possible: (1) The -e- is the animate in- -ipe- 'liquid' (cf. 6.32 for examples) we find an alter- transitive final; then we are dealing with a secondary nant -ipek- occurring in kisipekinam 'he washes it by derivative. (2) The -e- is a post-medial extension; hand'; in the particle atdmipek 'under water'; in then the verb is primary. A decision, in each case, iskopekdw 'the water goes so far'; etc. will have to depend on further semantic and combina- 6.333. The most typical post-medial extension is torial features. -e- which is particularly common where a medial occurs before a transitive final in primary stem 6.4. FINALS formation, e.g., sakicihcenew 'he seizes his hand (by Finals determine the word class of a stem and, in hand)' where the extended medial is -cihce- 'hand' and verbs, also the verb type (5.1). Or consider the stem the final -in- 'by hand.' It is convenient to distinguish between ABSTRACT 'run out of sacred stories' where mestdtayohkanesin-AI and CONCRETE finals, even though they do not con- is the extended medial -dtayohkane- (cf. dtayohkan stitute fully discrete classes. The meaning of concrete of sacred and the 'spirit animal,' subject stories) finals is readily statable, e.g., transitive animate -ahw-, complex final -sin- 'lie, rest.' transitive inanimate -ah- 'by tool.' Abstract finals is with The post-medial element -e- homonymous seem to have no further meaning than to determine final and this the animate intransitive -e-, homonymy the appurtenance of the stem to word class or verb as to the or often leads to indeterminacies primary type, e.g., the -e- typically found with animate The two well secondary status of a stem. -e-'s may intransitive verb involving a medial (cf. 6.333). related 1927: be historically (cf. Bloomfield, p. 399) Most finals are used in both secondary and primary but no attempt is made to untangle this complicated derivation. Some noun finals and some intransitive also the uncertain of Bloomfield situation; (cf. position verb finals seem to be restricted to secondary deriva- 1946: 1958: 1962: in 1927: p. 401, pp. 117, 119; p. 76; tion; see 6.413 ff. and 6.433 ff. pp. 305, 383). It is not the aim of this survey to attempt a listing For further illustration of this situation consider of the immense variety of final suffixes. Thus, only another instance of post-medial -e- in kaskdwikanehwew some or suffixes are 'he breaks his back by shot.' (Note that post-medial particularly productive typical that it is this -e- occurs very frequently but not exclusively with cited as examples. (Note practical medials denoting body-parts.) The root is kask- restriction which leads us to favor secondary suffixes 'break,' the final -ahw- 'by tool'; the medial -dwikan- over the more idiosyncratic primary ones.) VOL. 63, PT. 5, 1973] WORD FORMATION 69

It is often possible to further segment finals into a a number of stems which do not normally end in -w-, more restricted pre-final element and the more widely e.g. asiniy 'stone,' plural asiniyak, do in fact exhibit recurring final proper. The distinction between a final -w- when they function as the base of further pre-final elements and medials is primarily one of derivation, e.g. asinmwispwakan'stone pipe.' freedom of occurrence and in many cases no clear 6.414. Abstract nouns are freely formed from line can be drawn. As an example of a pre-final animate intransitive verbs by the suffix -win-, e.g. consider -s- 'lie, fall' (Proto Algonquian *-hO-)which acimo- AI 'tell a story': dcimowin 'story'; pimatisi- commonly combines with the animate intransitive and AI 'live': pimatisiwin 'life'; nanatawiho- AI 'doctor transitive animate abstract finals -n- and -m- to form oneself': nanatawihdwin 'remedy'; soskwaciwe- AI the complex finals -sin- and -sim-; (note the insertion 'slide downhill': soskwaciwewin 'slide' (action and of connective /i/). Thus, with the root pim- 'exten- place-name; Bloomfield instead recorded the alter- sion in time or space' we find pimisin 'he lies extended' native form soskwaciwdn, derived by the pattern of and pimisimew 'he lays him extended.' Compare 6.415). also pakamisimew 'he strikes him to the ground,' or 6.415. The final -n- is used to form nouns denoting wewekisimew 'he wraps him as he lies.' A middle action, instrument, product, etc. from animate in- reflexive -simo- which corresponds to the transitive transitive verbs. This type seems to be slightly more animate final, is found in kawisimow 'he lies down' or archaic than the one described in 6.414. pahkisimow 'he (sun) sets'; for further examples see Our examples all involve animate intransitive verbs also 6.439. ending in e-d (5.52). According to the treatment of the stem-final vowels before -n-, they fall into two 6.41. Noun Finals sets: one shows the alternant a, the other instead has a short a. The finals described in 6.411 and 6.412 may func- With a: apahkwe- AI 'thatch it': apahkwan 'thatch'; tion in primary or secondary derivation while those kistike- AI 'to farm': kistikan 'grain, wheat'; apwe- AI of the sections are secondary. remaining typically 'make a roast': apwdn 'roast.' Derived (deverbal) noun finals are exemplified in 6.13. With a: atdwake- AI 'sell': atawakan 'fur (literally: 6.411. nouns are and a zero Many unanalyzable what is sold)'; atayohkan 'spirit animal,' from dtayohke- final is conveniently posited, e.g. szsip- 'duck'; AI 'tell a sacred story'; spirit animals are the typical atimw- or horse.' 'dog subjects of sacred stories. Note also the noun 6.412. Concrete noun finals. -waydn- 'hide or atay5hkewin 'sacred story,' derived by the pattern garment made from it': mostoswayan 'buffalo robe,' of 6.414. amiskwayan 'beaver-skin.' 6.416. The suffix -kan- forms mainly nouns which -dhtikw- 'tree, stick': aydhkanahtikw- 'raspberry denote instruments; it is used primarily with transi- bush'; pipikwandhtikw- 'whistle-tree'; sokdwdtikw- tive inanimate stems or with animate intransitive 'sugar maple'; a noun kZkwahtikw- 'what tree' also stems which are syntactically transitive (5.12); it is underlies the verb kikwdhtik6wiw 'what tree is he.' often preceded by connective /i/. -kan- consists of -apoy- 'broth, soup': mihkwapoy 'blood soup'; the suffix -n- (6.415) added to the animate intransi- maskihkkwapoy'tea (literally: herb-broth)'; mdtowind- tive suffix -ke- 'action on general goal' (6.436). poy 'tear (literally: crying-liquid).' Examples: kwdpah- TI 'dip it up': kwdpahikan 6.413. Agent nouns are freely formed from animate 'ladle'; mamitoneyiht- TI 'ponder it': mdmitoneyihcikan intransitive verbs with a final -w- and the third person 'mind'; kisdkamis- TI 'heat it as liquid': kisdkamisikan prefix o(t)- (2.11, appendix A: 6), e.g., pimipici- 'tea-kettle.' Also, pahkwesikan 'bannock' from pahk- 'travel': opimipiciw 'a traveler'; maci- 'hunt': omdciw wes- TI 'cut a piece from it.' 'hunter'; ahkosi- 'be sick': otdhkosiw 'sick person, It occurs also with animate intransitive stems, e.g., patient.' The same prefix-suffix combination is also pZhtwa- 'smoke': pihtwakan 'pipe'; pimiha- 'fly': found with transitive inanimate stems; e.g., omdhamw- pimihakan 'airplane'; ospwdkan 'pipe' from an under- 'one who canoes downriver, voyageur' from mah- lying stem ospwa- 'smoke' which is not attested; 'canoe downriver' thematic suffix (with /am(w)/; cf. nimdskwe- 'carry weapons': nmmdskwakan'weapon.' or even with a 5.71); particle, e.g. onikaniw 'headman, 6.417. The suffix -ihkdn- /ehkan/ forms nouns from leader' from nmkan'ahead.' nouns. The derived noun denotes a substitute or In addition to the "agentive" type just mentioned, surrogate of the denotatum of the underlying stem; there is a multitude of nouns ending in -w-, e.g., also Ellis, 1960. -ihkan- to consist of the kaskitew 'gunpowder' from the root kaskit- 'black,' or cf. appears suffix -n- of 6.415 added to the animate intransitive sehkew 'car,' cf. the particle sehke 'by itself.' Their analysis in most cases is problematic (cf. Bloomfield, suffix -ihke- which forms verbs of making; cf. 6.435. 1962: p. 242). Not only is this -w- homonymous with Examples: pisimw- 'sun': pzsimohkdn 'watch'; the agentive -w-; there may also be some connection niskisikw- 'my eyes': niskZsikohkdna'spectacles (Op)'; with the inflectional -w- of 3.32 and 5.47. Finally, wzhtik5w 'Windigo': wzhtikohkdn 'member of the 70 WOLFART: PLAINS CREE [TRANS. AMER. PHIL. SOC.

