d AH3 PUBLICATIONS OF THE Cfc^T^SS^S | AMERICAN ETHNOLOGICAL SOCIETY Edited by Mabian W. Smith XXI GLADYS A. REIOHARD NAVAHO GRAMMAR J. J. AUGUSTIN PUBLISHER NEW YORK ALL RIGHTS RESERVED by Gladys A. Reiohard GIFT PRINTED IK GERMANY AT J.J. ATTGtrSTIN, GLUOKSTADT Est Us A«8 Preface A brief statement of the genesis of this grammar and an evaluation of the extant works on the Navaho language are outlined in the Introduction. The work is a byproduct of a project to investigate some of the inner meanings of Navaho religion, a study I felt could not be accomplished without some interpretation of the language. I therefore identified myself with a Navaho family, only two of whose members spoke English. I lived with this family for a total of more than three y% r s from 1930 to 1939. Later, in 1944, 1 spent a summer as a clerk v\ a trading post at Pinyon where hardly a word of English was heard because the Navaho who knew it were at war. By this time I knew the pattern of Navaho and had a large vocabulary; I used the opportunity to test the vernacular in practice. In the summer of 1934 I taught adult Navaho interpreters to write their language at the Hogan School which consisted of students and instructor only, the students being also the informants. The results of this experience, conducted under the auspices of the Department of Indian Affairs, are reflected in several ways: The students were made aware of the necessity of keeping forms to the same paradigms (which, incidentally, they loved). Since they came from many parts of the Reservation, the foundations were laid for the materials on the diversity of the language, and consequently, the reasons for similarities as well as differences and relationship of forms. Not the least of the results of the Hogan School was the interest aroused in many interpreters who have since been most helpful. Of these AB was the most analytic and his explanations have almost invariably checked with materials subsequently gathered. The following is a brief sketch of the interpreters who contributed most, and of their strength and weaknesses. Throughout the work I am writing initials for the source of forms and explanations. It should be remembered that many Navaho, whether educated in White schools or not, are tremendously interested in analyzing words and in constructing unusual and metaphoric utterances, which sometimes lead to punning. To this characteristic I attribute the explanation of their unusual ability to discern the composition and meaning of elements that have been thoroughly disguised by contraction. AB (Adolph Bitanny), orphaned in early childhood, lived with his maternal grandmother who, setting a high value on her language and M7101.83 ' VI PREFACE tradition, drilled her grandson in "classical" Navaho. They played games with forms and meanings, and the grandmother inculcated the interest in the language which so many years later became the key to my analysis. AB has a keen ear and is very particular about subtle differentiations for which he assigns reasons. His interpeta- tions almost always stand up according to old Navaho patterns, particularly the texts of old men ; he almost never gives a false or folk etymology. He is an ^-speaker, his texts and transcriptions being full of x J s and h's, not always accepted by others. The one weakness in his writings is his acceptance of some rules prematurely determined by his white instructors, especially his negligence in writing, though not in pronouncing, nasalized sounds and tone related to nasalization. He has some tendency also to reconstruct forms rather than to write the sounds influenced by contact, for instance, si'd for sa'q. The rules of assimilation and a great many other data serve as a check on this shortcoming. DD (Delia Degrote of Thoreau) gave me many expressions in the vernacular that later turned out to demonstrate diversity and to determine the differences between cessatives and repetitives. FH (Frank Harper of Klagito) has a large vocabulary in Navaho and English. His translations are thoughtful and usually reliable. His analyses sometimes leave much to be desired. He has contributed some interesting and amusing folk etymologies which are very illuminating. He was especially good at differentiating homonyms or near-homonyms. He does not have much practice in keeping paradigmatic forms in line, but he came into the picture after they had been worked out, and was used primarily for other purposes. His knowledge of Navaho tradition and lore, his love for the lan- guage, and his extensive experience on many parts of the Reserva- tion were of inestimable help. He is an x- and ^-speaker; his forms match almost invariably those of AB and HP's family which form the foundation of my analysis. JC (John Curley of Ganado) was primarily an interpreter of ethnology (except religion). He has a devious mind and I learned much from him because his reasoning was always indirect. In- direction is almost a Navaho (if not an Indian) tenet, and I consider it in many cases an infallible check on certain problems that defy formulation and direct questioning. MC (Marie Curley of Ganado) was my dependable mentor and guide. She is the late Red Point's daughter who taught me to weave and served as a buffer when I most needed it. She led me through my baby steps in Navaho. Like DD, she told me "how to say so and so." Neither she nor TC, her husband, tried analysis; both simply told me was what was what, and most important, corrected mistakes. RT (Ruby Tallman) lived at Red Point's for much of the time I was there. I have used her expressions primarily for comparison. PREFACE Vll TC (Tom Curley of Ganado) is MC's husband who took serious responsibility for my welfare and education in Navaho affairs. He direct^ m y travels, gave advice, and checked on terms I asked about w ith no attempt at giving reasons. If he did not know what I asked for, he found out and reported back. Wfy (William Morgan of Two Wells) has collaborated with Robert Youn* w hose work I have mentioned in 1. He is in the class of AB, j) U t his experience with old Navaho is not extensive. He has a a insatiable interest in extending his knowledge, understands analy 3 i s W el\ } and learns very rapidly. He does not differentiate in his mi n d some intricate overlapping forms, as the cessatives and repetitives, but the forms he writes naively without attempting to an a iy Ze are checks on these omissions. He, too, often neglects to wiite nasalization and some related tones, having been influ- enced by the same rules as AB. Otherwise his records are almost perfect. WM has saved me hours of time because of his understand- ing of the grammatical problems. We could go directly to the needed form instead of having to fish for it through interminable explanations and digressions. I am sure he would agree with me in attributing his value to the fine training given him by Robert XOUng g WM is next an ce speaker. Roman Hubbell is the only white man whose Navaho I have relied upon. I have abbreviated his name as RH for convenience, because he has to some extent served as an informant. He has a keen respect for evidence, and to him I owe certain insights on subtle questions which could never have been understood by a Navaho, particularly because the Navaho could not have under- stood the cultural aspects of the problem. The references to RH are few, but have proved extremely valuable, particularly as clues. To these persons and to many others with whom contacts may have been more brief, and also to the many medicine men whose instructions were often explained in Navaho, I acknowledge my indebtedness. I deeply appreciate the help of Professor George Herzog, now of Indiana University, who with unfailing patience taught me to distinguish Navaho tones. His greatest achievement was to make me aware of the glides in my own vowels, which are intolerable in Navaho. Despite the criticisms of 1.7-1.26., I feel deeply obligated to the late Edward Sapir who gave unsparingly of his time when I was first studying Navaho. The fact that I have come to different conclusions from his has no relation to his kindness and generosity. Harry Hoijer, and others of Sapir 's students at the University of Chicago, also helped greatly with their notes and discussions. I am grateful to Professors Alfred L. Kroeber and Andre Marinet of Columbia University for advice about the format of certain parts Vlll PBBFAOE of this work. The conclusions and final decisions are, however, my sole responsibility. For comfortable and stimulating circumstances under which the work was finally brought to a conclusion I wish to thank the Museum of Northern Arizona, Flagstaff, and all its staff. April, 1951 Gladys A. Reichard Barnard College, Columbia University Abbreviations If any apology is needed for what may seem to be an undue use of ab- breviations, it is because of the Navaho language itself. A reader will not get far in this description before he realizes that he must conceive of categories which may be new even to expert linguists. For convenience, certain symbols have been devised to save repetition. The following are abbreviations used in the phonology and grammar : ag. agent att. attitude C consonant C, glottalized consonant *C glottalized consonant preceded by glottal stop caus. causative cer. ceremony, ceremonial cess. cessative compl. completive cont. continuative cust. customary fut. future gen. general i some, someone, something I independent pronoun inc. inceptive mom. momentary, momentaneous myth. mythology, mythological O object of verb or postposition ob j . object opt. optative pass. passive pf. perfective pi. plural pres. present prog. progressive prol. prolongative rec. reciprocal rec. ef. reciprocal effect rep. repetitive rep. ac. repetitive action rep. asp. repetitive aspect spec. specific subj. subject s self t things (supernatural things) T type stem uni. uniform V vowel V* long vowel ^ vowel with high tone X ABBREVIATIONS V"* long vowel with high tone V nasalized vowel V high nasalized vowel V- long high nasalized vowel Numbers have been used to indicate the persons — 1, 2, 3, (3), 4 — all are explained in 10.64-10.73. Combinations of numbers, indicate a combination of subject-object pronominal prefix, for example, 3-3 "he moves it,*' "she hears him," etc. ; 3-i "he moves something;" i-3 "someone moves it;" (3) by i "it is moved by someone" (8.23 ff.). A number without a letter refers to the singular; preceded by D a number means the person in the dual ; preceded by P a number means the person in the plural. Abbreviations have also been used for textual material ; they refer to the initials of informants as described in the Preface, or to items of the Biblio- graphy : AB Adolph Bitanny (word of mouth), or Bitanny 1941. Ad Adah66niligfi (the first number refers to the year, the second to the month of publication, the last number to the page). BS Reichard, Big Star Chant (unpublished manuscript). DD Delia Degrote. EW llaile 1938 (Enemy Way).* FH Frank Harper. FS Young-Morgan 1948 (Function and Signification). FW Haile 1943 (Flintway).* HC Reichard 1944 (Hail Chant).* NT Sapir-Hoijer (Navaho Texts).* Ph Hoijer 1945c (Phonology). Pr Reichard 1943 (Prayer).* RH Roman Hubbell. SCE Reichard Shooting Chant Evil (unptiblished manuscript). WE Reichard, Chant of Waning Endurance (unpublished manuscript). WM William Morgan (personal communication). YM Young-Morgan, 1943, Navaho -English Dictionary. YME Young-Morgan, 1943, English-Navaho Dictionary. YMG Young-Morgan, 1943, Grammar. In references marked with an asterisk (*) the first number refers to the page, the second to the line. ALPHABETICAL ARRANGEMENT For lexical purposes the following arrangement is used : the short neutral vowel is placed first, next the long neutral vowel, next the high vowel, the high long vowel, the vowel with rising tone, the long vowel with rising tone, the vowel with falling tone, the long vowel with falling tone; the nasalized vowel, the nasalized long vowel, the high nasalized vowel, the long, high nasalized vowel, the nasalized vowel with rising tone, the long nasalized vowel with rising tone, the nasalized vowel with falling tone, the long na- salized vowel with falling tone. The vowels would thus appear in the following order, which is worked out for their occurrence in any position: a, a-, a, d«, &, &•, d, d' t q, q- t q, <£•, #, #•, q> §•, e, e-, i 9 4-, $, $•, e, e>, e, e-, £ f , & £•, (, £•, i, v, I, f , i, i*, i, £', % {•> {, i'> h V> i> t'> o, o*, <5, d-, <5, <5-, 6, 6; q, Q-, g, ?-, 0, (5-, $, 0-. The consonant arrangement is : -, h, 6, w, w, m, rrl, d, t, f, n t ri, y, y, g, k, #, Jew, y, x, yw, xw, z f s, dz, ts, ti, j, c, dj, ic, it, I, I, dl, tl, ti. Contents PREFACE V ABBREVIATIONS ix ALPHABETICAL ARRANGEMENT x 1-1.26. INTRODUCTION 1 1.7-1.26. The sapir school of athabaskan 4 2. GRAMMATICAL PROCESSES 12 3-3.140. PHONOLOGY 13 3.1. JiawELS 13 3.7. Syl labic n " 15 3.8. Con sonants 16 3.9-3.14. Glottalization . . 17 3.15-3.19. Aspiration and non -aspiration 19 3.20-3.21. Labialization 20 3.22-3.140. Assimilation 20 3.26-3.33. Assimilation due to mechanical change 22 3.34-3.38. M echanical change of tone 24 3.39. Gl ide consonants 26 3.40-3.44. Jlowel or consonant loss 26 3.45-3.133. _ Sou nd changes due to morphology 28 3.45-3.48. Stem and stem complex 28 3.49-3.53. Rel ationship between vowel and consonant. 29 3.54-3.189. Consonant combination 31 3.82-3.97. -c-first personal pronoun 35 3.98-3.111. ^-perfective 37 3.112-3.133. Final h 40 3.134-3.135. Tone change and assimilation 43 3.136-3.140. Interrelationship of phonetic processes 43 4-4.36. THE WORD 46 5-5.114. THE NOUN 56 5-5.2. ' Possession 56 5.3-5.19. Monosyllabic nouns 57 5.20-5.113. Compounding 61 5.22-5.37. Nominal suffixes 62 5.38-5.70. Nominal prefixes 66 5.71-5.113. Composition of nouns 71 5.114. Borrowed nouns 78 6-6.38. JTHE PRONOUN 80 6-6.12. Person and number 80 6.13-6.14. Independent pronouns 82 6. 15. Possessive pronouns 83 6. 1 6. Possessive pronominal prefixes 83 6. 17. Emphatic possessives 84 6.18-6.18a. Subjective and aoentive pronominal prefixes 84 xii CONTENTS (5.19-6.28. Objective pronominal prefixes 85 0.29-6,32. Other pronominal prefixes 89 0.33-6.34. Demonstrative pronouns .' 90 6.35-6.38. Indefinite pronouns 90 7-7.116. BOUND FORMS 92 7-7.10. Bound forms initial position 92 7.1. Demonstratives 92 7.2. Interrogatives 92 7.3-7.10. Adverbial elements 93 7.11-7.103. Postpositions and enclitics ,,«,,, t s s ...... : s s s 95 7.104-7.116. Compounded© of postpositions and enclitics 115 8 8.104. THE VERB 119 s, 7-8.30. Intransitive and transitive 120 8.31-8.35. Static verbs 127 8.36-8.81. Active verbs 129 8.38-8.42. Progressive-continuative system 129 8.43-8.47. Inceptive system 131 8.48-8.50. Cessative system 132 8.51-8.53. Customary 133 8.54-8.61. Perfective 134 8.62-8.72. Repetitive system 135 8.73-8.76. Imperative 137 8.77-8.81. Optative 137 8.82-8.84. Interrelationship between tense, aspect, system, and mode 138 8.85-8.91. Phonetic character of vbrb stem 139 8.92. Augmentative 141 8.93-8.94. Diminutive 142 8.95-8.104. Irregular verbs 143 9-9.23. THE ADJECTIVE 147 9.4-9.12. Comparison 148 9.13-9.22. Numerals 150 9.23. Money 153 10-10.124. PREFIXES 154 10.1-10.14. Position of pronomlnal prefixes 154 10.15-10.20. Objective prefixes 157 10.21-10.24. aspecnve-inflectional prefixes 158 10.25-10.26. Inflectional prefixes 159 10.27-10.34. Pre -paradigmatic prefixes 159 10.35-10,64. Principles of conjugation 162 10.47. Glide prefixes 165 10.48. Retroactive influence 166 10.49-10.54. Contraction 166 10.55-10.58. Position of prefixes 168 10.59-10.64. Voicing 170 10.65-10.73. Arrangement of paradioms 171 10.74. List of prefix paradigms 173 10.75-10.124. Paradigms 177 1 1 -11.118. SYNTAX 293 1 1-1 1.25. Position of elements 293 11.26-11.54. Syntactic elements 300 11.55-11.86. Negative 307 CONTENTS XU1 11.87-11.100. Interrogatives 315 11.98-11.100. Interrogative* with "Be" 320 11.101-11.118. Connectives and clauses 322 11.104-11.111. Clauses 323 11.112-11.118. Cause 329 12-12.60 . USAGE AND VOCABULARY 332 13.1-12.18. Time and place 332 12.19. Thus , 336 12.20-12.27. Number and quantity .- .- ... 337 12.28-12.60. Verbs ; .. . 339 12.28-12.43. Type verbs 339 12.44. Verbs of force and speed' 351 12. 45-12. 46a. Verbs of animated motion 352 12.47. . Verbs of doing and making 357 12.48-12.53. Verbs of being and becoming 361 12.54-12.60. Verbs of communication 365 13-13.54. SPEECH DIVERSITY 369 ANALYZED TEXT 382 BIBLIOGRAPHY 392 1-1.26. INTRODUCTION 1. This Navaho Grammar was begun in 1930 as a means to the end of investigating Navaho Religion. 1 In a sense the task was an in- voluntary undertaking because at the time, and for several years after, it seemed that the language could be learned by a practical application of available sources, or of papers proposed for immediate publication. At that time the late Professor Edward Sapir had studied Navaho for some years and had a bevy of students working on it, so that Navaho was classed as ' 'a recorded language." For some weeks I attended Sapir's class in Navaho at the University of Chicago, and there got an outline of his analysis, which was later supplemented by conferences at Yale University. At the same time I was seeking a basic pattern for the language, I was trying to speak it. As time went on, I realized, too slowly, that the structural pattern I was struggling with did not have a practical application, that is, the forms were too theoretical to be understood by the Navaho. This unsatisfactory result was not due to mis- pronunciation, for I passed most tests designed to differentiate tone, quantity, and the like. The forms simply did not fit the formulas given. I was primarily concerned with ethnological research, so that it was not until 1937 that I finally decided to start the language work as near the beginning as seemed necessary, especially since by this time the publication of the grammar was more remote than in 1930. Another circumstance contributed to my decision. Adolph Bitanny (AB) was one of my interpreters in 1934 and later, under the in- fluence of Professor Hubert Alexander of the Department of Philos- ophy, University of New Mexico, had become interested in the analysis of Navaho and its relation to philosophy. AB was parti- cularly intrigued by the concept of "aspects" and attempted to apply it to Navaho. Interestingly enough, he came up with an outline of principal parts and prefixes that resembled more closely than anything P. E. Goddard's analysis of Hupa and Kato, a but AB's was of necessity much extended. In 1937 he came to New York where he spent the winter working with me on transcriptions, vocabulary, and morphology. To AB then, I owe the foundation of this work. We spent hours isolating principal parts and their signifi- 1 Beichard 1950. a Goddard 1910, pp. 112ff.; 1912, pp. 42ff. 2 NAVAHO GRAMMAR 1 . 