Windigo Society'; okimdw 'chief': okimahkdn 'elected nete 'yonder.' The full range is exemplified by the or appointed chief.' interrogative tan-: tdnite 'whither, where,' tdnita 6.418. The suffix -dkan- occurs primarily with 'where,' tdnisi 'how', tdnispi 'when,' tdntahto 'how transitive animate verbs. It forms nouns which many.' denote the goal of the action, e.g. notindkan 'enemy' 6.423. The finals -wdw '(so many) times' and from notin- 'fight him'; wZkimdkan 'spouse' from -wayak 'in (so many) ways or places' represent the wikim- 'live with him or her'; mandcimdkan 'parent- simple (vs. derived) types of concrete particle finals. or child-in-law' from mandcim- 'avoid speaking to him.' Examples: nikotwdw 'anytime' from nikot- 'some, The relation of -dkan- to the inverse theme sign of any'; mihcetwdw 'many times' from mihcet 'many'; verbs /ekw/ and to various other /ek/-suffixes in- tahtwdw 'so many times, every time' from tahtw- 'so volving "passive" meanings is a fascinating problem; many'; finally also tdntahtwdw 'how many times' cf. 5.422. from tdntahtw- 'how many' (cf. 6.422). -wayak: mihcetwayak 'in a lot of ways' from mihcet 6.42. Particle Finals 'many'; tahtwayak 'at every place' from tahtw- 'so many'; and finally tdntahtwayak 'in how many places, The vast majority of particle finals occurs with from how many sides,' cf. tdntahtw- 'how many.' roots which are paralleled by either verb or pronoun Both -wdw and -wayak also occur with numerals: roots (cf. 6.21). Particles whose root does not peyakwdw 'once' (both as numeral and as temporal recur elsewhere, e.g. kanihk 'of course,' are usually adverb), peyakwayak 'in one place,' cf. peyakw- unanalyzable. 'one'; nmswdw 'in two or Medials in fixed association with a final 'twice,' niswayak ways particle nisw- nikotwdsikwdw 'six times,' zero are for see 6.32. places,' cf. 'two'; cf. frequent; examples nikotwdsik 'six.' 6.421. The most characteristic and productive abstract final is /i/, e.g., oht- 'thence': ohci 'thence'; 6.43. Intransitive Verb Finals it- 'thus': isi 'thus'; kandt- 'clean': kandci 'clean'; cf. also appendix A: 5.1. Intransitive verb finals form animate intransitive Also very common is the final -dc (perhaps to be (AI) and inanimate intransitive (II) verbs. Some analyzed into pre-finai -at- plus /i/; e.g., soskwdc intransitive finals come in derivationally and ety- 'right away' from soskw- 'smooth'; nitawac 'as the mologically related pairs, for animate and inanimate best thing to do' from nitaw- 'go to do .... .' In at actor. For obvious semantic reasons, however, least one instance a particle with final -dc is con- paired finals are much less common here than with sistently paralleled by a variant with -dk: e (ya)kwaydc, transitive finals which differ according to the gender e(ya)kwaydk 'just then.' of the goal. A large number of particles end in -taw or -aw, e.g., There is in fact a large subset of inanimate intransi- ohcitaw 'on purpose, expressly' from oht- 'thence.' In tive verbs where to speak of an "actor" is nmisleading many instances, however, the root remains obscure, syntactically as well as semantically; while morpho- e.g., pihtaw 'in the actual outcome'; tipiyaw 'in logically inanimate intransitive, verbs like kimiwan person'; nayestaw 'only'; etc. 'it rains' or yotin 'it is windy, wind' are more aptly labeled 6.422. Concrete particle finals seem to be relatively "impersonal." Most intransitive finals can be used in both few in number. The most common ones are paralleled primary and formation. The suffixes described by the following independent particles: ite 'thither, secondary in sections 6.433 to 6.439 are there'; ita 'there'; isi 'thus'; ispZ 'then'; tahto 'so typically secondary. 6.431. Of the of abstract many.' great variety finals, only one is which is treated in some detail. nikotita somewhere,' nikot- example given Examples: 'just cf. The animate intransitive abstract final -isi- 'some, dstamita, dstamite 'on, to the side /esi/ any'; higher is often inanimate intransitive -d- or time or istam 'come awasita, paralleled by (in space),' cf. here'; -an-. Root sek- 'scare': sekisiw 'he is scared.' Root awasite 'on, to the further side (in time or space),' kaskite- 'black': kaskitesiw 'he is black.' Extended cf. awas 'go away.' The underlying forms are not root wdpisk- 'white': 'he is white,' known in napate 'to one side' and patote 'off alone, wdpiskisiw wdpiskdw from the band.' 'it is white.' Extended root mayat- 'bad (of charac- away 'he is 'it is bad.' the main use of these finals is not with ter)': maydtisiw bad,' nmdytan However, -isi- and -i- are added to all manner of stems. the handful of forms so far but freely underlying given From the misiwe all' we with roots which also occur in pronominal stems. particle 'everywhere, get misiwesiw 'he is all in one from the Thus, with the stem eyakw-, ekw- 'the selfsame' (4.41), piece, entire'; we find ekote 'just thither, just there,' ekota 'just particle nandtohk 'different kinds' (ultimately derived there,' ekosi 'just thus,' and ekospi 'just then.' With from the root nitaw- 'go to do . . .' ?) there is nandtoh- the demonstrative stems (4.11) one finds mainly the kisiw 'he is many different kinds.' Consider also locative particles ote, ta, 'hither, here,' anita 'there,' namakwkwdw 'it has disappeared' which ultimately VOL. 63, PT. 5, 19733 WORD FORMATION 71 derives from the phrase nama kzkway 'nothing, well': miyopayihew 'he makes him fare well'; nohtepayiw absent' (cf. 4.32); since namakikwdw is not a com- 'he runs short': ndhtepayihew 'he causes him to run pound (6.5), we have to assume an underlying stem short'; pimipayiw 'he, it moves along': pimipayihtdw namakikw- which is not attested; cf. 6.433. 'he conducts it'; koskopayiw 'he, it bursts' (not -isi- and -an- also occur freely in complex finals. attested): koskopayihtdw 'he makes it burst.' Thus, -isi- is part of the complex final -dkosi- which As with the various complex suffixes discussed in derives "medio-passive" verbs from transitive in- 6.431, the complex suffixes built on -payi- seem to be animate stems, e.g., iteyihtdkosiw 'he is thus thought treated as units and may be found even where an of' from iteyiht- TI 'think so of it'; nisitaweyihtdkosiw underlying form in -payi- or -payih- is not likely to 'he is recognized' from nisitaweyiht- TI 'recognize it.' occur. This is particularly obvious with the "middle (The other constituents of the complex final are the reflexive" complex final -payiho- in a form like inverse or "passive" marker /ekw/ (5.422) and a kitdpipayihow 'he turns to look,' cf. kitdpi- AI 'look.' pre-final element -d-; -d- most likely is a back-forma- Other examples, with or without underlying forms, tion based on contracted -d- from /aw-e/, e.g., are extremely frequent, e.g. nohtepayihow 'he is in kiskinohamakosiw 'he is taught' from kiskinohamaw- want (for himself),' cf. the underlying form above; TA 'teach (it to) him.' Thus, while the -a- arises kweskipayihow 'he throws himself around,' cf. the regularly in most forms it is part of the complex root kwesk- 'turn'; sdkiskwepayihow 'he throws him- final in others, such as the examples cited above.) self so that his head sticks out,' cf. sdkiskwe- AI The complex finals -ikosi- -dkosi- and -ikwan- 'stick one's head out'; kipipayiw 'he tumbles over,' --dkwan- are often suffixed to finals which denote kipipayihow 'he throws himself over'; nmhcipayiw'he sensory perception, such as TA -naw-, TI -n- 'see,' TA comes, falls down,' nihcipayihow 'he throws himself -htaw-, TI -ht- 'hear'; e.g. ohcindkosiw 'he is seen down'; etc. from there,' wiyasindkwan 'it looks funny' (but 6.433. Verbs of being are freely formed, primarily consider also miyondkohew 'he makes him look well'); from nouns. The animate intransitive and inanimate itihtdkwan 'it is thus heard, it sounds thus,' kiti- intransitive stems are homonymous. mdkihtdkosiw 'he sounds pitiable'; consider also Where a noun stem ends in a vowel followed by miyomdkosiw, miyomdkwan 'he, it smells good'; etc. -w-, the verb final has the shape -i- /i/: iskwew Verbs in -ikosi- -dkosi- and -ikwan- -akwan- 'woman': iskwewiw 'she is a woman'; mosdpew 'single generally denote single actions and thus differ from man: mosadpwiw 'he is a widower'; nehiyaw 'Cree': the "middle reflexives" of 6.439, which denote a nehiyawiw 'he is a Cree'; etc. (Cf. also 3.32 and 6.413.) general, habitual action. Bloomfield (1962: p. 299) All other noun stems take the verb final in the shape also speaks of "a weakening of the passive sense." -iwi- /ewi/, e.g., kon- 'snow': koniwiw 'it is snowy'; -isi- and -an- further combine with a longer alter- mahzhkan 'woll': mahihkaniwiw 'he is a wolf'; kihci- nant of the inverse marker /ekw/ (5.422) to form the -mohkomdn 'American': kihci-mohkomdniwiw 'he is an complex final -ikowisi- 'action by supernatural American'; amiskw- 'beaver': amiskowiw 'he is a powers.' This final is suffixed to transitive animate beaver.' Note that nouns ending in -iy show con- stems, e.g., iteyim- 'think so of him': iteyimikdwisiw traction (/iy-e/ becomes /i/; cf. appendix A: 4.2): 'he is thus thought of by the powers'; pakitin- 'set nipiy 'water,' nipiwiw 'it is water'; askiy 'earth, year,' him down by hand': pakitinikowisiw 'he is set down askzwiw 'it is earth, year.' by the powers'; pakitinamaw- 'set it down for him Verbs of being are not only derived from nouns but (by hand), permits it to him': pakitinamdkowisiw 'he may be formed from particles as well, e.g., kehte 'old': is permitted it by the powers.' kehtewiw 'he is old'; kisiwak 'near': kisiwdkiwiw 'it 6.432. Concrete finals. Example -payi- (animate is near,' etc. and inanimate) 'move': nitopayiw 'he goes seeking, The animate intransitive fiiial -iwi- also appears in goes on the war-path' (cf. niton- TA 'seek him (by namakikwayiwiw 'he comes to nothing'; cf. 6.431 and hand)'); miyopayiw 'he fares well,' miyopayiw 'it goes the references given there. The underlying stem well'; cf. miyo- 'well'; kiwepayiw 'he goes home,' cf. namakikway- which ultimately derives from the kiwe- AI 'go home'; kokzpayiw 'he goes under water,' phrase nama kikway 'nothing, absent,' is not attested. koki- AI 'he with cf. 'dive'; pikwaskisinepayiw goes Example: T534p8 ka-namakzkwayiwindnaw 'we shall torn AI 'tear one's moccasins,' cf. pzkwaskisine- come to the form which occurs in T534p28 'the wind nothing'; moccasin'; kweskitinipayiw changes,' cf. even be as a further derivative kweskitin- II 'the wind turns' and the root kwesk- might interpreted yet will be 'turn.' (cf. 5.85): ka-namakikwayiwindniwiw 'there to Stems with the final -payi- are subject to further coming nothing.' 6.434. Verbs of are formed from derivation, for instance with the causative finals -h- possession freely and -htd-; -h- is transitive animate and -htd- forms possessed themes (3.2) of nouns with the suffix animate intransitive verbs which are syntactically -i- /i/; the possessed theme shows the third person transitive (5.12). Examples: miyopayiw 'he fares prefix o- (2.11). Thus, mohkomdn 'knife': omohko- 72 WOLFART: PLAINS CREE [TRANS. AMER. PHIL. SOC.

mdniw 'he has a knife'; miciwin 'food' (from miciw -ike- /eke/ occurs with transitive animate stems, 'he eats it'): omiciwiniw 'he has food.' e.g., paskiyaw- TA 'win from him': paskiydkew 'he This type of verb also occurs with transitive wins from people'; note the contraction of the stem- meaning, e.g., nikosis 'my son': okosisiw 'he has a final /aw/ with the suffix-initial /e/. Also, tepwes- son, he has him as son'; nimanitom- 'my god': omani- tamaw- TA 'act as announcer for him': tepwestamadkew tomiw 'he has a god, he has him as god.' Consider 'he acts as announcer (for people)'; wihtamaw- TA the following example: T105p15 namoy ihtdw kotak 'tell it for him': wihtamdkew 'he makes predictions'; m(anitow, . . ., t-omanitomihk, . . . 'There doesn't etc. -ike- by no means closes the construction; thus, exist another spirit, . . ., (for a person, indf) to have from notin- TA 'fight him' there is notinikew 'he as god.' Further transitive animate derivatives are fights (people)'; from this may be derived another frequently made with a final -m- ('by speech'?), e.g., transitive animate verb, notinikestamawew 'he fights omanitdmimew 'he has, addresses (?), him as god'; or (people) for him' (cf. 6.446), and then even a further okosisimew 'he has him as (adopted?) son.' derivative, namely the reflexive notinikestamasow AI 6.435. The animate intransitive final -ihke- /ehke/ 'he fights (people) for himself.' is freely suffixed to nouns to form verbs whose meaning With transitive animate verbs there is also another is 'make, gather, produce such-and-such.' Thus, suffix, /iwe/, of similar meaning. Thus, ndt- TA menisk 'trench': meniskehkew 'he digs a trench'; 'fetch him': ndsiwew 'he fetches people'; takohtah- TA matotisan 'sweat-lodge': matotisdnihkew 'he builds a 'bring him': takohtahiwew 'he brings people'; etc. sweat-lodge'; soniydw 'gold, money': soniydhkew 'he The meaning of /iwe/ seems to be emphatically, makes, creates money.' expressly general; consider the following sequence of -ihke- seems to be particularly common with derivations: AI nawaswe- 'pursue' (which is itself derived nouns denoting communal activities, e.g., derived); TA nawaswat- 'pursue him'; AI nawaswasiwe- nitopayiwin 'raid': nitopayiwinihkew 'he arranges a 'pursue people.' raid, leads a war-party'; piciwin 'trek, moving of 6.437. Reciprocal verbs are freely formed from camp': piciwinihkew 'he arranges for the moving of transitive animate verbs with the final -ito- /eto/. camp'; pasakwdpisimowin 'shut-eye dance': pasak- Thus, nipah- TA 'kill him': nipahitowak 'they kill wdpisimowinihkew 'he gives a shut-eye dance.' each other'; kitimah- TA 'ruin him': kitimahitowak 6.436. Morphologically intransitive verbs of action 'they ruin each other'; wihtamaw- TA 'tell it to him': on a general goal are formed with the suffix -ke- wihtamatdwak 'they tell it to each other'; miyowicew- -ike_.83 TA 'get along well with him: miyowicetowak 'they get -ke- seems to occur most typically with transitive along well together.' inanimate verbs or animate intransitive verbs which In some cases we find a slightly divergent formation. are syntactically transitive (5.12); note the insertion Thus, beside wdpam- TA 'see him' there is the recipro- of connective /i/. Thus, mdkwaht- TI 'chew it': cal wdpahtowak 'they see each other.' Bloomfield mdkwahcikew 'he chews, he chews things'; mestin- TI described this form as derived from the transitive 'use it up': mestinikew 'he uses things up'; pakamah- inanimate pendant wdpaht- 'see it' by a shorter TI 'strike it': pakamahikew 'he strikes'; sakah- TI alternant (Proto Algonquian *-wi-) of the suffix 'fasten it (by tool)': sakahikew 'he drives nails.' (1946: p. 108). However, if we assume an alternation Also, kinipota- AI 'sharpen it': kinipocikew 'he of -m- with -h- before -t-, as well as a form of the sharpens things'; nocihtd- AI 'hunt for it': nocihcikew suffix without the initial /e/, namely -to-, then the 'he hunts'; (note that syntactically transitive AI reciprocal verb can be derived directly from the verbs here show an alternant without their final transitive animate stem. There is good independent vowel, cf. 6.444 and 6.446). evidence for such a morphophonological alternation; With other animate intransitive stems, -ke- either cf. 5.73, 5.74, and appendix A: 1.3. Not only would such an the characteristic emphasizes the generality of the goal (which is not, interpretation preserve consonant of the it would of course, expressed syntactically), as in metawdkew suffix; also avoid the semantic of 'he plays with things,' from metawew 'he plays, complications deriving an animate in- contends'; note the further TA derivative metawdkdtew transitive reciprocal verb from a transitive inanimate stem. This latter would have 'he plays (with things) with him.' Or -ke- may serve difficulty been particu- obvious in cases like the where to mark the meaning of the derived stem in a fairly larly following, from idiosyncratic way, as in atdwdkew 'he sells' from a transitive animate verb wikimew 'he lives with her, is married to her' live atdwew 'he trades, buys.' we get wikihtowak 'they with each other, are married.' 83 In his discussion of Algonquian palatalization, Piggott For obvious semantic reasons the reciprocal verbs (1971a: p. 27) sets up only one suffix, which begins in /i/, in spite in forms. of their different environments. The that this is one usually appear plural However, singular assumption back-formations are not uncommon. besides suffix which takes part both in palatalization (of TI stems in t) Thus, and contraction (of TA stems in aw) is crucial to his analysis of wikihtowak 'they are married' we find wikihtow 'he gets palatalization. married.' From the TA stem ohpikih- 'raise him' VOL. 63, PT. 5, 1973] WORD FORMATION 73 we get S67-1 ekwah aylsiyiniw wi-ohpikihitow. 'Now makweyimow 'he feels oppressed,' cf. makohew 'he mortal man is to come into being.' And besides presses hard on him.' nimihitowak 'they make each other dance, they dance,' there is nimihitow 'he dances'; note that the under- 6.44. Transitive Verb Finals lying stem of the TA verb nimihew 'he makes him Transitive verb finals mostly come in pairs, for dance,' namely the AI stem nmmi- 'dance' does not animate or inanimate goal. Most transitive finals seem to occur in Plains Cree. seem to function in both secondary and primary 6.438. Explicit reflexives are freely formed from formation; some exclusively secondary finals are transitive animate stems with the suffix -iso- /eso/. discussed in 6.445 and 6.446. Thus, nipah- TA 'kill him': nipahisow 'he kills Only a very small sample of transitive finals can be himself'; wvh- TA 'name him': wihisow 'he names considered here; we present fairly concrete, instru- himself'; phtaw- TA 'hear him': pehtds5w 'he hears mental finals in 6.441 to 6.443 and more abstract himself'; kiskindhamaw- TA 'teach it to him': kiskino- finals in 6.444 to 6.446. hamdsow 'he teaches it to himself.' 6.441. One of the most common instrumental finals 6.439. Of the great variety of "middle" reflexives is TA, TI -in- 'by hand,' which has the same shape only very few examples can be given. in transitive animate and transitive inanimate stems. One of the most common pairs of finals is animate Many transitive animate finals show a w which is intransitive -o- and inanimate intransitive -e- which lacking in the transitive inanimate pendant, e.g., occur in both primary and secondary derivation. TA -ahw-, TI -ah- 'by tool,' TA -isw-, TI -is- 'by The AI final -o- palatalizes a preceding /0/; cf. heat,' TA -isw-, TI -is- 'by cutting edge,' etc. (These appendix A: 2.4. Thus, beside kdt-: kdtew 'he hides last two pairs are homonymous in Cree but distinct him' and kata-: kataw 'he hides it' (cf. 5.12) we find elsewhere, cf. Proto Algonquian *-esw- 'by heat' and the middle reflexive pair kaso-: kdsow 'he hides' and *-esw- 'by cutting edge.') kdte-: kdtew 'it hides, it is hidden.' Similarly, besides -in- /en/ 'by hand': root it- 'thither, thus': TA TA tahkopitew, TI tahkopitam 'he ties him, it fast' we itinew, TI itinam 'he moves him, it thither or thus by find AI tahkopisow 'he is tied fast' and II tahkopitew hand.' Root oht- 'thence, therefore': TA ohtinew, TI 'it is tied fast.' ohtinam 'he takes him, it thence or therefore.' Root The same suffixes recur in the complex finals AI kis-- 'agitate,' medial -ipek- 'liquid': kisipekinam 'he -kdso- and II -kdte-; (whether these are in any way washes it.' -in- occurs as a primary final in the AI related to the suffix -ke- (6.436) which forms verbs verb kiszpekinisitew 'he washes his (own) feet' which with general goal, remains to be seen). Thus, from is derived from the stem kisipekin-, above. It is a the transitive inanimate stem masinah- 'mark, write secondary final in TA kisipekistikwanenew 'he washes it' and besides the animate intransitive verb masina- the other's head' which is derived from the AI verb hike- 'write' we find masinahikdssw 'he is marked, kisipekistikwdnew 'he washes his (own) head.' pictured' and masinahikdtew 'it is marked, pictured, TA -ahw-, TI -ah- 'by medium': Root pat- 'miss': written.' Or besides wihtam 'he names it, tells it' TA patahwew, TI pataham 'he misses him, it by tool we find wihcikasdw 'he is named or told' and wihci- (or shot)'; cf. patinew 'he misses him (ball) by hand katew 'it is named or told.' (in catching).' -ahw-, -ah- also occur with pre-final Another very frequent type of "middle reflexives" -at- which specifies the tool as being stick-like or adds an animate intransitive final -o- to transitive having a handle, e.g., ksskataham 'he chops it off by animate stems. Thus, with the complex final -sim- axe,' kiskikwetahwew 'he severs his neck by axe,' root 'make him lie' cited in 6.4 we get a middle reflexive kzsk- 'sever,' medial -ikw- 'neck.' For the meaning complex final -simo- as in pahkisimdw 'he (sun) sets,' 'by external medium' consider ndtahwew 'he fetches it cf. pahkisimew 'he lets him fall.' This complex final by water,' root nat- 'fetch'; sipweham 'he utters it,' also occurs freely in the specialized meaning 'place root sipwe- 'out, off'; ohosimdwaham 'he sings the oneself, dance,' e.g., pimisimdw 'he dances along,' Owl Dance,' stem jhdsimo- AI 'dance the Owl Dance.' newosimow 'he dances as one of four (in a fiddle- TA -isw- /esw/, TI -is- /es/ 'by heat': Root kis- dance),' pasakwapisimow 'he dances with eyes shut,' 'complete': TA ksiswsw, TI kisisam 'he completes ylwacayesimow 'he loses his belly by dancing,' etc. him, it by heat, i.e., he cooks him, it done.' Root -o- also occurs with the transitive animate stem pdsk- 'open' TA pdskiswew, TI pdskisam 'he shoots dcim- 'tell of him' to form dcimo-: dcimow 'he tells of him, it with gun; he hits him, it with lightning' himself, narrates.' It is frequently found with stems (literally: 'he opens or uncovers him, it by heat'). ending in the complex transitive animate final -eyim- Root mihkw- 'red,' medial -apisk- 'stone or metal': 'think of him,' e.g., kitimdkeyimew 'he takes pity on TA mihkwdpiskiswew 'he reddens him by heat as stone him': kitimakeyimow 'he takes pity on himself, he or metal.' feels pitiable.' As a complex final, -eyimo- means TA -isw- /esw/, TI -is- /es/ 'by cutting edge': 'feel in such a way' and may be used even where no Root kisk- 'sever': TA kiskiswew, TI kiskisam 'he cuts transitive animate pendant seems to occur, e.g., him, it through or off.' Root man- 'take, get': TA 74 WOLFART: PLAINS CREE [TRANS. AMER. PHIL. SOC. maniswew, TI manisam 'he cuts him, it to take; he (food) goes through his body with good effect, does mows it (grass).' Root kisk-, medial -ikwe- 'neck': him good,' TI miyoskam 'he has a good fit of it.' TA kiskikweswew 'he cuts his throat'; consider also Root past- 'break': pdstiskam 'he breaks it by foot.' the AI verb kiskikwesikew 'he cuts throats' which is 6.444. The most common abstract final is transitive derived by the suffix -ke- (6.436) of general action animate -h-. -h- is paralleled by -hta- which forms, from an unattested TI stem kZskikwes-'cut its throat.' not transitive inanimate stems but syntactically 6.442. TA -m- 'by mouth, by speech; (by thought)' transitive AI stems (5.12); cf. also 6.436. is particularly frequent in the latter meanings. (For Thus, from the root kzs- 'complete' we get TA various other finals of the same shape cf. Bloomfield, kzsihew 'he completes him' and AI kisihtdw 'he 1946: p. 113.) completes it'; also TA kaskihew, AI kaskihtaw 'he Examples: Root nito- 'seek': nitomew 'he calls or manages, controls him, it'; TA wanihew, AI wanihtdw invites him.' Root kis- 'hot, angry': kisimew 'he 'he loses him, it'; TA mosihew, AI mosihtdw 'he angers him by speech.' Root sihk- 'push': sihkimew perceives his, its coming or presence'; kdskohew 'he 'he urges him by speech.' kzhkihkimew'he persuades startles him,' cf. koskomew 'he startles him by call,' him against his will,' particle kihkihk 'in spite, kdskonew 'he startles him by hand'; etc. nevertheless.' -h- and -htd- also function in secondary derivation: TA -m- is often paralleled by TI -ht- (to be analyzed manatisiw 'he acts discreetly': TA mandtisihew 'he as -m- plus -t-; cf. appendix A: 1.3), e.g., root makw- spares him,' AI manatisihtdw 'he is careful of it.' 'press': TA makwamew, TI makwahtam 'he bites or waweyiw 'he gets ready': TA waweyihew, AI waweyih- chews on him, it.' Root paskw- 'clean, clear': TA taw 'he gets him, it ready.' paskomew, TI paskohtam 'he cleans or clears him, it by A few verbs show what seems to be an extended mouth.' Root (?) pahkwe- 'break into pieces': TA alternant of -h- and -htd-: TA isihew, AI isihtaw 'he pahkwemew, TI pahkwehtam 'he bites a piece from makes him, it so,' cf. it- 'thus'; TA osizhw, AI osihtdw him, it.' 'he makes, arranges him' (cf. os- in osapiw 'he looks -m- and -ht- typically combine with the pre-final from there'); AI misihtaw 'he makes it big,' cf. the -eyi-; the complex suffixes -eyim- and -eyiht- denote particle misi 'big.' the action of the mind. Examples: Root it- 'thus': Secondary verbs which are formed with a transitive TA iteyimew, TI iteyihtam 'he thinks so of him, it.' animate suffix -h- often have a causative meaning, Root oht- 'thence, therefore': ohteyimew 'he is jealous e.g., nikamohew 'he makes him sing,' cf. nikamo- of him.' Root kwetaw- 'impatiently': kwetaweyimew 'sing.' (Whether this suffix can actually be identi- 'he misses him.' TA mamitoneyimew, TI mami- fied with the -h- discussed above is yet to be fully toneyihtam 'he ponders over him, it'; reduplicated, cf. determined.) Further examples: miyondkohew 'he the particle mitoni 'really.' TA ayiwdakyimew, TI makes him look well,' cf. the stem miyondkw-isi- ayiwakeyihtam 'he thinks more of him, it', cf. the 'look well'; kiskeyihtamohew 'he makes him know it,' particle ayiwdk 'more'; etc. cf. kiskeyihtam 'he knows it.' Some of the underlying 6.443. Transitive animate finals often differ from stems show different alternants (with e and short a) their transitive inanimate pendants by the addition when combining with this suffix; AI metawe- 'play': of -aw-. TA metawehew 'he makes him play'; AI tipahike- TA -ihtaw- /ehtaw/, TI -iht- /eht/ 'hear, by 'pay': TA tipahikehew 'he makes him pay'; AI hearing': Root it- 'thus': TA itihtawew, TI itihtam pzhtoke- 'enter': TA pihtokahew 'he makes him go 'he hears him, it so.' Root nito- 'seek': TA nito- inside'; AI takohte- 'arrive': TA takohtahew 'he brings htawew, TI nitohtam 'he tries to hear him, it.' Root him,' etc. miyw- 'well': TA miyohtawew, TI miyohtam 'he likes 6.445. The abstract finals TA, TI -t-; TA -staw-, the sound of him, it.' Stem kitimdk- 'pitiable': TI -st-; TA -totaw-, TI -tot- are used to derive transi- kitimiakihtawew'he hears him with pity.' tive verbs from animate intransitive stems. If the TA -naw-, TI -n- 'by vision': Root it- 'thus' (plus derivational suffixes add a further meaning to the connective /i/): TA isinawew, TI isinam 'he sees him, resulting stem, it is yet to be discovered. it so.' Root nito- 'seek': TA nitonawew, TI nitonam -t- /0/ derives both TA and TI stems. kito- 'call': 'he seeks him, it.' Root (?) nisitaw- 'recognize': TA TA kitotew, TI kitotam 'he talks to him, it.' With AI nisitawinawew, TI nisitawinam 'he recognizes him, it stems ending in e-d, -t- is added to the a-alternant: by sight'; kitimakinawew 'he looks on him with pity,' pikiskwe- 'speak': TA pikiskwatew, TI pikiskwatam cf. above. 'he speaks to him, it.' sakowe- 'call, yell': TA TA -iskaw- /eskaw/, TI -isk- /esk/ 'by foot or sakowatew 'he whoops at him.' nawaswe- 'pursue': body movement': Root miskw- 'find': TA miskos- TA nawaswatew 'he pursues him.' notiskwewe- 'he kawew, TI miskoskam 'he finds him, it with his foot courts a woman': notiskwewdtew'he courts her.' or body.' Root kik- 'have along; with': TA kikis- TA -staw-, TI -st-: acimo- 'narrate': TA acimostawew kawew, TI kikiskam 'he wears him, it; he goes having 'he narrates to him.' kwaski- 'turn': TA kweski- him, it.' Root miyw- 'well': TA miyoskawew 'he stawew 'he turns to him.' Where -staw- is added to a VOL. 63, PT. 5, 1973] WORD FORMATION 75 stem ending in -i-, that -i- is lengthened, e.g. wasaka- However, many verbs which appear to be formed mepayi- 'move in a circle': TA wdsakdmepayistawew this way have a different meaning; that is, the goal of 'he circles him'; consider also pimitehcipayistawew 'he the hypothetical TI stem which would form the rides along with him' (pimi- 'along,' -tehci- 'on intermediary stage between the AI stem and the horseback'). With AI stems ending in e-d, -staw- is eventual TA stem, does not seem to appear in the added to the e-alternant, e.g., pimitehcikocikdwe- meaning of the TA stem. Thus, a different analysis 'race (-kocikdwe-) along on horseback': pimitehciko- seems indicated: that there is also a complex final cikawestawew 'he races along with him on horseback.' -stamaw- which derives verbs of action on a general In the following two examples the underlying goal with a transitive animate beneficiary from AI forms are not attested; in each case the root it- stems. Examples: tepwe- 'speak, call': TA tepwe- 'thither or thus' is followed by a medial, and the -e- stamawew 'he acts as announcer for him.' pikiskwe- is ambiguous (cf. 6.333). itiskwe- AI 'have one's 'speak': TA pikiskwestamawew 'he speaks for him'; a face thither or thus': TA itiskwestawew 'he faces him'; further AI stem may then be derived by the reflexive isiniske- AI 'have one's arm thither or thus': TA suffix -iso-: AI pikiskwestamdsow 'he speaks for isiniskestawew 'he makes hand signs to him.' himself, he prays.' Consider also AI nikamo- 'sing'; Transitive animate verbs formed by -staw- are while the TA stem nikamostamaw- 'sing for him' is paralleled by transitive inanimate verbs in -st-, e.g., not attested, we find an AI stem of action on a nahapi- 'sit down': TA nahapistawew, TI nahapzstam general goal which is derived from it by the suffix 'he sits down by him, it'; moskH- 'come forth': TA /eke/ (6.436): nikamostamdkew 'he makes music for moskistawew, TI mosktstam 'he attacks him, it.' How- people.' Finally consider the root not- 'pursue' and ever, syntactically transitive AI (5.12) parallels also the stem TA notin- 'fight him' (-in- 'by hand'); AI occur, e.g., nepewisi- 'be bashful': TA nepewisistawew, ndtinike- 'fight (people)'; TA notinikestamaw- 'fight AI nepewisistdw 'he is bashful about him, it.' (people) for him'; and finally AI notinikestamdsow TA -totaw- and TI -tot- also occur with the e- 'he fights (people) for himself.' alternant of AI e-stems, e.g., kzwe- 'go home': TA The final -aw- also derives transitive animate stems kzwetotawiew'he goes home to him'; sdkewe- 'come into from syntactically transitive AI stems (5.12). It is view': TA sdkewetotawew, TI sdkewetotam 'he comes added to an alternant of the underlying stem which into view of him, it.' Further examples: dkayaszmo- lacks the final vowel. Thus, nahastd- 'place it right, 'speak English': TA dkaydsimototawew 'he speaks put it away': TA nahastawew 'he places it right for English to him'; kawisimo- 'lie down': TA kawisimoto- him.' Consider also the pair of derived stems already tawew 'he lies down with him'; etc. cited in 6.445: TA nepewiststawew, AI nepewisistdw 'he is bashful about it.' 6.446. Transitive animate double-goal verbs (5.11) him, are derived from transitive inanimate verbs with a This is obviously an area of extreme productivity suffix -aw- which follows the transitive inanimate and considerable fluctuation. It is not too surprising, theme sign -am-. then, that parallel to syntactically transitive AI verbs we also find transitive verbs derived with The meaning of these verbs clearly reflects their AI kimotiw 'he steals it': TA kimotamawew morphological structure: the inanimate goal of the -amaw-, e.g., 'he steals it from him.' both formations underlying stem, although not cross-referenced in the Indeed, may be found with the same from AI derived verb, is still the primary object, and the stem; thus, nipahtdw 'he kills it' we both TA and TA animate goal of the derived stem is the secondary get nipahtawew 'he kills it for with no object; since in the great majority of instances it is nipahtamawew him,' apparent difference in the beneficiary of the action, we may also call these meaning. verbs also 5.814. benefactives4;cf. 6.5. COMPOUNDS Examples: dtot- 'tell of it': TA dtotamawew 'he tells of it for him.' nakat- 'leave it': TA nakatamawew 'he Compounds combine certain characteristics of leaves it for him.' manis- 'cut it': TA manisamawew phrases and of unit words. (The term "compound" 'he cuts it from or for him.' in this specific use is adopted from Bloomfield.) The This suffix may also be added to TI stems which are members of a compound are separated by a hyphen. derived from AI stems by the suffix -st- of 6.445. Compounds differ from unit words and resemble Thus, from moski- 'come forth' there is a derived TI phrases in that the sandhi between compound mem- stem moskist- 'come forth towards it, attack it'; TA bers is of the external type; that is, even though not moskistamawew, in turn, means 'he attacks it for him.' all compound members actually occur separately as free forms (6.52), they are nevertheless treated like 84 As a historical aside, it may be noted that the missionary words phonologically. For example, the particle isi fathers who consideredthis formation part of inflection, regarded 'thus' and the verb atoskew 'he works' form a it as unique among languages: "C'est une richesse que la langue may Crise seule possede" (Faraud in Lacombe, 1874b: 186). Faraud compound is-dtoskew 'thus he works' which clearly also provides a term which beautifully suggests his sentiments: shows the effect of external sandhi in the loss of the -i cas vicaire. and the lengthening of the a-. (The existence of a 76 WOLFART: PLAINS CREE [TRANS. AMER. PHIL. SOC.