1 cance, classifying stems, and working on prefixes. AB's contribution was so basic that, had events permitted, h§ WOllld have been & collab- orator in this work. As it happened, however, he went into the Army and slowly the Grammar moved so far beyond his scope that he would not understand much of the analysis as it now stands. Another effect upon written Navaho goes back indirectly to Sapir. Robert Young, a student at the University of New Mexico, was also influenced by Alexander (one of Sapir's students) and by AB. Later Young was put in charge of Navaho language work for the Department of Indian Affairs ; he trained William Morgan (WM), and the two collaborated in producing a Navaho grammar and dictionary for Navaho and laymen who might want to learn Na- vaho. 3 The prefixes of the Young-Morgan book are treated in the Sapir manner, but are of great value because full (or nearly full) paradigms are given. The principal parts of the verbs are also interpreted according to Sapir's principles, but underneath each verb many prefix paradigms not included in the grammar are written out in full. Furthermore, some of the most commonly used expres- sions are listed, and the transcription is almost faultless. Usually, therefore, the material could be adjusted to my analysis, and when that was impossible, I was able to work with WM, who has by this time a remarkable training and facility in Navaho analysis. The foregoing explains the reason for this grammar. Actually there is no Sapir grammar — it is in the form of various papers on the Apachean languages, largely theoretical, published by Hoijer since Sapir's death. 4 Since my view of certain basic principles differs greatly from Sapir's a section will be devoted to the differences in the analysis of the Sapir school and mine (1.7-1.26.). 1.1. A word is in order to explain why earlier efforts at written Navaho are not adequate for learning the language. An Ethnological Dictionary of the Navaho Language (1910) and A Vocabulary of the Navaho Language (1912) by the Franciscan Fathers, and Dineh Bizad by Rev. F. G. Mitchell (1932) are unsatisfactory because tone, quantity, and other important phonetic details are not distinguished. Trying to read Navaho is therefore a guessing game for those who already know it, rather than a means of communication. Father Berard Haile who was a motivating force in the Franciscans' work, has since published excellent texts in an accepted orthography which have been extensively used in my work, both linguistic and ethnological. In 1926 Father Berard published A Manual of Navaho Grammar, and since 1941 a series of four phrasebooks Learning Navaho. I cannot agree with his analyses. 5 3 Young-Morgan 1943. 4 Hoijer 1945a, b; 1946a, b; 1948a; 1949. s Cp. Hoijer 1961. 1.2.-1.6. tffT&omjctfiotf ' $ 1.2. A major lack in all the works mentioned is the fact that the verbal prefixes have been treated only cursorily and have hardly been translated. The result is the same as if Latin were explained on the basis of stems with formal suffixes, none of which were differ- entiated by translation. Since Navaho is a living language spoken by some 60,000 persons two-thirds of whom do not and perhaps never will speak English, and since there are cultural reasons why it should be written, various attempts have been made to reduce the language to a medium that would allow written communication. Anthropologists have been able to record many dying Indian languages in phonetic writing despite the protests of those untutored in linguistics, and the limitations of the English printing press. Navaho, however, has been subject to pressure groups of all kinds, most of whom consider that "if it cannot be written in English symbols, it may as well not be written." A few, however, have realized that linguistic principles not found in English prevail in Navaho and must be indicated. Among them is Father Berard Haile who has changed his orthography at least three times to my knowledge. He now publishes in the last revised system employed by Sapir and Hoijer. 1.3. The Department of Indian Affairs has devised a special format for its publications. In the process of simplification several symbols were eliminated as unnecessary which, in my opinion, vitiate the recording for historical purposes. Most people who can read it can read the other systems as well. A great advantage is that several groups interested in Navaho have agreed to use it. 1.4. The system favored by linguists is that devised by Sapir and used by Father Berard. It is complicated, including several Greek symbols, and symbols for the sibilants which seem arbitrary to the layman — as indeed they all do! Actually they follow the Inter- national system of phonetic transcription. In his Navaho Phonology, Hoijer has made a few capitulations to the Press, which must be confusing to one trying to learn Navaho. In short, there is so little agreement about writing Navaho that the novice experiences ex- treme frustration at the outset, and it is small wonder that he gives up trying to learn Navaho almost at the first lesson. 1.5. Since it is impossible to choose a system which everyone may use and read, since it was necessary to be uniform about the grammar and contemplated dictionary, and more particularly, since I started with the system that was mechanically well worked out on the printing press, I have adhered to the system used by American linguists before Sapir's last revision. Besides saving time, it has the advantage that the symbols of the well-developed sibilant system 2 Reichard 4 NAVAHO GRAMMAR 1.5.-1.7. are more suggestive than the later system, for instance, ts instead of c, j instead of z, etc. I have transcribed all references in this system no matter how they were originally written. 1.6. I have cited many examples from published texts — Sapir, Hoijer, Father Berard Haile, Young-Morgan, and I have even revised some of Matthews' writing (not used in this work), but I have not always adopted the author's interpretation, in fact, I have often used his example to illustrate my own point. In presenting this grammar, which may seem unduly complicated, I have tried to elucidate the principles of the language, principles which I myself do not find simple. I have tried to make the grammar as complete as possible, though there is scarcely a subject with whose analysis I am completely satisfied. Consequently, it cannot be expected to be elementary ; it contains too many ideas that are not common linguistic knowledge. Nevertheless, I hope it may furnish a source from which Navaho forms and idioms may be drawn. I have tried to keep historical questions and references to a minimum, the major purpose being to present the Navaho language. However, I have had in mind the need for genetic reconstruction of Athabaskan and have tried to include all data that may contribute to knowledge of family relationships — my insistence on differentia- ting h and x, for example, and my reluctance to reduce y and y to y (3.8.). 1.7-1.26. The Sapir School of Athabaskan 1.7. Because of Sapir's influence on Athabaskan studies it is necessary to discuss these materials, particularly the points of dis- agreement between them and this work. The works are by Sapir himself, by Hoijer for Southern Athabaskan, Apache and Navaho, and by Li for California (western Athabaskan) and northern Atha- baskan (Sarsi and Chipewyan). Some of the major conflicts are due to approach. The reports of the Sapir school indicate as a primary purpose the reconstruction of primitive Athabaskan; as another, the demonstration of the method of what has come to be called "structural analysis," purposes which are largely theoretical. The interpretation of a particular language as a living, cultural pheno- menon seems to be almost incidental. 6 Most of the criticism of Navaho will of necessity be cited from Hoijer's works since it fell to him to publish much of Sapir's work. It is impossible to dissociate their contributions. Hoijer seems to fl Hoijer 1946a, b, c; 1947, 1948a, b, 1949; Li 1930a, b, 1933; Sapir, see bibliography. INTBODTJCTION O have hewe which is most common in Navaho, is "organic" o " "inorganic." The question should be left open so that the student does not accept a neat, but possibly incorrect conclusion which prevents him from further research on the subject. 