variant form isi-atoskew is due to the optional nature Conversely, it- is clearly a verb root in itatoskew 'he of external sandhi and does not affect the status of works so'; cf. the compound is-atoskew, above. Simi- compound words.) Furthermore, the final vowels of larly, mahkaess 'fox' is clearly a compound member, compound members are subject to the same gradual rather than a derived (deverbal) suffix, in mdski- devoicing as those of simple words; in Bloomfield's -mahkesis 'lame fox' since it is paralleled by a non- texts this is symbolized by h, e.g., S44-4 ey-isih- initial alternant -ahkesis-, as in wdpahkesis 'white fox.' -tapasit 'he thus fled.' Compound words differ from phrases by showing 6.51. Nominal Compounds unit words. This prefixation and suffixation just like Nominal consist of a noun as second in the case of compounds difference is particularly striking member and either a noun or a as first verbs where the verb stem follow a particle compound may member. oski-minisa 'fresh berries' (minis are Examples: series of preverb particles. There the prefixes is the non-initial attached to the first rather than to the initial 'berry' paralleled by -min-); soniydw- preverb -okimdw Indian (okimdw 'chief' Sll-23 kiwih- 'money-boss, agent' element of the verb stem itself, e.g., is a non-initial alternant -ikimdw- so to kill me.' paralleled by -kakwe-nipahin 'you are going to try that a unit word would have A: corresponding probably Furthermore, initial change (5.332, appendix 7) the first member of the thus shape (contracted) soniyakimaw). also affects the compound, Bloomfield a -i- which to the its constituent posits special ending /i/ attesting tight linkage among noun stems take when functioning as the prior member members. For the preverb wi 'intend to' example, of a compound (1946: p. 103, 1930: p. 5; cf. also 1958: in changed form in S247-17 wdh-pimdcihdt- appears p. 41); e.g., paskwawi-mostos 'buffalo,' cf. paskwdw wdwi 'whenever they were going to revive him.' 'prairie'; atimo-kisayiniw 'dog of an old man,' cf. While are distinct from unit words compounds atimw- 'dog'; maskeko-sdkahikan 'Muskeg Lake,' cf. the heuristic of theoretically, problem distinguishing maskekw- 'muskeg.' However, this "suffix" seems to the two is considerable. The use of morphological occur only where the second member begins in a criteria is often vitiated two sets of by homonymies: consonant (otherwise it would result in a lengthening with their non-initial of initial elements (suffixal) of the initial vowel of the second member); thus it alternants and of connective an (6.131); /i/, empty greatly resembles connective /i/ which is typical of morph, with the near-ubiquitous particle final /i/ internal sandhi. Whether these forms should be The semantic which well be (6.421). criteria, might regarded as unit words (so far, no distinctly initial the most reliable and the easiest to the use, require elements have turned up as second members), or services of a informant. fairly sophisticated whether noun composition needs to be treated differ- When such homonymies occur and semantic criteria ently from verb composition, remains to be studied cannot be applied, a decision can be made only if the in detail. (Note that the distinction of unit word effects of external sandhi are obvious (which they and compound "has been troubled in Cree" (Bloom- need not be), or if one of the members is otherwise field, 1930: p. 5) in any case; the details go beyond distinctly marked as free or bound. In a large the scope of this survey but see also Bloomfield, 1930: number of cases no such criteria are available; p. 72, footnote). consider the root it- /eO-/ 'thus' and the stem tehtapi- 'ride on horseback.' it- shows the alternant is- both 6.52. Verbal Compounds in the particle isi and when followed by connective /i/, as it would be in combination with tehtapi-; Verbal compounds consist of one or more PREVERB tehtapi- has the same phonemic shape as (initial) stem particles combined with a verb stem. Preverbs and as noninitial element. Thus, in the absence of belong to two position classes. The preverbs of unambiguous phonological evidence (such as the position 1 are few in number and mutually exclusive. devoicing of the final -i- of isi: isih), the phonemic The preverbs of position 2, by contrast, constitute sequence isitehtapiw cannot be unambiguously inter- an open class of particles several of which may occur in succession. preted as either unit word or compound. The loosest of is after the last In many instances, however, one of the compound point linkage preverb and before the other material be inserted members is clearly marked as free (or, conversely, stem; may at this after such an of a unit word is point. Usually, however, part obviously bound). Thus, insertion is the breaks off and forms mistahi is a free form in the begun, speaker 'big, great' compound the compound all over again; so S239-6 ka-kape-kisik mistahi-maskwa the root 'Big Bear'; corresponding kd-tah-tehci-kwdskohtiyit 'all day jumping down (on mist- 'big' occurs in mistasiniy 'big stone.' The free them)' where the insertion is, or would have been, status of these compound members is also evidenced kape-kisik 'all day.' by the fact that the monosyllabic noun stems retain Although treated like words with regard to sandhi, their suffix vowel (cf. 3.311), e.g., wdkayosi-wati not all preverbs actually occur as independent words. 'bear-den,' kihci-mihti 'big club.' Those of position 1 occur only as preverbs. Some VOL. 63, PT. 5, 1973] WORD FORMATION 77 preverbs of position 2 also occur only in this function, S42-23 kiya nama nantaw ka-kih-ohtinamdson ta- e.g., pe 'hither' or ati 'progressively'; the vast majority, -mZciyin. 'you, you wouldn't be able to get anything however, also occur freely as mere particles, e.g., isi to eat, anyway'; S243-12 nimise, taniteh maka ke-kih- 'thus,' ohci 'thence, therefore,' newo 'four,' kamwdci -miskaman minisah? 'Big sister, where, however, will 'quietly,' etc. you be able to find berries ?' Some of the preverbs of position 2 may even be 6.522. More than one of the preverbs of position 2 reduplicated, e.g., S43-22 niwah-wani-kiskisin 'I may be present. remember very dimly' where wdh adds emphasis to Contrary to the claim of Edwards (1954: p. 17), wani 'dim, dark.' no order of occurrence has been established among 6.521. The preverbs of position 1 are mutually position 2 preverbs, although kil 'past,' k2 'able to,' exclusive. and wZ 'intend to' tend to precede, and isi 'thus' to e indicates subordination in an entirely neutral follow all others. (But consider P266-20 kit-si-ki- way. It is formed by initial change from an under- -ohci-nipahikawiyahk 'by what means each of us can lying form /a/ which, though reflected in Ojibwa be killed' which runs counter to both the above (Bloomfield, 1958: p. 62), does not occur in Cree. assertions at once.) At least when following the We interpret it as an empty "vehicle" for initial position 1 preverb kd, o 'from there, therefore; change since it seems to be the latter which actually originally' precedes all other preverbs. does the subordinating; cf. 5.332. Position 2 preverbs are semantically ranged along a Examples: P254-8 oki mac-dyTsiyiniwak eh-mihce- continuous scale from abstract to concrete; consider, ticik 'these evil men who were many'; P266-1 acosisah for instance, isi 'thus' or ktl 'past' versus matwe mitoni eh-apisdsikih 'very small arrows' (literally 'audibly' or misi 'much, big.' Such a semantic 'arrows being very small'); S12-28 e-kih-wihtamdkot, classification also seems to correlate with the relative nipahew ohi. 'When he (3') told him (3), he (3) freedom of occurrence of position 2 preverbs and, killed this one (3').' perhaps, with features of internal syntax yet to be kd is historically the changed form of the preverb explored. For practical purposes, position 2 preverbs kil 'past' but its primary role now is that of a sub- are treated as if they constituted two discrete classes. ordinator, in which function it may in fact be followed Examples of "abstract" preverbs. dta 'although, by kli. The term "relative," applied to it by Ellis in vain': S41-31 eh-dtah-kit6ott 'although he spoke to and others, is applicable to only part of its range. him'; S13-2 iydtah-pey-itohteyici 'when in vain they Examples: S237-37 oskinmkiskwewkd-pesiwak 'the came there.' young woman (3) I have brought (1-3)'; S244-14 k1 'past': S246-11 nikth-pe-maskamikawin 'he was ohi oskinikiwah kd-kih-wdpamdt 'that young man (3') taken away from me.' kil rarely occurs with a whom she (3) had seen (3-(3'))'; S236-10 . . . eh- negative particle, e.g., S13-32 ekotah ohcih nama -wepinahk, kd-peh-pzhtoketdcimoyitiskwesisah. '. wthkdc esah kth-nipiwak aySsiyiniwak. 'From then when he threw it out (3), there came crawling inside on people never died'; the preferred preverb for a past (3') a little girl (3')'; S239-27 ". . ." eh-itweyit, negative statement is ohci. mekwdc ostesah kd-pehtawdt. '". . ." they (3') said, kH2'able to' normally occurs with a negator (ekd, while she (3) listened to her brothers (3').' namoya, etc.) or after the future markers ka and kita, ka and kita, the latter optionally reduced to ta, ta; it does not undergo initial change. S14-1 namoya mark subsequence or futurity. In preterit forms or nika-kih-itwan. 'I cannot say'; S63-3 tanisi ke-kih- in simple conjunct clauses, the future preverbs com- -totamdn? 'What shall I be able to do?' S63-22 bine with the position 2 preverb k;i 'past' to indicate ewak ohci ek dwiyak k-5-kih-nipahit. 'That is why irreality; cf. 5.322. no one can kill me.' Although ka seems to be more intimately associated o, ohci 'from there, therefore; originally' is based with forms involving speaker and/or addressee, all on a relative root (6.23) and thus usually has an three preverbs are freely interchangeable in most antecedent. o and ohci differ tactically: (1) & func- contexts. However, kita, ta does not occur with the tions only as a preverb while ohci is freely used as a personal prefixes ki- and ni-. Also, only ka is subject particle, as in P10-34 below. (2) While ohci is not to initial change, yielding ke; the unchanged ka does attested with kd, o does not seem to occur with e. not occur in conjunct forms. The shape ka also (3) Occurring immediately after kd, 6 precedes all occurs as a contraction (haplology) of the personal other preverbs; the relative position of ohci is not prefix ki- followed by the preverb ka; only this fixed. P2-5 acosis pikoh kih-ohcih-nipahewak 'with shorter form occurs before the preverb ki2 'able to.' merely an arrow they killed them'; P262-21 ekotdwah Examples: S243-9 ekwah kitahtawe kd-kiskeyihtahk nik-dh-nipahdwak 'with such I shall kill them'; kitah-pimihdt. 'Then, presently, she knew that she P10-34 eyak ohci yospisiwin k-oh-aydt nehiyaw. 'That would fly'; P4-33 nika-miyweyihten niya mina wdhyaw is why the Cree has gentleness.' When it occurs kit-etohtemakahk pikiskwewin,. ... 'I myself shall with a negator (ekd, namoya, etc.) 6, ohci takes the be glad that far away my speech will go, . . '; place of kil which does not normally occur with one; _ _ _P_ I_ 78 WOLFART: ]PI LAINS CREE [TRANS. AMER. PHIL. SOC.