10 This criticism may be leveled at many reconstructions which, on the basis of Navaho, seem to me to be founded on deceptive cognates or false etymologies. 1.11. 1 cannot agree with the Sapir school in accepting as readily as they do the theory of alternants, the principle that several forms are interchangeable in meaning. 11 Often it has proved an evasion. Usage does not allow doubt ; one can no more compose a form with one or another element than a compositor can strike a letter half a half i on a linotype machine. Navaho texts prove this contention. It is true there are some alternants but, like all overlapping, there are limitations which must be discovered. 1.12. The principle of alternants, together with the lack of distinction between prefix and stem meanings, has led to confusion in the determination of aspect and mode — some so-called alternant forms were generalized, but the result is uncertain, whereas the actual principal parts are much simpler, and patterns can be dis- cerned through the recognition of all the aspects. The shortcut of alternants was accepted not only for verbs, but for other elements as well. For example, Hoijer ascribes to alternants, without indicat- ing change of function, the forms tcq-' "excrement" and bitca-n "his excrement," neglecting to mention bitcq*' "his excrement." 12 One explanation of these two forms lies in meaning and usage: tcq m ' "excrement" (possessive bitcq-') is an inelegant "household" word, but bitca-n "its manure, ordure" may be used as a polite form. And not only is there a difference in usage, but there is also a rule that an unpossessed noun with form CV has a possessive of form -CV*n to signify that it is not a mere possession, but a possessed part in relation to a whole (5.13-5.16.). There is therefore a phonetic, morphological and semantic differentiation. 9 Hoijer 1945c, pp. 7, 8, 10, 11, 14, 15, 16, 17, 19, 24, 27, 28, 30, 34, 40, 41, 42, 47. 10 Ibid., p. 39. 11 Ibid., pp. 32ff; Hoijer 1943, pp. 39-40; 1946c, p. 72; Li 1930a, pp. 53ff; 1930b, pp. 15ff ; 1933, pp. 126ff. 12 Hoijer 1945c, p. 36. 1.13-1.15. INTRODUCTION 7 1.13. As the discussion of prefix and verb stem and the inter- pretation of their combination as aspects implies, much more can be done than has so far been accomplished by insisting on a more rigid analysis of morphology through the investigation of meaning. The formulas of the Sapir outline often attempt to explain several prefixes as if they were one. This result may seem to be due to insufficient data. Since Sapir and Hoijer had a large mass of materials, such a conclusion is false. The effect is due rather to the extreme difficulty we all encounter in holding Navaho informants to paradigmatic forms of the same series. The language is rich in forms and many have meanings so similar as to seem the same in English, so that informants mix paradigms and give a single mean- ing for the distinctive forms. The Young-Morgan grammar some- times suffers from the same affliction. There is, to be sure, a great deal of overlapping, but there is also some distinction, usually phonetic, which shows the independence of many elements, especially of prefixes. If these distinctions are isolated, there remain relatively few morphological irregularities. This result is contrary to former analyses in which forms have so often been described as "irregular" that the student gets the idea they are more common than regular forms. 13 It is hoped that this grammar will show that, though the patterns are somewhat unusual, the language is quite regular once the linguistic scheme is realized. 1.14. Sapir's interest in "pattern phenomena" is well known, and I have always wondered why he and his students failed to apply this useful theory very extensively. They worked out some of the elementary assimilations and contractions, but failed to push them to the obvious conclusions dictated by the pattern. 14 It seems to me that one of the lessons pattern has to teach is that once it be- comes established, in language as in other cultural phenomena, it is often carried far beyond what may seem to be "reasonable" limits. If therefore this happens in a language, we may properly extend the analysis as far as the language allows. I may be accused of having expected too much of the rules I have found, on contraction, for example. I feel justified by the results for which I think there is proof. 1.15. The limitations so far discussed apply to the study of Athabaskan. Let us now examine some more specific details of Hoijer's analysis with which I am forced to disagree. One reason another Navaho grammar seems appropriate is the peculiarity of the examples cited, although, to be sure, neither Sapir nor Hoijer 13 Ibid., p. 50; 1938a, p. 89, n. 1 :22. 14 Li utilized the concepts more than anyone, but even he stopped long before realizing the suggested possibilities (cp. Sapir, 1925, p. 194). 8 NAVAHO GBAMMAR l-lfe.-l.18. has attempted to aid a speaker of commonplace Navaho. For in- stance, they give many examples derived from the stems for "go." These stems are distinctive for singular, dual, and plural — this is in itself a curious adaptation for the English speaker to make. And in addition, the singular stem is one of the very few Navaho stems that change consonant initials in an "irregular" manner. Formally therefore the treatment of "go" is atypical. Morphologically too the stems for "go" may seem strange. Although to the English speaker "go" is an intransitive verb, in Navaho it may be treated as a transitive with forms in the active and passive voices. This possi- bility, though by no means uncommon in Navaho, makes the verbs for "go" additionally difficult to the novice. 1.16. From the semantic viewpoint the stems for "go" are extremely important because with various prefixes and in com- bination with other elements they determine many fundamental idioms. The criticism here is, not that "go" should not be treated, but rather that it is a single exception which has been used to illustrate a type. 1.17. Another example constantly cited by Hoijer is regular, but it has become so generalized in meaning and so extended in form that it is far from typical of the class of verbs it is used to illustrate. This is the stem ~&l "round, convenient object moves." 15 It is the most generalized of all the stems in the class called the "type stem" (abbreviated T), and therefore only rarely demonstrates what purports to be its primary or essential meaning. It is a great tempta- tion to use this stem as a type form, one to which I myself some- times yield, yet it should be understood that this stem is funda- mental in contributing many idioms, and that consequently modi- fications of the rules may be encountered. The stem -a*l also illustrates a point previously mentioned, namely, it is a stem which has alternant forms in all principal parts save the progressive. This fact makes it confusing to the novice, and "atypical" of the verb class which must be understood from the very first. Therefore other stems from the list of type stems often illustrate the prefix com- binations and the usage more simply and clearly. 1.18. The assumption of "inherent tone" seems to me not to be validated by the behavior of Navaho elements, particularly the prefixes. This assumption implies that some "syllables" are essenti- ally low, others high, and some neutral. 16 1 do not know of evidence to support the assumption. We ought to know, for instance, why 16 Unless otherwise noted the progressive stem is cited as basic in the discussion of the verb. 16 Hoijer 1938a, p. 74; 1943, p. 39; 1945c, pp. 50, 56, 68; Sapir 1925. 1.18.wl.21. INTRODUCTION 9 there are radical differences in tone structure of Navaho and Sarsi, and why tone in Chipewyan is often just the opposite of that in Navaho. Since Sarsi and Chipewyan are the only two major northern Athabaskan languages for which pitch has been worked out — -and I for one look to the north or northwest as the place of origin of grammatical pitch — it seems premature to rely on a principle so insubstantial. These remarks do not mean that rela- tionships should not be examined and proposed; they merely warn against final acceptance with the resulting cessation of inquiry. 1.19. The problem of inherent tone is one with the definition of the syllable. I operate with the concept that the Navaho prefix syllable is of the form Ca-, Ci-, or Co-, and the basic stem syllable -CVC. The tone of these basic syllables is assumed to be neutral (not differentiated in writing from low). Any variation of these simple vowel forms — e, some o's, lengthened vowels, any tone except neutral, nasalization — are due to contractions, many of which have been ascertained, more of which may be discovered by comparative analysis. The reduction of the syllable to such simple forms has led to the isolation of many prefixes. It may even explain such stem forms as -CVC progressive or present, in contradistinction to -CV'C inceptive, or -CV*' perfective. It is possible too that the final con- sonant of the stem syllable of form -CVC may prove to be a stem contracted with a tense or aspective suffix. 