such expressions usually indicate a very remote past. Examples of "concrete" preverbs. kamwaci P2-4 nama kekway ohc-dydwak iskotew. 'Originally 'quietly': S13-33 kih-kamwaci-pimdtisiwak 'they lived they had no fire'; P12-8 ayisk namoya niyanan noh- quietly'; matwe 'audibly': S237-19 matweh-nipaha- were to -nipahanan manitow okosisa; 'For not we have ever ciyiwa 'they (3') audibly freezing death'; slain God's son.' mecimwaci 'permanently': S44-16 ta-mecimwdcih- -nahapiw 'let him sit down for good'; mesci 'exhaus- wi intend to': S43-37 wi-mescihdwak 'will, 'they tively': S244-37 eh-kih-mesci-sipwehteyit 'when every- will all be killed' (indf-3p); S248-2 e-wZh-pe-nipahikot one of them had departed'; newo 'four': S245-25 'as he (3') was about (wi) to come (pe) and kill him eh-kih-newo-tipiskayik 'when the fourth night had (3).' passed.' APPENDIX A: MORPHOPHONOLOGY latter is used only where it is immediately relevant to the discussion.86 0.1. Transcription In the morphophonological representation we use The transcription system used in this study is all the symbols of the phonemic representation, plus essentially that of Bloomfield. While it is adequate the special symbols /0/, /e/, /L/. as a practical orthography, it does not reflect a com- prehensive and detailed analysis of Cree phonology. 0.3. Summary the issues of Cree Among many problematic For the task at hand, namely paradigmatic analysis, phonology, the status of the , syllable statements of internal combination are most con- the delimitation of the structure, phonological word, veniently formulated as replacement rules.87 and the and intonation seem stress, pitch, system Thus, rules (1) to (4) are replacement rules oper- most in need of In urgently investigation. presenting ating on morphophonological symbols. They are his Bloomfield mentions orthography (1930: pp. 1-6) partially ordered: (1) precedes all others. Rule (1) some of these for the of points, example interplay is also ordered internally. stress and syllabification, and Longacre (1957) has After the rules of internal combination have been published a brief of vowel But far study length. applied, (R 1) and (R 2) which are ordered with too much of Cree phonology has remained unexplored. respect to each other, yield the actual phonemic The following symbols are used in the phonemic transcription: shapes. This summary includes only the more generally consonants: ptckshmn applicable statements in a general form. For details semivowels: w y see section A: 1 through A: 5, below. This summary vowels: i a o a also omits prefixation (A: 6) as well as initial change and indicate and word c ranges from a blade-alveolar to a dorso-laminal (A: 7). Hyphen # morpheme affricate. Word-final h appears to be non-distinctive boundary, respectively. (for its morphophonological role cf. A: 5.1); there is (1) (a) w-w-- w some fluctuation in Bloomfield's texts which appear - h "as they were actually taken from dictation." For a (b) (m,n)-(k,t) (k,t) brief discussion of stress, external sandhi, and surface (c) C-C -> C-iC variations see Bloomfield, 1930: pp. 1-6. (2) 0-(i,i,y) - s(i,i,y) 0.2. Scope t-(i,I,y) - c(i,i,y) The morphophonological statements85 given in this appendix are based primarily on the alternations of (3) V-V -+ VyV inflectional affixes and of entire stems. Most of them will also reflect the alternations which take place in Vl-V2 -V2 derivation but no attempt is made to fully incorporate the latter. Within these restrictions, Bloomfield's V1-V2 -> V1 treatment of Menomini internal sandhi (1962: pp. V-L- V 78-100) is followed closely. o-w - ow Phonemic representation is indicated by italics, morphophonological representation by slashes. The 86Although this policy may result in loose formulations, as when phonemicallyrepresented stems are combinedwith morpho- 86The terms "phonemic"and "morphophonological"are used phonologically represented suffixes, it is nevertheless adopted throughoutas a matter of convenience. Their use is not intended because the underlyingforms, especiallyof stems, are by no means as a commitment for a "taxonomic" approach to phonology or always known. Even with the suffixes,only internal evidence is against, say, a distinctive feature approach; the particularchoice used; historicalevidence could not usefully be introducedwithout of words is considered to be largely accidental and other terms, a full historical account of the various paradigms;cf. 5.6. e.g., "segment" and "underlyingrepresentation," might be sub- 87The present sketch of internal combination is, of course, stituted without greatly affecting the discussion. In short, given analysis-oriented. In a more detailed study of morphophono- the morphological-semantic emphasis of the present study, logical alternation, the insertion of connective /i/ would be dis- "phonemic" and "morphophonological"are seen as convenient tinct from the remainderof rules (2) to (4); /0/ would be treated levels of phonological representation which may later be de- as the higher-levelrepresentation of the alternation of phonemic veloped in one of several different directions. t and s; etc. 79 80 WOLFART: PLAINS CREE [TRANS. AMER. PHIL. SOC.

(4) Cw-iC, Cw-eC --, CoC is kinasin 'you fetch me'; with the corresponding 1-2 the form is kinatitin 'I fetch Vw-e, Vy-e -- V ending /etin/, resulting you.' The palatalizing suffix may be subject to apocope 1) V -- 0 in env. _# (R (A: 5.1); for instance, when the dependent noun w -- 0 in env. C_# stem /-iwa0-/ 'pack' is followed by the inanimate proximate singular suffix /i/, the result is nmwas'my with the locative suffix we (R 2) 6 -- t pack'; /ehk/, by contrast, get niwatihk 'in my pack.' 2.2. /t/ is replaced by /c/ before /i,i,y/. Else- p, t, c, k . . ->p, t, c,k . . . where, e.g., before /e/, /t/ remains. For example, consider the third person conjunct 1. CONSONANT SEQUENCES suffix /t/; when it is followed by the plural marker /k/ and connective /i/ (A: 1) occurs, the resulting 1.1. Connective When a /i/. morpheme ending form is /cik/. in a is followed a non-syllabic by morpheme-initial Conversely, from the dependent noun /nisit-/ 'my consonant, a CONNECTIVE is inserted /i/ normally foot' and the locative suffix /ehk/, we get nisitihk between them. Consider the transitive animate con- 'on my foot.' for 1-3 and and when junct endings 2-3, /ak/ /at/; However, is not without exception. are followed the marker palatalization they by third-person plural Thus, the stem wat- 'hole' is followed by the inanimate is In the case of /k/, /i/ inserted, e.g. 1-3p /ak-i-k/. proximate singular suffix /i/ in wati 'hole.' Whether the shows 2-3p /at-i-k/, sequence /t-i/ palatalization the t corresponds to /t/ or /0/, palatalization would (A: 2.2), yielding /acik/. be but does not occur. Conversely, consider 1.2. Connective does not occur before semi- expected /i/ the combination of the root not- /not-/ 'hunt' with vowels. For the third example, person conjunct -acaskwe- in nocacaskwew 'he hunts muskrats'; unless suffix followed the marker /t/ by plural /waw/ yields this is simply a matter of non-contiguous assimilation, Or consider the third /twaw/. person plural ending a model for analogical leveling is easy to find: nocih- with a stein in . . we /wak/; ending ./n/ get 'hunt him' from /not-/, /-h-/, and connective /i/. pimisinwak lie.' ./nwak/, e.g., 'they While the alternation of t and c is "partly troubled" Where a in is followed morpheme ending /w/ by (Bloomfield, 1930: pp. 5, 72) in Cree, that of t and s one in one remains. Con- beginning /w/, only /w/ "had suffered analogic disturbance [even] before sider the which marks the in- morpheme /ahkw/ Proto Central Algonquian time" (Bloomfield, 1925: clusive in the when followed plural conjunct order; p. 144). the third marker the by person plural /waw/, resulting 2.3. A further type of palatalization is found in form is /ahkwaw/. . 1.3. In certain combinations which have to be when a word contains one of the diminutive listed insertion of connective does Usually, individually, /i/ suffixes all t's in that word not take The combinations which concern /es/, /esis/, etc., preceding place. are nitem horse': nicemisis us here are of or with or Where an replaced by c, e.g., 'my 'my /m/ /n/ /k/ /t/. little 'his blanket': 'his little intransitive stem in in- horse'; otakohp ocakohpis ending /n/ or a transitive mistahi 'lots': miscahis a lot.' For animate stem in is followed the blanket'; 'quite /m/ by third-person verbal see 5.82. Note that this suffix the form is examples palatali- /k/, resulting /hk/ (5.73, 5.74). zation does not t/t/ and in the combination of the animate distinguish t/6/, e.g., /aOemw-/ Similarly, transitive atimw- acimosis 'little stem 'see' with the suffix 'dog': dog.' /wapam-/ reciprocal /to/, to indicate diminution is the forms is 'see each Pervasive palatalization resulting /wapahto-/ other,' not restricted to stems with a diminutive suffix. etc. (6.437). Thus, for example, the stem wat- 'hole' (cf. A: 2.2) in the diminutive wac- in waca 'little 2. PALATALIZATION (MUTATION) appears shape holes (Op)'; consider also yotin 'it is windy': yocin 'it t/6/ alternates with s, and t/t/ alternates with c, is a little windy.' Palatalization of t throughout before i/i/, i, and y. This is the major type of entire sentences or even speeches "makes them sound palatalization or MUTATION.88 pitiful" or overly sweet and effeminate. It is charac- 2.1. /0/ is replaced by /s/ before /i,i,y/; elsewhere, teristic of the culture hero Wisahkecahk to occasion- e.g. before /e/, /6/ remains. ally speak this way. Thus, when the stem /na0-/ 'fetch' occurs before 2.4. Derivation, finally, presents many instances of the ending /in/ '2-1 independent,' the resulting form palatalization. Few of these are very productive, 88 and some are definitely archaic. A great deal of A fairly detailed and highly readable account of Algonquian palatalization is given by Piggott, 1971a, even though the evi- work remains to be done before a comprehensive dence does not always seem sufficient to support his conclusions. account can be written. VOL. 63, PT. 5, 1973] APPENDIX: MORPHOPHONOLOGY 81