1.20. The acceptance of CV as a syllabic form establishes the ideal that if a form differs in any respect from CV it needs explanation, and that the elements composing it may be susceptible of meaning. Questions arising are: Why is a, i, or o long ? Why is there no vowel at all ? Why is the vowel high ? Nasalized ? High and long ? High, nasalized and long ? Why is the vowel e or e, or any of their lengthened or high-toned variants ? Why is n syllabic ? Is it equi- valent to na- or ni- ? Why do we find -Vn instead of -y- ? Many of these questions and others are answered in the chapter on Prefixes (10-10.124.); it is essentially a chapter on phonetic processes. 1.21. In relation to "inherent tone" some vowels combine with others in a way entirely different from others with the same forms. Compare, for example, 'a-beyond and 'a-indefinite pronoun (10.76b, 10.103.). What is inherent in 'a-beyond that makes it combine with yi-continuative to become 'r-, whereas J a-indefinite pronoun ab- sorbs 2/z-continuative with hardly a trace ? Or is the inherent tone of ?yi-continuative the reason for the difference ? Is di-start from completely equivalent to di-emit ? Are they distinct by origin, or have they become so by development ? Is di-start from related to d£-start from against, or is the tone inherent ? If the latter ? when 10 NAVAHO GRAMMAR 1- 2 1.-1.24. and where was it inherent ? In Navaho ? In primitive Athab as k an ? In a language transitional between the two ? 1.22. In connection with these and other related questions Hoijer posits a theory that the vowel of certain high-toned syllables assimilates to a preceding high-toned prefix (syllable). 17 This rule is not thoroughly substantiated because the effect occurs in only a few restricted circumstances — in the continuative and perfective aspects and, more important to note perhaps, only in the singular and dual forms, not in the plural. Hoijer does not tell us which prefixes require this form of assimilation nor does he mention the fact that the effect is variable. An attempt to test the rule gave rise to the determination of the "inflective prefix," an element affecting tone, and explaining the compounding of several prefixes in the perfective as well as in the continuative forms. The rules for the use of the inflective prefix have few exceptions, although some meanings still remain doubtful, and the ultimate reason for the inflective prefix is undetermined (cp. 10.25-10.26.). 1.23. The isolation of the inflective prefix, which immediately precedes the pronominal prefix of the verb complex invalidates Hoijer's place numbering of prefixes. 18 Moreover, he disregards the difference in position of the subjective and agentive pronouns, an omission that changes the place number of the pronouns and leads him to a different pronominal system for the continuative and perfective. These omissions are major shortcomings, for place has no meaning if only one element is in error. Still another point of dis- agreement is that, if place numbers are to be used satisfactorily, they should be from right to left, instead of the reverse, counting backward from the stem. The stem is always ascertainable and the three prefixes immediately preceding are essential, whereas any or all the prefixes Hoijer has in places 1-9 may be lacking. 1.24. A matter of determining phonemes seems to me of great importance. When a language has as many similar forms with distinctive functions as Navaho, how can the phonemicist do it justice, and at what point in the course of his study may he trust his judgment that two (or more) sounds may be safely reduced to a single phoneme? The answer to the question is difficult, but a warning may be sounded to the effect that a premature deter- mination of the character of sounds and particularly of their functions, may obscure or eliminate a morphological process. 17 Hoijer 1943, pp. 39-41 ; 1945c, pp. 30-1. In the work last cited in note 13 Sapir asks similar questions, but I know of no place where he answered them except by assumption. 18 Hoijer 1946a, p. 1. 1 ;24.-1.26. introduction 11 An exa/mide from "Na.vnTin illnflf.rfl.+.Afl this point. The definition of the voiceless stops t and k includes a discussion of aspiration, which in my opinion is of historical significance. I agree with Hoijer in omitting from writing the secondary phases of the pronunciation of t and k and related sounds — aspiration, palatalization, labializa- tion. In considering t and tx as alternants, however, the morpho- logical function of x has been overlooked, x in some cases and its voiced counterpart y constitute a phoneme that forms a consonant cluster with a whole series of sounds, unrelated in certain respects: tx, sx 9 zy> dzy, t$x y ex, tcx, Ix (cp. 8.92.). The clustering of x or y with a consonant initial is a true infixing process, and denotes an augmentative or pejorative. Hoijer gives an inkling of this process in the statements: u t8 is not as strongly aspirated as the phonemes t and k, and is never followed by an #-glide." But to this a note is affixed: ' 'There is one exception to this rule: if a word containing ts is pronounced very emphatically, as in a command or exhortation, the ts phoneme may be followed by an #-glide." 19 Since the aug- mentative function of x was disposed of for phonemic expediency, the texts, for the most part, lack words that indicate the augment- ative or pejorative. 1.25. Hoijer mentions a rule to the effect that a stem of the form -Ct' may take the form -CVnV when a suffix of the form -P is added. 20 He fails to show, however, that there are bisyllabic stems of the form -CVnV or -CVCt which are diminuatives (8.93.). Almost certainly such stems became crystallized by an historical process different from that now recognizable as suffixation, or at least by a process that had a distinctive result. 1.26. If the student be irked by the extreme stress on small details in the analysis of Navaho, he must realize that such emphasis is functional, and more particularly, that the details define pro- cesses and significance that now often seem unique, but may doubt- less be found in other languages once attention is directed to them. 21 The major question is not only what forms exist, but also where the lines are drawn within a single category of form — what is mechanical, what is morphological, and what is historical or genetic. Meaning seems to be the key that can open these doors. Not etymology, semantics, phonetics, phonemics, or morphology alone, but all in their fascinatingly intricate association. 19 Hoijer 1945a, p. 12. 20 Ibid., p. 34. 21 Cp. Reichard 1938, pp. 553-9. 2. GRAMMATICAL PROCESSES 2. The grammatical processes by means of which Navaho words are modified are: affixing, including prefixing, suffixing, and in one case, infixing; of these prefixing is most common. The affix frame, that is, prefix with suffix is also a common syntactic device. Phonetic changes, particularly those caused by contraction (sandhi) are of great importance in morphology. They include change of consonant by assimilation and juxtaposition, change of vowel with various combinations of vowel quality, quantity, and tone, and even change from consonant to vowel, indicated by change of tone. Still another phonetic change with morphological significance is voicing. Although position is relatively free in some respects, it is never- theless an important device, particularly in indicating the relation of pronouns, verbs, and postpositions. 3-3.140. PHONOLOGY 3. For various reasons some of Hoijer's work must be repeated here: His recording differs from that in this work. It is essential to present the system here used. Some qualifications will be stipulated, some modifications and additions will be made to the discussion of phonology. Hoijer's analysis has been adopted to a large extent although some differences are noted because of interpretation. 1 3.1-3.7. Vowels 3.1. The Navaho vowels are the following: Low-central unrounded a as in English odd Mid -front unrounded e as in English met High-front unrounded i as in English bit Mid=back rounded o as in French mot All vowels have continental rather than English values, that is, they are pure vowels, and when primary, they are very short. The vowel a is so short that it is often heard as the obscure vowel of English "about, above;" the variation is non-phonemic. Although a, e, and i are near the English equivalents if not diphthongized, o varies considerably. It is between o and u as in English "look" and is often heard as u ; this variation is not phonemic. 