Examples: tahkopit- TA 'tie him fast': tahkopiso- AI vowel of the sequence, if not already long, is length- 'be tied fast' (cf. 6.439); pitoteyiht- TI 'think it ened, and the semivowel and the /e/ disappear. strange': pZtosisi- AI 'be a stranger'; mitdtaht 'ten': In the formula /Vw-e/, the following values are mitatasi- AI 'be ten.' attested for /V/: /i, a, i, e, a, 6/. Examples: kise- The non-initial alternant -dpam- TA 'see him' yiniw 'old man,' /epan/ 'former, absent': kiseyinipan provides a particularly interesting example. Not 'old man no longer alive'; mahkesiw 'fox', diminutive only are there both palatalized and unpalatalized /es/: mahkesis 'little fox, '; pesiw- 'bring him,' initials, e.g., pakisdpamew 'he lets him out of sight' /ehkok/ (2p-3p imperative): pesihkok 'bring them!' (cf. pakitinam 'he sets it down') vs. kitdpamew 'he nehiyaw 'Cree Indian,' vocative plural suffix /etik/: looks at him.' There are also clear chronological nehiyatik 'oh you Cree'; wihtamaw- 'tell it to him,' 1-2 sequences; for example, consider the archaic kosd- ending /etin/: kiwihtamdtin 'I tell it to you'; kikinaw pahtam 'he conjures' (root kot- 'try,' e.g., kotdpacihtaw 'our (21) house,' locative suffix /ehk/: kzkindhk 'in 'he tries using it') and kocihtdw 'he tries it' which our house.' sdsvw 'Sarci Indian,' distributive locative shows that the original /0/ has been re-interpreted suffix /enahk/: sdsnadhk 'in the land of the Sarci, at as /t/. Sarci Reserve.' ndpew 'man,' possessed theme suffix /em/: ninapem 'my husband'; wicew- 'have him 3. VOWEL COMBINATIONS along,' 1-2 ending /etin/: kiwicetin 'I have you along.' 'woman': 3.1. When two long vowels come together, /y/ is moniydw 'Canadian', /eskwew/ moniyd- skwew 'Canadian diminutive inserted between them, e.g., wayawi- 'outside' and woman'; ispatinaw 'hill', 'little locative -dmo- 'run, flee': wayawiydmow 'he runs outside'; /es/: ispacinas hill'; /ehk/: ispatinahk 'on a hill.' manitow theme kisikd- 'be day,' -dpan- 'be dawn': kisikdydpan 'it is 'spirit, god', possessed suffix nimanitom askJw- day-break.' (Cf. also 5.453.) /em/: 'my god'; 'follow him,' suffix askotowak follow 3.2. Before or after a long vowel, a short vowel reciprocal /eto/: 'they each other.' disappears, e.g., pimohte- 'walk,' /eyiwah/ (AI 3' In the formula the values are ending): pimohteyiwa 'he (3') walks'; pzhtokwe- /Vy-e/, only /i, a, o/ attested for 'inside,' -akocin- 'fly, hang': pihtokwekocin 'he comes /V/. Examples: pimiy 'lard,' possessed theme suffix lard.' flying inside'; nmpd- 'in the dark,' -ohte- 'walk': /em/: nipimam 'my natay 'my locative suffix nipdhtew 'he walks in the dark'; kask- 'break,' -ihcik- belly,' /ehk/: natdhk 'on my belly'; 'his locative suffix wane- 'knee,' -ahw- 'by tool, by shot': kaskihcik- wanoway check,' /ehk/: wanowdhk; diminutive kikwds wanehwew 'he breaks his knee by shot.' kikway 'something,' /es/: 'some- little.' 3.3. In a of short the second thing nisoy 'my tail,' diminutive /es/: nisos sequence vowels, little tail.' disappears; this situation seems to occur with 'my only Contraction is one of the clues for the of suffix-initial /e/, e.g., ositiyi- 'his foot or feet,' identity (3') i as or the other is the alternation of t locative suffix /ehk/: ositiyihk 'on his foot or i/i/ i/e/; (3') with s and t with c contraction and feet'; nikamo- 'sing,' /eyiwah/ (AI 3' (A: 2); palatali- ending): zation to be exclusive.89 Contrast nikamoyiwa 'he (3') sings'. appear mutually the following forms with a stem ending in . . ./aw/: 3.4. Where a morpheme beginning in /L/ is kiwihtamdtin 'I told it to you' and kiwzhtamawin 'you preceded by a short vowel, that vowel is lengthened, told it to me'; the 1-2 suffix is /etin/, the 2-1 suffix, e.g., stem nipi- 'die' plus suffix /Lmakan/ (5.86): /in/. Similarly, with a stem ending in . . . nipzmakan 'it dies.' /w/: askok (ending /ekwa/) 'he (3') followed him 3.5. When a stem-final is followed it (3'),' /o/ by /w/, askowin (ending /in/) 'follow me!', etc. is thus, dcimo- 'narrate': dcimow 'he lengthened; There are, however, also a few exceptions to the narrates'; contrast nitdcimon 'I narrate.' contraction rule. Contraction does not take place if the stem is monosyllabic, e.g., -lw- 'wife' 4. CONTRACTION (de- pendent), /ki- -enaw/ 'our (21)': kiwinaw 'our wife.' 4.1. Interconsonantal /w-i/ or /w-e/ are replaced TA stem miy- 'give him or it to him'; (3')-3 suffix by /o/, e.g., pakamahw- 'strike him,' /in/ (TA 2-1): /ekwa/: miyik. TA stem mow- 'eat him,' same kipakamahon 'you strike me'; the same stem com- suffix: mdwik.90 It seems reasonable to assume that bined with the 1-2 corresponding suffix, /etin/: exceptions of this type occur wherever the contraction kipakamahotin 'I strike you.' would excessively distort the stem; cf. also 5.813. In some instances, interconsonantal /w-i/ or /w-e/ There are several instances of disyllabic stems where may remain as a matter of surface variation, e.g. pahkekinw- 'hide,' /eyi/ (obviative possessor), /ehk/ 89 For an analysis based on a somewhat different interpretation (locative): T58-11 pahkekinwiyihk 'on his (3') hide.' of these phenomena, see Kaye, 1971a, and Piggott, 1971a; cf. 4.2. When a in a of also section 6.436, fn. 83. morpheme ending sequence 90As might be expected, there is some fluctuation which may vowel and semivowel is followed by another beginning also reflect the difficulty of distinguishing some of these forms, in /e/, contraction takes place. That is, the first e.g., mowikwak and mokwak 'he eats them (TA (3')-3p).' 82 WOLFART: PLAINS CREE [TRANS. AMER. PHIL. SOC. contracted and uncontracted forms occur side by side, kwaydc, kwaydci 'ready,' ndspic, naspici 'for good, e.g., wiyaw- 'his body,' locative suffix /ehk/: wiydhk, beyond return,' etc. wiyawihk 'on his body'; watay 'his belly,' locative 5.2. Postconsonantal word-final /w/ is lost, e.g., suffix /ehk/: watdhk, watayihk 'in his belly.' the 21-suffix of the conjunct order /ahkw/ in The full range of applicability of the contraction e-apiyahk 'when we were sitting.' rules remains to be determined. For Proto Algon- This final /w/ may arise from loss of final vowel, quian, Bloomfield stated them to be restricted to e.g., /a0emwa/ 'dog': atim; contrast the plural form instances where the second morpheme is an inflectional atimwak 'dogs.' ending (1946: p. 92). It also seems to hold for Cree 5.3. Otherwise, i.e., when the rules of 5.1 and 5.2 that contraction is mostly absent when morphemes have been applied, the morphophonological symbols are joined in derivation. Consider these examples: are phonemically manifested as follows: /0/ is realized kaw- 'prostrate,' /en/ 'by hand': kawinew 'he pros- as t, /e/ as i, and /p, t, c, k, . ./ as p, t, c, k, . . . trates him by hand'; particle kahkiyaw 'all,' animate intransitive abstract final /esi/ (6.431): kahkiyawisiw 6. PREFIXATION 'he is 'have him entire'; wzcew- along,' /eskwewe/: In general, when the personal prefixes ki-, ni-, o-, his wife etc. wzcewiskewew'he has along,' and mi- occur before a stem-initial vowel, the normal contraction is also found where However, widely manifestations of vowel combination (A: 3) do not are in and thus the morphemes joined derivation, appear; instead, /t/ is inserted. Thus, nitapin 'I in need of further situation remains investigation. sit,' etc. A few be added to those examples may already given: Instead of this /t/, the insertion of /h/ or /w/ has okimaw 'woman': okimaskwew 'chief,' /eskwew/ been observed in isolated instances, e.g., nihaydn 'I 'queen'; okimdw 'chief,' /ehkan/ 'substitute' (6.417): have it' or kiwdtoten 'you tell it,' etc. okimdhkdn 'elected or chief.' Consider appointed 6.1. Before certain dependent stems (3.2) which also the multitude of instances which a few were (of begin in a vowel, the prefixes have the alternants k-, which are the combination already cited) provided by n-, w-, and m-; katay 'your belly,' natay 'my belly,' in . with the diminutive suffix of noun stems ../Vw/ watay 'his belly'; kZki 'your dwelling,' niki 'my different discussed in /es(is)/. (A analysis, 3.32, dwelling,' wiki 'his dwelling,' miki 'a dwelling.' would the final of these nouns as inflec- regard /w/ Before dependent stems beginning in /o/ or /o/, even the above would then tional; though examples the third person prefix disappears, e.g., nohkom 'my fall in the domain of A: there is evidence 3.2, ample grandmother,' ohkoma 'his grandmother (3').' instances referred to For in the verbal below.) 6.2. Before stems with initial /o/ or /6/ there is a which involve the final suffix examples (derivational) great deal of what seems to be free variation. When of verbs see 6.43. As a representative example, the regular pattern (cf. above) appears and /t/ is stems in . . when followed consider here only ./aw/ inserted, /o/ is lengthened, e.g., okimdw 'chief': the reflexive or suffixes by secondary reciprocal /eso/ nitokimdmindn 'our chief'; ospwakan 'pipe': otospwa- or wthtamaw- 'tell it to him': /eto/ (6.437, 6.438): kana 'his pipe (3'),' etc. (This pattern does not 'he tells it to himself' and wihtamdtowak wihtamdsow normally occur with dependent nouns.) tell it to each other.' 'they More often, however, no /t/ is inserted and the prefixes are directly followed by /6/ (short /o/ is 5. PHONEMIC REALIZATION lengthened); thus, okimdw 'chief': ndkimdmindn 'our Certain word-final sounds do not appear in the chief'; ohtapiwin 'seat': k5htapiwin 'your seat'; phonemic realization of our morphophonological otinew 'he takes him': notindw 'I take him'; preverb o forms. 'past (in negative statement)': T113-6 moyihkdc no- 5.1. Final short vowels are subject to apocope, e.g. -wapahten. 'I had never seen it.' With the third proximate singular animate /sisip-a/: sisip 'duck.' person prefix o-, only dependent nouns show this In words whose stem is monosyllabic, the final pattern, e.g. ohkomiwdwa 'their grandmother.' vowel remains, e.g., /nisk-a/: niska 'goose'; /waw-i/: wdwi 'egg.' The final vowel is dropped sometimes 7. INITIAL CHANGE is (but not always, cf. wdwi) if the stem vowel long, INITIAL CHANGE is a systematic alternation of the e.g., /na0-/ 'bring him,' 2-3 imperative ending /i/: first vowel of a stem or compound (6.5); it occurs in nds 'bring him!'; cf. the same ending with the stem the changed and iterative modes of the conjunct /eO-/' say so to him': isi 'tell him so!' (Although order (5.33).91 incomplete in Cree, this distinction between stems i_e: itdpiw 'he looks thither or thus': T45p6 with long and short vowels reflects the Proto Algon- etPpihki 'wherever one (indf) may look'; pimohtew 1946: quian situation; cf. Bloomfield, p. 93.) 91Its function in word formationis yet to be investigated syste- With regard to the final /i/ of particles (6.421), matically; as an example, consider the reduplicative formation the application of this rule seems to be optional, e.g., mamwesahkiw'he goes naked' from mosawaham'he bares it.' VOL. 63, PT. 5, 1973] APPENDIX: SAMPLE TEXT 83

'he walks along': T55p87 pemohteyahk 'as we walked "awin ita ki-osapamikoyahk ?" itik. (10) "a, namoya, along.' awiyak k6sapamikonaw." a-e: takohtew 'he arrives walking': T61p13 tekoh- (11) mak esa witimwa eyakw awa napew awa, e- tecik 'when they arrived'; apiw 'he sits': S22-46 -m6siskweweyit, mak es e-ocawasimisiyit e-tahkopi- epiyani 'whenever I sit down.' tawasoyit. (12) kitahtaw etok 6ma, ekota e-ayayacik o-we: otawasimisiw 'he has children': wetawdsi- 6ma, e-wapamat esa man eyakw awa. (13) kitahtaw misicik 'those who have them as children'; ohtinam etokwe e-mamatot eyakw awa awasis awa. (14) aw 'he takes it thence': S244-1 wehtinahk 'taking it'; etokwe peyak awa, iskwew awa, osimisa 6hi, "kika- postaskisinahew 'he makes him put on moccasins': -kakwe-pisiskeyima ana man eyako ka-tatwewitahk," S126-12 pwestaskisinahdt 'putting on moccasins.' itew etokwe. (15) "yaw," k-etwet es aw iskwew. i-a, iyi. In Plains Cree both d and iyz occur but a (16) at an 6htawiya e-wicewat, "ta-ki-papisiskeyi- is attested only in the preverbs ki-kd and wv-wd mikot, min e-kiskeyihtamiyit," k-etwet es awa. (6.52); cf. also Geary, 1945: p. 171 and Bloomfield, (17) pehtawew awa kiseyiniw tanis e-itweyit ota- 1946: p. 101. For James Bay Cree, by contrast, Ellis nisa. (18) "eha," k-etwet esa. (19) "kispin tapwe, reports only the a-variant. Examples: wi-kapesiw wicewew ohtawiya n6sisimis. (20) nam6ya wahyaw 'he is going to camp': T103p8 wd-kapesici 'whenever ekot astew, ayahciyiniwak ta-nipahicik" k-etwet esa. he was going to camp'; kisi-nikamow 'he completes (21) ekosi. (22) nam6ya mihcet tipiskaw, es ana his singing': P14-24 kiyisZ-nikamotwdwi 'whenever ayahciyiniwa ki-nipahik eyakw ana kiseyiniw. (23) they have sung their song'; miciw 'he eats it': S244-20 eyak om 6ta namoya wahyaw, "Battle Lake" isi- miyicitwdwi 'whenever they had it to eat.' yihkatew, ekot es ana e-ki-nipahiht eyako kiseyiniw. e-iye: tehtapiw 'he rides (on horseback)': T123-5 (24) "George Maskwa" esa ki-isiyihkas6w. (25) tiyehtapit 'as he rode'; preverb pe 'hither': T105p8 ekota es ani kekac ki-mescihawak min ekonik aniki. piye-sdkeweydni 'when I rise (sun speaking),' T120-4 (26) ekonik mana nohkom, ekosi e-ki-itatotahkik. (27) piye-nipdtwdwi 'whenever they slept on the way.' ekosi. adiya: apasapiw 'he looks back': T80p8 iydpasapici TRANSLATION 'whenever he looked back'; wdpahtowak 'they see each other': T54pl4 wiydpahtoydhki 'whenever we see (1) I too will narrate some more. (2) But I don't each other.' know how many years it has been since that which I am oJiyo: totam 'he does it so': T53p6 tiyotahkwdwi going to tell about. ... 'whenever they did this,' S58-44 tdnisi tiyotahk? (3) Here at Hobbema some place these people must 'How did she do it?'; notinito- AI 'fight one another': have lived. (4) A certain Stoney Indian, his children, P60-39 niyotinitohki 'whenever there is a battle.'92 his wife, and his mother-in-law; and also three men [apparently his sons-in-law]. (5) He must have B: SAMPLE TEXT made a trip this way, and in a steep valley they The text which is here presented was narrated by traveled one behind the other, along this Battle River. CL in February, 1968. It is a kayds-dcimowin, a (6) Thus, I guess, presently this one who was at the historical narrative. For a brief characterization of end, this man, saw someone ducking up and down. the informants for whom CL is representative, (7) So it was. (8) He told it to his wife, "Someone see 1.3. is watching us." (9) "Who would be able to watch us there?" she answered him. (10) "Oh, no, someone TEXT: AN ENCOUNTERAT BATTLELAKE is watching us." (1) eyapic nista nik-acimon. (2) mak eyako (11) But this man's sister-in-law, she was single, namoya nikiskeyihten, taniyikohk oma ta-ki-ispihc- but she had a baby which she had tied in a moss-bag. -askiwik aspin eyak oma ka-wi-atotaman.... (12) Presently, I guess, they were there where this (3) 6t etok oma e-ayayacik, maskwacisihk oki one had seen him [someone]. (13) Suddenly this peyakwayak ayisiyiniwak. (4) asinipwatis etokwe baby screamed. (14) This one, this woman, told peyak, otawasimisa, owikimakana, osikosa; ekwa her younger sister, "You should have tried to look nisto mina napewa. (5) etokwe ka-pe-piciwinihket, after this one who makes the noise," she must have ekwa posiskahcaw 6ma nanaway pimipiciw oma, told her. (15) "Why," said this woman. (16) But notinito-sipiy. since she had her father along, "he should have taken (6) ekos isi etokwe, kitahtaw etokw awa peyak, care of him [the baby], and he knows it," this one naway e-ayat awa, napew awa, ka-wapamat awiya said. e-osiskwepayihoyit. (7) eyakosi. (8) wihtamawew (17) The old man heard his daughter how she esa owikimakana, "awiyak k6sapamikonaw." (9) spoke. (18) "Well," he said. (19) "If this were true, my grandchild will follow his father [i.e., die]. 92For James Bay Cree, Ellis reports 5 ~ wd, e.g. posiw 'he embarks': pwdsit 'as he embarks'; notamesw 'he is fishing': (20) Not far from here, there is the place where the nwdtameset 'as he is fishing.' Blackfoot will kill me," he said. 84 WOLFART: PLAINS CREE [TRANS. AMER. PHIL. SOC.

(21) So it was. (22) Not many nights [passed], transcription), and personal prefixes, are separated and this old man was killed by the Blackfoot. (23) from the stem by =. The person-number-gender- Not far from this here, Battle Lake it is called, there obviation codes given for inflected forms also serve this old man was killed. (24) George Maskwa he was to indicate cross-reference, especially with the numer- called. (25) Right there these [other people] were ous demonstrative pronouns. also almost all killed. (26) This is how my grand- The frequent occurrence of the demonstrative mother's people always used to tell it. (27) This is it. pronouns, e.g. awa, oma, eyako, and especially of such particles as etokwe and esa, is characteristic of narra- ANALYSIS tive style; the full meaning of esa and mdna is not Since numerous examples of complex word forma- known. tion are given in chapter 6, the present interlinear The sequence of spans (2.2) and the focus assign- analysis is restricted to identifying the word class ment within each span are summarized separately. and the inflectional form of each word. In nouns (The phrases listed under "proximate" are the first and verbs, endings (which appear in a phonemic or clearest indication of who is in focus.)

SENTENCE SPAN PROXIMATE (FOCUS) OBVIATIVE 1, 2 a (2) eyak oma ka-wi-atotaman 'that which I am going to tell about' 3-5 b (3) oki ayisiyiniwak 'these people,' then (4) asinipwatis 'the Stoney' (4) otawdsimisa 'his child- ren,' etc., including ndpewa 'men' 6-12 c (6) ndway e-aydt awa, ndpew awa 'this man who was at (6) awiya 'someone' the end' (8) owikimdkana 'his wife, (d) (direct speech is inserted: (9) awiyak 'someone' and (10) awina 'who') c (8) itik 'she answered him' shows that the man is still in focus (11) wztimwa 'his sister- in-law' (12) e-ayaydcik 'they were there,' then eyakw awa (12) goal of e-wdpamdt 'he 'this one' had seen him' 13 e awasis awa 'this baby' 14 f iskwew awa 'this woman' osimisa 'her younger sister' (g) (direct speech: ana eyako ka-tatwewitahk 'that one who makes the noise') f itew 'she told her' 15-16 h (15) aw iskwew 'this woman' (the younger sister) (15) ohtdwiya 'her father' (i) (direct speech, but note that the father remains obviative: the baby is the goal of (16) ta-ki-papisiskeyimikot 'he should have taken care of him') ((16) e-kiskeyihtamiyit 'he knows it') h (16) k-etwet es awa 'this one said' 17-25 j (17) awa kiseyiniw 'this old man' (17) otdnisa 'his daughter' (k) (direct speech: (18) nosisimis 'my grandchild') ((18) ohtawiya 'his father') (1) ((19) ayahciyiniwak 'the BIackfoot') j (20) k-etwet 'he said,' then (25) ekonik aniki 'these (22) ayahciyiniwa 'the (people)' Blackfoot' 26 m ekonik nohkom 'those (including) my grandmother' VOL. 63,:PT. 5, 19731 APPENDIX: SAMPLE TEXT 85

In the interlinear analysis which follows, these special abbreviations occur: NA, NI animate, inanimate noun PV preverb NDA, NDI animate, inanimate dependent noun red reduplication syllable PR pronoun indep independent indicative PC particle cj changed conjunct PP personal prefix simple cj simple conjunct For the symbols TA, TI, AI, II see 1.41, and for the number codes of the person-number-gender-obviation categories see table 1 of 2.01. In the interlinear version of the text, the modifications of external sandhi have been removed. (1) eyapic nista ni = ka-acimo = n. PC personal PR, emphatic PV ka 'future,' AI 1 indep in due course I too I will narrate

(2) maka eyako namoya ni = kiskeyiht = en, PC 0PR PC TI 1 indep but just this not I know it taniyikohk oma ta-ki-ispihci-askiwi = k PC (conjunction) 0PR PV ta 'future,' PV khl 'completion,' to what extent this it will have been years since aspin eyako oma PV ispihci 'meanwhile,' II 0 simple cj PC 0PR 0PR away just this this kd-wi-atot = aman .. PV kd (subordinator, 6.521), PV wi 'intend to,' TI 1 cj what I am going to tell

(3) ota etokwe oma e-ayaya =cik, PC PC 0PR PV e (subordinator, 6.521), AI 3p cj, red ay- here I guess this they were, they lived = ihk oki peyakwayak locative NI 3p PR PC at Hobbema ("Bear Hills") these at one place ayZsiyiniw =ak. (4) asinipwatis etokwe peyak, 3p NA 3 NA PC PC people Stoney () I guess a certain ot = awdsimis = aL, o = wikimakan = a, o = sikos = a; ekwa nisto 3' NA (3 poss) 3' NA (3 poss) 3' NDA (3 poss) PC PC his children his wife his mother-in-law and three

mina napw = a. (5) etokwe ka-pe-piciwinihke = t, PC 3'NA PC PV ka, PV pe 'hither,' AI 3 cj also men I guess he made a trip hither ekwa posiskahcdw oma nanaway PC 0NI 0PR PC, red nd- and long, narrow valley this one behind the other pimipici = w oma, notinito-sipiy. AI 3 indep OPR 0NI they traveled (general singular) this Battle River

(6) ekosi isi etokwe, kitahtawe etokwe awa PC PC PC PC PC 3 PR just thus thus I guess presently I guess this 86 WOLFART: PLAINS CREE [TRANS. AMER. PHIL. SOC.

peyak, ndway e-aya = t awa, ndpew awa, PC PC PV e, AI 3 cj 3 PR 3 NA 3 PR one behind he was there this man this

ka-wapam = at awiya e-osiskwepayiho= yit. (7) eyakosi. PV ka, TA 3-(3') cj 3' PR PV e, AI 3' cj PC he saw (him) someone he ducked with his head up just thus (8) wihtamaw = ew esa o = wkimdkan = a, "awiyak TA 3-(3') indep PC 3' NA (3 poss) 3 PR he told it to her then his wife someone

k = osapam = ikonaw. ((9) "awina ita PP ki-, TA 3-21 indep, -osdpam- 3 PR PC he is watching us who there kl-osapam = ikoyahk," it = ik. PV ki2 'able to,' TA 3-21 simple cj TA (3')-3 indep he would be able to watch us she answered him

(10) "a, namoya, awiyak k = osapam = ikonaw." excl. PC (see ab)ove) Oh no

(11) maka esa witimw:-a eyako awa PC PC 3' NDAi (3 poss) 3 PR 3 PR but then his cros s-cousin just this this napew awa, e-mosiskwewe= yit, maka esa 3 NA 3 PR PV e, AI 3' cj PC PC man this she was single but then e-ocawasimisi = yit e-tahkopitdwaso= yit. PV e, AI 3' cj PV e, AI 3' cj she had a child she had it tied up in a moss-tie (12) kitahtawe etokwe oma, ekota e-ayay = cik PC PC 0PR PC PV e, AI 3p cj, red ay- presently I guess this just there they were there oma, e-wapam =dt esa mana eyako awa. 0 PR PV e, TA 3-(3') cj PC PC 3 PR 3 PR this where he had seen him then then just this this (13) kitahtawe etokwe e-mamdto = t eyako awa PC PC PV , AI 3 cj, red ma- 3 PR 3 PR presently I guess he cried loud just this this awasis awa. (14) awa etokwe peyak awa, i skwew awa, 3 NA 3 PR 3 PR PC PC 3 PR 3; NA 3 PR child this this I guess one this v,volan this o = simis = a ohi, "ki = ka-kakwe-pisiskeyim= i 3' NDA (3 poss) 3' PR PP ki-, PV ka 'future,' PV kakwe 'try her younger sister this you should have taken care of him ana mana eyako to,' TA 2-3 indep h-preterit 3 PR PC 3 PR that always (?) just that kd-tatwewit = ahk," it = w etokwe. (15) "yaw" PV ka, TI 3 cj TA 3-(3') indep PC exclamation who makes noise she told her I guess why VOL. 63, PT. 5, 1973] APPENDIX: SAMPLE TEXT 87

kd-itwe = t esa awa iskwew. (16) dta ana PV ka, AI 3 cj PC 3 PR 3 NA PC 3 PR she said then this woman however that o = htdwiy = a e-wicew = at, "ta-kz-papisiskeyim = ikot, 3' NDA (3 poss) PV e, TA 3-(3') cj PV ta 'future,' PV kh her father she had him along he should have mmna 'completion,' TA (3')-3 simple cj, red pa- PC taken care of him and e-kiskeyiht = amiyit," kd-itwe = t esa awa. PV e, TI 3' cj PV ka, AI 3 cj PC 3 PR he knows it she said then this