3.2. The vowels may be lengthened, but lengthening does not cause diphthongization. Quantity is very important in Navaho because it has morphological significance. The symbol - indicates length, so that one type of vowel modification is: a*, e*, r, o\ Vowels may have any of three quantitative values, short, half- long, and long. Of these short and long are phonemic. The long vowels or vowel clusters are only half-long before some voiceless consonants, particularly h 9 x, s, c, I. Length is important because it indicates contraction, but it is often difficult to differentitate half long from short vowels in this position. 3.3. Lengthening a vowel may be a means of securing emphasis: '&din 'd'din 'ddvn "there was none, none, none;" 'ani'dr 7 (< 'anidi- i-') "very fresh (tracks)" (NT 132:1). Such lengthening occurs 1 Hoijer 1945c. 13 14 NAVAHO GRAMMAR 3.^.-3.6. particularly before the final glottal stop of a word. The glottal stop varies in strength; it may be very weak, but when used for emphasis after a lengthened vowel, it is articulated with a deliberate release which sounds much like the exaggeration unconsciously achieved by speakers first learning to pronounce the glottal stops. The word for "Let it alone!" is no-we' ; it may be used to a child the first time it is admonished. If, however, the child persists in investigating a for- bidden object, the adult may say, no'wt^ with an exaggerated lengthening of the second vowel and a release of the glottal stop which is truly a "catch." A conventional ending of a prayer is the phrase, xqj$ ndxdsdlf'' "it has become beautiful again." Prayers are generally repeated perfunctorily with considerable elision and speed, although all words must be articulated. The first three times the phrase sounds as it has been written, but the last time, it is x$j§ ndx&sdU" 9 with the exaggerated release of the glottal stop. 3.4. All four vowels may be nasalized; they resemble French nasalized vowels. Nasalization is indicated by a cedilla under the letter: q, % y i, p. The nasalized vowels may be long: qr, g", |-, g\ Navaho has pitch instead of stress accent; consequently every vowel must be thought of in terms of tone. When no accent is used, the vowel has a neutral tone, that is, the tone is selected arbitrarily. If two unmarked vowels occur in succession both should be pro- nounced without change of pitch. The significance of pitch is the relationship of one tone to another. Consequently a high tone, indicated by an acute accent over the vowel — a, e, i, o— signifies that the tone is high in relation to the speaker's chosen neutral (low) level. Vowels may be long and high; a-, i\ v t 6', and they may be long, nasalized and high, #\ $•, (', 4'- 3.5. When tone is grammatically important, as it is in this language, change of tone cannot indicate emphasis. Women, how- ever, scold or indicate surprise by raising the level of the neutral tone, and consequently the relative level of all the others. This device is quite as effective on children, husbands, and dogs as the elaborate glides in English. 3.6. When two vowels with differing tone come into contact, they may merge into a single vowel which retains the tone of both, as di'd- (< di-i'd-), -&-- (< -a-i). The circumflex £ indicates falling tone, the inverted circumflex ¥ rising tone. These compound, and there- fore secondary, tones are important morphologically — they indicate meaning. When two vowels come into contact they may contract to one, or they may form a "vowel cluster" — Hoijer rightly prefers this term 3.6.-3.7. PHONOLOGY 15 to "diphthong." If the tones of the vowels forming the cluster differ, they are usually retained in a rising or falling accent, written on the first vowel: naydi "one who goes about;" cinai "my older brother." Vowels that combine to form vowel clusters are: ai, ao, ei y eo, io f oi, qi, qo, gi. Two tendencies compete for vowel combination — the tendency for two vowels to combine into one, and the tendency to form vowel clusters ; the function of both is therefore in the same class as length, changed tone, nasalization. In other words, vowel clusters are a result of combination, most commonly of contraction. 2 My researches have not confirmed the statement that all vowel clusters are long. 3 I therefore conclude that the same rules of quantity apply to vowel clusters as to single vowels. For instance, citcai or citcei means "my grandfather," so called out of respect, not necessarily a relative, but citcai', or citcei' "my mother's father," and by extension, "my mother's parent, parent's sibling, my daughter's child." In prefixes, dai- or dei- "they pi.... it," but dard- "we pi ," or "we pi. . . .it." The long vowel cluster may indicate that either component is long, or the cluster may be long because two short vowels or a short and a long vowel have combined. 3.7. Syllabic n 3.7. A vowel, derived from CV, often na- or ni-, which must be added to the vowel series, is syllabic n, a sound that has at one time consonantal and vocalic values. It should be treated as a vowel in the following respects: it takes the place of a vowel; it may be low or high. Hoijer considers syllabic n a phoneme and insists that it be written with an accent — he uses the grave accent {h) for this purpose. 4 Since n is derived from na- as well as ni-, since n- is equivalent to na- or ni-, and since na- becomes ni- or n- in certain settings and, correspondingly, nd- becomes ni-, syllabic n must be treated in relation to its setting. Since syllabic n does not always stand for the same thing, it seems reasonable to treat it as we do other modified vowels which are not always phonetically equivalent. I do not therefore mark the low tone. If n stands in syllabic position and has no accent, it is low. Two ways in which a vowel may be affected are lacking in the treatment of syllabic n. Obviously n is not nasalized, and when it is long, it usually retains the vowel, for instance, not -n*- or -w- but na-, ne'-, nv-, no*-, na--, ne--, ni*-, or no*-. The contraction of two interconsonantal vowels may give rise to -»'-. If so, the resulting syllable may involve merely the glottalized n which may be syllabic and stands for -ria- or -ni- (cp. 3.41-3.42.). 2 Hoijer 1943, p. 39. * Ibid., p. 30. A Ibid. s p. 11. 16 navaho grammar 3.8. Consonants 3.8. 3.8. The consonantal system i§ summarized ill the following table : Stops Nasals Spirant s T3 V § ■o -d to >- 1 *o > i 3 'o > 43 O 3 o O > Bilabial b m m Alveolar d t t n n Prepalatal 9 k U Labialized kw Palatal y Postpalatal y X Labialized yw,w xw Sibilants Alveolar * z 8 Blade alveolar J C Lateral alveolar I I Glottal stop - Aspiration h(x) Affricates 00 •a 8 o o o > > y dz ts a tc dl tl ts 16 ti According to position of articulation the stops are: bilabial b, alveolars d, t, t\ palatals g, k, ti, labialized palatal kw, and the glottal stop -. The classes of articulation are : the voiceless, lenis, unaspirated stops b, d, g; the voiceless, fortis, aspirated stops t, k, kw, and the voiceless, fortis, glottalized stops f, ti, and the glottal stop -. The two nasal spirants m and n are essentially like those of English; m is a rare sound in Navaho, and rd is a result of contrac- tion, ri, also a secondary sound due to contraction, is frequently found. Because the contractions are a developing phase of Navaho, not yet thoroughly crystallized, the glottalized nasals, like $, and, in prefix position, t, ri, ti, td, are preceded, rather than followed, by the glottal closure, or the two may be combined, for example, z and s are alveolar sibilants quite similar to the same sounds in English. The corresponding affricates are: dz voiced, lenis and unaspirated; ts voiceless, fortis and aspirated; and U voiceless, fortis, and glottalized. j is a voiced blade alveolar sibilant similar to medial s in English "measure;" j may occur in all positions, initial, medial and final. c is the voiceless blade alveolar similar to sh in English "ship." The blade alveolar affricates are : dj voiced, lenis and unaspirated ; tc voiceless, fortis, aspirated; and td voiceless, fortis, glottalized. I is a voiced, spirantal alveolar lateral; I is its voiceless counter- part. 3 \-3.9. PHONOLOGY 17 fhe three lateral affricates are: dl, voiced or semi-voiced lateral, aC ^imllj7 rf. pmnmmpoH wi+.k 7 roloae©. Sinoo d is unaspftated ill Navaho, its manner of articulation affects I. The corresponding voiceless lateral affricate is tl, pronounced t with I release ; ti is the glottalized form of tl. y is an unrounded prepalatal semi-vowel pronounced with enough friction to produce in some settings confusion with y; both are nevertheless phonemes. y, an unusual sound due to contraction, is pronounced with the stop slightly preceding y. Hoijer has y as a stem-initial before a and o only. 5 My vocabulary yields -ye as well as -ye "marry;" -ye-l and -ybl "move ropelike obj.;" -yil and -yil "push." Hoijer also states, "Neither sound (y or y) has any noticeable variants." 6 My prefix analysis and the texts indicate that y and y are sometimes as difficult to distinguish as x and h. Morgan often reduces y to y. He evidently considers them equivalent in certain forms where I find them distinguished phonetically and morphologically. This is an example of the y- problem pointed out in 1.6. Morgan would hardly distinguish the stems I mention above. y is a voiced back palatal spirant with a labial tinge, sometimes quite strong before o ; it is fronted before e and i, before e sometimes so exaggerated as to sound yy. Labialized yw may lose its y character almost entirely to become w preceded by a slight spirantal attack : 'awe*' or J aywe* J "baby;" ywalya\ or uwlya' "jail;" biyo tJ or biwo* 9 "his tooth." Hoijer and I have discussed x and h, 7 arriving at different con- clusions. I think x is best regarded as an initial phoneme, h as a final: xa-out, xastvn "man;" daA-forth; -o^-second person dual pronoun. Navaho interpreters do not consider these two sounds interchangeable and morphology seems to confirm theft protests against a single phoneme. However, this is a problem of overlapping, and we shall doubtless never have the material to come to an agree- ment about its significance. x is the voiceless spirant of y. When y or x precedes o, it is pronounced ywo- or xwo — w is written only if y is not pronounced. yw and xw before a, e, and i, however, are written because the resulting form is due to contraction, o survives in w. 3.9-3.14. Glottalizatton 3.9. Certain phases of the sound system have been rearranged to bring out values deriving from different viewpoints, and to explain why the system cannot be strictly phonemic. Overlapping is very 5 Ibid., p. 57. * Ibid., p. 15. 7 Ibid., pp. 15ff.; Reichard 1948. 18 NAVAHO GRAMMAR 3.9.-3.14. impressive in Navaho — there is scarcely a sound which i^ ex _ clusively this or that, in certain settings it may be both this an( j that. Is such a sound then a single phoneme or several ? 3.10. The glottal stop is a case in point : It is a sound in it^ own right: VeVA "west;" 'i'i^a "sun has set/' 3.11. When the sounds t\ U, U, td, tt are defined, glottalization refers to the manner of artipulation ; it is simultaneous or near- simultaneous with the closure of the sound of which it is a part. Incidentally, these sounds vary greatly from speaker to speaker. Some have a very strong glottal enunciation, others glottalize the sounds so weakly that the glottalized sound can only with difficulty be differentiated from the unaspirated sonant. These glottalized consonants belong in a group of sounds unaffected by preceding -d- (cp. 3.55ff.). 3.12. The sounds rd, ri, and y are preceded, rather than followed by the glottal release. This effect may be due to the character of m, n, and y, or it may be a reflection of the setting in which they occur, in other words of their function: -d-m > -rd-, -d-n > -»'-, and -d~y- > -y-, the first and last being rare, -d-n > -ri- being quite common. 3.13. The sounds ri and y as initials are not confined to stem syllables, as is rd, but belong to a larger class of glottalized sounds occurring as the result of contraction differing from that of 3.12. If the prefixes of type Ca- and 'a-, or Ci- and 'a-, or Ci- and 'i- occur in juxtaposition, as they often do, and there is some prefix, such as di-, ni-, or yi~ between them and the stem, they may contract with the following consonant to form either -'C- or 'C-. There is much dis- cussion among Navaho thinkers as to which is correct. As in other cases of overlapping, both forms are so frequent that both deserve consideration. Prefix initial consonants affected by preceding -a-a- or -i-i- are d (> 'd or t), n (> 'n or ri), s (> '«$), dz- (> ti or 9 tS), dj (> td or 'td), y (> "y or y), x (> '#). In prefix combination with -a'a- j becomes -fC (cp. 3.41.). Certain inconsistencies have been tolerated to avoid undue com- plication. For instance, it would be convenient to have -ri- stand for -d-n- as a stem complex initial, and y n for -a'a-w, but both are written -ri-. '£, J £i, and Hd have been retained to indicate -a-'a- plus d, dz, and dj, respectively. On the other hand, t, ti, and td have not been altered to indicate d- > f, d-ts> ts, or d-td> td. The position of the sound indicates its type, 3.14. When stems are paired in "light" and "heavy" syllables, the glottal stop ends the heavy syllable compared with -h of the light one: -tah "among," -ta? "between;" -tah (prog.), -/a' (pf.) "be in series." 5.15.-3.19. phonology 19 3.16-3.19. Aspiration and non-aspiration 3.15. The matter of aspiration is as complex as that of glot- talization ; the pronunciation of the aspirates is not as definite as that of the glottal stop. The consonant h varies from a long and emphatically articulated spirant to an almost imperceptible breath. I have noted elsewhere that x after a consonant t or k may be so emphasized that the features defining them as stops are lost in favor of the spirant x which remains — tadidvn > txddidvn > xadidi'n. % The strength or weakness of h or x depends upon the speaker's habits, as the Navaho recognizes when he talks about "x-speakers;" AB and FH belong with these; WM does not. The character of h as related to x is important in many respects, but particularly in determining historical relationships : x- is preferable to h as an initial -h is preferable to x as a final Consonants are aspirated, by some speakers very weakly, by others so strongly as to form consonant clusters — tx, kx, tsx, tcx — the second consonant of which seems to have no function. 3.16. The voiceless stops t and k differ according to the vowels that follow them. Before a, the aspiration tends to be notable, but not exaggerated; before e and i, the aspiration is detectable, but the stop sounds as if palatalized along with the aspiration, the full forms might be indicated as ihye, thi, or khye t khi. The vowels have the same effects on the spirants y and x. 3.17. The sounds t, k t y, and x, as well as U before o take on a cluster character of the type txwo y kxwo, ywo, Kwo* The sounds t, k, y, and x are written without aspirate or labial symbols, it being under- stood that the rules of aspiration and labialization function reg- ularly. 3.18. A cluster is formed by combining several sounds — tx, sx, tsx, tcx, lx y zy> and dzy — most frequently surds, followed by x. In cases of this kind the strong aspiration is a morphological element expressing an augmentative or pejorative (8.92,). In this work when x is written after a consonant it indicates the consonant cluster, that is, the augmentative form. 3.19. It is interesting and perhaps significant that the voiced stops are not aspirated, in contrast with their paired surds, which are often somewhat, frequently strongly, aspirated. The sonants 8 Reichard 1945, p. 162. 3 Keicbard 20 NAVAHO GRAMMAR 3.19.-3.22. have the usual distinction of unaspirated consonants ; their brevity and lack of aspiration sometimes cause English speakers to hear b as p, d as t, g as k, dz as ts, and dj as tc. It may be helpful to note that b is not common in the language and that neither aspirated p nor any other form of p occurs ; aspirated d and g are lacking. 3.20-3.21. Labialization 3.20. The infrequency of the bilabials 6, m and rri has been noted, as has the occurrence of w for yw. Hoijer lists w as a phoneme, unrelated to yw, and yw a variant of y before o. 9 w is also found in borrowed words — wdxindo-n "Washington, the U.S. Government" —but, like the other labias, w is not often found. We have already seen that labialization is related to aspiration; aspirated sounds retain their character while anticipating by lip rounding a following o. Since this is true for t and all palatals preceding o, they need not be written with w. On the other hand kw, and xw are retained because they appear before e and i. They may often be contractions of the type Owe < Co-a-i or Cwi < Co-i. Probably they are all contracted forms, but so far the elements of such stems as kwv and kwe'e have not yet been convincingly established, yw is probably a contracted form which in Navaho is sometimes further contracted to w. Neither sound occurs often enough to corroborate a theory. 3.21. The processes, nasalization, glottalization, aspiration, and labialization are related to various parts of the phonetic system. To understand them it is necessary to cut across the system from bilabial stops to glottalized lateral af fricatives, and to consider the relationship between consonants and vowels. 3.22-3.25. Assimilation 3.22. Assimilation is one of the fundamental processes in the study of Navaho. It is closely related to contraction; both are carried so far that I omit the term "sandhi." Hoijer has based his phonetic-phonemic analysis upon a subdivision of stems, prefixes ("prefinals," he calls them), and suffixes. For this reason he has in some cases failed to arrive at some generalizations that apply at least to prefixes and stems, others that apply to all three parts of the verbal complex. For example, the interrelationship between vowels, consonants, and vowels and consonants is fundamentally similar. Differences encountered have to do with the absence of some sounds 9 Hoijer 1945c, p. 18. 3.22.