(17) pehtaw = ew awa kiseyiniw tdnisi TA 3-(3') indep 3 PR 3 NA PC (conjunction) he heard her this old man how

e-itwe =yit o = tanis = a. (18) "eha," kd-itwe = t esa. PV e, AI 3' cj 3' NDA (3 poss) PC PV ka, AI 3 cj PC she spoke his daughter yes he said then

(19) "kispin tapwe, wzcew = ew o = htawiy=a PC PC TA 3-(3') indep 3' NDA (3 poss) if truly, indeed he accompanies him his father nsisimis. (20) namoya wahyaw ekota ast =w, 3 NDA PC PC PC II 0 indep my grandchild not far just there it is there ayahciyiniw = ak ta-nipah = icik," kd-itwe = t esa. 3p NA PV ta 'future,' TA 3p-1 simple cj PV ka, AI 3 cj PC the Blackfoot they will kill me he said then (21) ekosi. (22) namoya mihcet tipiskdw, PC PC PC ONI just thus not many nights (general singular) esa ana ayahciyiniw =a ki-nipah =ik PC 3 PR 3' NA PV kHl'completion,' TA (3')-3 indep then that the Blackfoot they killed him eyako ana kiseyiniw. (23) eyako oma ota namoya 3 PR 3 PR 3 NA O PR OPR PC PC just this that old man just this this here not wahyaw, "Battle Lake" isiyihkdat = w, ekota esa ana PC II 0 indep PC PC 3 PR far it is named so just there then that e-kZ-nipah= iht eyako kiseyiniw. PV e, PV ki 'completion,' TA indf-3 cj 3 PR 3 NA he was killed just this old man (24) "George Maskwa" esa ki-isiylhkdso = w. PC PV ki1 'completion,' AI 3 indep ("Bear") then he was called so

(25) ekota esa ani kekac PC PC PC PC just there then (emphasizes preceding words) almost 88 WOLFART: PLAINS CREE [TRANS. AMER. PHIL. SOC. ki-mescih = dwak mina ekonik aniki. PV kil 'completion,' TA indf-3p indep PC 3p PR 3p PR they were annihilated also just these those (26) ekonik mdna n = ohkonm kosi e-kl-itatot = ahkik. 3p PR PC 3 NDA PC PV e, PV ki 'completion,' just these always (?) my grandmother just thus (27) ekosi. TI 3p cj PC they used to thus tell it just thus REFERENCES CITED

BLOOMFIELD,LEONARD. 1924, 1927. "Notes on the Fox Lan- Lane (University of North Carolina Studies in the Germanic guage." International Journal of American Linguistics 3: pp. Languages and Literatures, 58, Chapel Hill), pp. 137-145. 219-232; 4: pp. 181-219. HANZELI, VICTOR E. 1961. "Early Descriptions by French . 1925. "On the Sound System of Central Algonquian." Missionaries of Algonquian and : a Study Language 1: pp. 130-156. of 17th- and 18th-century Practice in Linguistics "(unpublished . 1927. "The Word-Stems of Central Algonquian." dissertation, Indiana University). Festschrift Meinhof (Hamburg), pp. 393-402. . 1969. Missionary Linguistics in New France: a Study . 1928. "The Plains ." International of Seventeenth- and Eighteenth-Century descriptions of American Congress of Americanists, Proceedings 22, 2: pp. 427-431. Indian Languages. (The Hague). A revision of Hanzeli, 1961 (Rome) which omits some parts which are relevant to Cree. . 1930. Sacred Stories the Sweet Grass Cree of (National HISEY, HELEN, and CHARLESE. FIERO. 1969. "A Report on Museum of Bulletin Canada, 60, Ottawa). Coastal Montagnais-Naskapi" (MS). 1933. Language (New York). HOCKETT, CHARLES F. 1948. "Potawatomi." International 1934. Plains Cree Texts (Publications of the American Journal of American Linguistics 14: pp. 1-10, 63-73, 139-149, Ethnological Society, 16, New York). 213-225. . 1946. "Algonquian." Harry Hoijer et al., Linguistic .1958. A Course in Modern Linguistics (New York). Structures of Native America (Viking Fund Publications in . 1966. "What is Iike." International Anthropology, 6, New York), pp. 85-129. Algonquian Really Journal American 32: 59-73. .1958. Eastern Ojibwa: Grammatical Sketch, Texts, and of Linguistics pp. Word Lists (ed. Charles F. Hockett, Ann Arbor, Michigan). HOWSE, JAMES. 1844. A Grammar of the Cree Language . 1962. The Menomini Language (ed. Charles F. Hockett, (London). New Haven, Connecticut). HUNTER, JAMES. 1875. A Lecture on the Grammatical Con- MS(a). Cree Texts (Syllabary texts, item #785, Free- struction of the Cree Language (London). man-Smith, American Philosophical Society Library, Phila- JACOBSEN, WILLIAM H., JR. 1967. "Switch-Reference in delphia). Hokan-Coahuiltecan." Dell H. Hymes, ed., Studies in South- . MS(b). Cree-English Lexicon (MS). western Ethnolinguistics (The Hague), pp. 238-263. CANADA, Indian Affairs Branch. 1970. Linguistic and Cultural JONES, DAVID J. 1971. "Odawa Noun Morphology." J. D. Affiliations of Canadian Indian Bands (Ottawa). Kaye, et al., ed., Odawa Language Project, First Report (Anthro- CHAFE, WALLACEL. 1962. "Estimates Regarding the Present pological Series, 9, University of Toronto), pp. 39-79. Indian Interna- Speakers of North American Languages." JONES, WILLIAM. 1911. "Algonquian (Fox)." (Revised by tional Journal of American Linguistics 28: pp. 162-171. Truman Michelson.) Franz Boas, ed., Handbook of American Course EDWARDS,Mary. 1954. Cree, An Intensive Language Indian Languages (vol. 1, Bureau of American Ethnology, (Meadow Lake, Saskatchewan). Bulletin 40, 1, Washington), pp. 735-873. ELLIS, C. DOUGLAS. 1960. "A Note on okimahkan." Anthro- KAYE, JONATHAN D. 1971a. "A Case of Local Ordering in 2, 3: 1. pological Linguistics p. Ojibwa." J. D. Kaye, et al., ed., Odawa Language Project, . 1961. "The So-Called Order in Cree." Interrogative First Report (Anthropological Series, 9, University of Toronto), International Journal American Linguistics 27: 119-124. of pp. pp. 3-10. . 1962. Spoken Cree, TVest Coast of James Bay (Part I, .1971b. "Selectional Restrictions and the Algonquian Toronto). Animate-Inanimate Classification." J. D. Kaye, et al., ed., .1971. "Cree Verb Paradigms." International Journal Odawa Language Project, First Report (Anthropological Series, of American Linguistics 37: pp. 76-95. 9, University of Toronto), 80-91. FABVRE, BONAVENTURE. 1970. Racines Montagnaises (1696) pp. KURYLOWICZ, 1964. The of Indo- (ed. Gerard E. McNulty, Travaux divers, 29, Centre d'Etudes JERZY. Inflectional Categories Nordiques, Universite Laval, Quebec). European (Heidelberg). FARIES, RICHARD. 1938. A Dictionary of the Cree Language LACOMBE, ALBERT. 1874a. Dictionnaire de la langue des Cris (Toronto). (Montreal). FRANTZ, DONALD G. 1966. "Person Indexing in Blackfoot." .1874b. Grammaire de la langue des Cris (Montreal). International Journal of American Linguistics 32: pp. 50-58. LAMB, SYDNEY M. 1966. "Epilegomena to a Theory of Lan- GEARY, JAMES A. 1943. "The Proto-Algonquian Form for guage." Romance Philology 19: pp. 531-573. 'I-thee'." Language 19: pp. 147-151. LEHMANN,WINFRED P. 1958. "On Earlier Stages of the Indo- .1945. "The Changed Conjunct Verb (without ni) in European Nominal Inflection." Language 34: pp. 179-202. Fox." International Journal of American Linguistics 11: pp. LEMOINE, GEORGES. 1901. Dictionnaire Francais-Montagnais 169-181. et Grammaire Montagnaise (Boston). HENRY 1961. An Introduction to GLEASON, A., JR. Descriptive LESSARD,JEAN. MS. "M6thode de la langue Crise." Linguistics (2nd ed., New York). LONGACRE, ROBERT E. 1957. "Quality and Quantity in Cree GODDARD,IVES. 1967. "The Algonquian Independent Indica- Vowels." Journal of the Canadian Linguistic Association 3: tive." National Museum of Canada, Bulletin 214: pp. 66-106. pp. 66-70. GODDARD,PLINY EARLE. 1911. "Athapascan (Hupa)." Franz LYONS, 1968. Introduction to Theoretical Linguistics Boas, ed., Handbook of American Indian Languages (vol. 1, JOHN. Bureau of American Ethnology, Bulletin 40, 1, Washington), (Cambridge). MARGUERITE. 1971. "The Eastern pp. 91-158. MCKENZIE, (Mistassini) Cree Verb: Derivational M.A. HAAS, MARY R. 1967a. "Roger Williams's Sound Shift: a Morphology" (unpublished McGill Study in Algonkian." To Honor Roman Jakobson (The thesis, University). Hague), pp. 816-832. MCNULTY, GERARD E. 1971. Petite grammaire du parler 1967b. "The Development of Proto-Algonkian *-awe-." Montagnais en ts-n (Collection Notes de Cours, 4, Centre Studies in Historical Linguistics in Honor of George Sherman d'Etudes Nordiques, Universite Laval, Quebec). 89 90 WOLFART: PLAINS CREE [TRANS. AMER. PHIL. SOC.

MANDELBAUM,DAVID G. 1940. The Plains Cree (American 1963. "Survey of Round Lake Ojibwa Phonology and Museum of Natural History, Anthropological Papers, 37.2). Morphology." National Museum of Canada,Bulletin 194: pp. MEILLET,ANTOINE. 1937. Introductiona l etude comparative 92-154. des langues indo-europeennes(8th ed., Paris). SCHMIDT, JOHANNES. 1889. Die Pluralbildungen der indo- MICHELSON, TRUMAN. 1912. "Preliminary Report on the germanischenNeutra (Weimar). Linguistic Classification of Algonquian Tribes." Bureau of SIEBERT,FRANK T. 1967a. "The OriginalHome of the Proto- AmericanEthnology, Annual Report28: pp. 221-290. Algonquian People." National Museum of Canada, Bulletin . 1926. "The Fundamental Principles of Algonquian 214: pp. 13-47. Languages." Journal of the WashingtonAcademy of Sciences 1967b. "Discrepant Consonant Clusters Ending in 16: pp. 369-371. *-k in Proto-Algonquian." National Museum of Canada, . 1939. "LinguisticClassification of Creeand Montagnais- Bulletin 214: pp. 48-59. Naskapi Dialects." Bureau of American Ethnology,Bulletin VOEGELIN,CARL F. 1946. "Delaware, an Eastern Algonquian 123: pp. 67-84. Language." Harry Hoijer et al., LinguisticStructures of Native PIGGOTT,GLYNE L. 1971a. "Some Implicationsof Algonquian America (Viking Fund Publications in Anthropology, 6, New Palatalization." J. D. Kaye, et al., ed., OdawaLanguage Pro- York), pp. 130-157. ject, First Report (Anthropological Series, 9, University of VOORHIS,PAUL H. 1967. "Kickapoo Grammar" (unpublished Toronto), pp. 11-38. dissertation, Yale University). . 1971b. "Odawa Intransitive Verb Morphology: Inflec- 1971. "New Notes on the Mesquakie (Fox) Language." tional Suffixes." J. D. Kaye, et al., ed., OdawaLanguage Pro- InternationalJournal of AmericanLinguistics 37: pp. 63-75. ject, First Report (Anthropological Series, 9, University of WOLFART,H. CHRISTOPH. 1971a. "Plains Cree Internal Syn- Toronto), pp. 92-106. tax and the Problem of Noun-Incorporation." International PITTMAN,RICHARD S. 1965. "The Fused Subject and Object Congress of Americanists, Proceedings 38, 3: pp. 511-518 Pronouns of Red Pheasant Cree." Linguistics 13: pp. 34-38. (Stuttgart). POSTAL, PAUL M. 1966. "A Note on 'Understood Transi- 1971b. "Boundary Maintenance in Algonquian: a Lin- tively'." International Journal of American Linguistics 32: guistic Study of Island Lake, Manitoba." American Anthro- pp. 90-93. pologist[in press]. ROGERS,JEAN H. 1960. "Notes on Mistassini Phonemics and WOLFART,H. CHRISTOPH,and FRANCISPARDO. 1972. "The Morphology." National Museum of Canada, Bulletin 167: Cree Project: Towards Automation in Linguistics" (Canada pp. 90-113. Council Research Report, 1, MS)

Top View