-3.25. PHONOLOGY 21 in certain positions, or with their frequency. In other words, dif- fered oas are a matter of probability rather than of essential phonetic makeup. 10 3.23. To illustrate, the d-agentive classifier affects the stem initial in the same way as -d- of -vd- first person dual pronoun. Both of these d's affect the stem initial y in different ways: If the first person dual form is yvdzol (prog.) "we are blowing," we may conclude that the zero stem is -yol related to -sol (< -l-zol) and -l-zol, rather than to -yol which might be related to ~^ol (non- existent), or to -yol with affiliations with -dol (not found). Again, yidzil "we 2 are pushing prog." would suggest seeking other forms of the verb complex in -yil (zero), -dzil (-d-form), -ail (-Worm), and -l-zil (-Z-form), rather than as -yil (zero-form), -dil ( d-form), -l-xil (-Worm), or -l-yil (-Z-form). As a matter of fact, the first series appears, the second has -gil as the -d-form. 3.24. Several processes of contraction are closely related to assimilation, since the dichotomy between vowels and consonants is not always preserved. These processes have been called "absorp- tion" and "saturation." Absorption refers to the combination of two or more prefixes of similar pattern with resulting form like one or the other, or both, if they happen to be exactly alike to begin with. For example, yi-3-3 continuative < yi-3 object-^i-continuative; or ni-2 continuative < i/i-continuative-m-2 subject. 3.25. A prefix is said to be "saturated" when it can absorb no other prefix without change — of length, tone, vowel, consonant, or position. For example, rfo*-3 future < d^-future- ^-progressive; but yido--3-3 future < yi-3 object-di-future-i/i-progressive. do*-, though very stable, cannot absorb any more prefixes and, when others are involved, some change must be evident in the result. Other ex- amples are: diyo*-3 repetitive aspect future < dt-future-^/i-pro- gressive-^-repetitive aspect; ditfo'-S-i future repetitive aspect < di- future-'a-indefinite object-^-progressive-^/i-repetitive aspect. The last example shows a change of position of 'a-indefinite object, which in the simple future precedes do*- as 'adcr-3-i future. The change in position indicates a closer relationship between 'a-indef- inite pronoun and -yi-repetitive aspect than between 'a- and di-. 10 It would be satisfactory to ascertain the reasons for the numerical dominance of n- f y-, and d- prefixes, for example, as compared with other sounds which might have been used. And we cannot help wondering why so few vowels — a, i, and o — are basic in the prefixes, causing such extensive overlapping. A plausible reason, of course is that the processes of nasalization, lengthening, tone change, vowel change due to combination, glottalization with its attendant effects on vowels and consonants, all substitute for such variation in vowel development as occurs in Indo-European and other languages. 8* 22 NAVAHO GBAMMAIt 3.25.-3.29. The principles of absorption and saturation will be used in this analysis of assimilation as they apply to single sounds ; they will be more fully demonstrated in the section on Prefixes which is essenti- ally phonetic (10-10.124.). 3.26-3.33. Assimilation Due to Mechanical Change 3.26. Changes are called "mechanioal" when they have no morpho- logical or semantic significance. If an unnasalized vowel precedes a nasalized vowel, the former may become nasalized (-V-CY > YCV) : 8<£§ (< 8i'4) round object is, lies x$$ (< xo-n-j$) it is agreeable, satisfactory, beautiful x$l§ (< xo-n-lg) there is, there are, it is available Some speakers carry this process back to two or three syllables preceding the stem ; others do not use it extensively. I have never heard the Navaho mention "n-speakers" as they do "^-speakers. They should, for differences in the use of nasalization are marks of speech diversity. Other recorders have unfortunately omitted the distinctions, probably because of a premature reduction of Y and V to a single phoneme. Some indication of the differences comes out in texts, for instance, -djV or -dj{' "to a point;" -ni\ -n% J "mind;" -ni't 9 -nyl "say to, tell;" fa- 'altso, id' "altsq "all" (cp. 12.57.). 3.27. AB had a notion that a vowel following a nasal consonant must be nasalized and high in tone ; he therefore refused to indicate the nasalization of vowels such as q, in cimd, camd "my mother." Hoijer, Young and Morgan follow AB'& practice, though they give no reason. The rule is not borne out by -mas (prog.), -mqs (pres.), -mq,*8 (inc.), -mq,'z (pf.), principal parts of the stem "spherical object moves;" ma'v, mq,'i' "coyote," and other stems. 3.28. The example sq'4 "there is a round object" illustrates another common assimilation — of * to a — from sVq, which al- though written by Young and Morgan, I have never heard spoken. 3.29. The effect of one vowel upon another in adjacent syllables is progressive or retrogressive ; any vowel may be so affected : bayan (< bi-yan) his home boxo*yan, b6*oyan (< bi-xo-yan) where his house, home is cittctfr (< cika'fr) my arrow that very one (NT 238;21) u ni da'dc ( ) I have combed it cidjo4 ■(< si-djo-l) bunchy substance is, lies yicdjo-l (< yisdjo-l) he has, keeps bunchy substance cidja-'* (< ai-dja,'') there is a granular mass yicdja^ (< yi&dja-*) he has, keeps granular mass cidji-' (< si-dje^) plural obj. are tc6t6il, tsdtdil hard oak; rock-plant de-cjah (< desjah) it is jagged, curved 24 NAVAHO GRAMMAR 3.33.-3.34. 3.33, The influence of one type of sibilant upon another varies in extent : cido-sq-l, aido'sq-l he will have my affection; he will love me bfrsistogi-, bi'cisfogi- ceremonial arrow dziditsxiz (< djiditsxia) he(4) is shaking in a paroxysm tsiade'zkt-z {< tsb}d&zk&-z) she(4) began to consider it, she thought it over dazdo-Usa-l (< dajdo-ltsa-l) he(4) is going to die (FH) t&il tain biya- xaddnisq, vegetation that grows under trees 'aze-'' t6il bd-xozini- herbalist; one-who-knows-plant-medieine (not y aze- y t$il bi-x6*zini' which would mean "medicine fragments, medicine broken like pottery") 'attfyh aodizin protective prayer (cp. fd' do* sits*q,- sinizini [FH] and fd* do* cittfyh sinizini [YMG 55] "don't stand in my way") 8*}'8 lijini' mole; the-wart-that -is -black i oainiai*hic (DD), "> aainiai-yic (FH) are you making a mistake? nirli'tsd-nic can't you do better than that ? tsi de-cjahi Jagged Rocks (place name) stfcdjo-l Old Age; old-age-lies-in-a-heap (NT 128:13) fd' Hi ' q\t$efyzni'zi*' > (< , dfjci-jni-zi' i ) "that must be the one," she (4) thought (EW 90: 17) Apparently the reciprocal influence of alveolar on blade alveolar sibilants is limited by the attempt to avoid ambiguity, for example, } at6a* xo'dzoh "boundary line, danger line, line beyond which one is not safe;" but 'altsd 'asdzoh "two lines spread, branch from one point." 3.34-3.38. Mechanical Change of Tone 3.34. Several rules of tonal effect may be called mechanical. A short vowel with high tone may be lengthened and its tone changed to falling when the syllable is followed by certain other syllables with low tone: 'a-demonstrative, "there remote, there near third person", xa-interrogative, "what in remote space or time, who, which of all possibilities," are examples of bound forms that so behave (7.1-7.2.). Many verb stems are in the same category. A partial list of elements before which the change occurs is: -r "the particular one that;" -r' "after . . .ing;" da m Ui "perhaps, maybe;" -di "in place, at; times;" do- future (abbreviation of do-le*l)\ -dah of the negative frame do- . . . -dah; ni "for a fact;" ndi "although;" -gi "in, in place;" -go subordinating suffix; la J exclamation of surprise, finality, conviction; la'na' "desire, wish;" leh "customary," -dji* "to a point." Note that the list includes independent words as well as bound forms : 'd-di (< *d-di) in place there remote xd-dji* ( -w, in addition to the tone and quantity changes). Examples may be found with^ alternant forms— either -V-CV > -V-CV, -tf-CV > -VCV, or -V-CV, or they may retain the original form. 3.36. Before certain syllables, however, the s>me kind of vowels -V- and -V'- retain their accent. Among them ar6 =e* ' 'concerning, custom, way;" -e* 9 future subordination; -r' "after having ..-,;" ~dah "for example, among others that might be mentioned;" -ni* "deceased, past, gone" (cp. ni "for a fact"): y ati6'h~e' concerning weaving nl6-e- Hail Chant xa*ct66-ltihi-dah Talking God among others (BS) litci** de-z'&i-nV the late Red Point Compare xaya* xayvkd'n-i*' (< xayvkq-v') after taking contained substance off (fire). 3.37. When a syllable with a short high vowel is followed by certain syllables with a high vowel, the first may be lengthened. 12 Elements that may so influence a preceding vowel are: -dfr' "from a point toward the speaker; along the way;" -do* "from a point away from the actor;" nte\'\ rite-' "past, used to be;" -cf* "doubt, probability ;" -dji "side ;" Id "evidently" (pres.) ; lei' "surely, indeed, as expected; a certain": 'd-d^*' ( 6£i-c{*) whatever it may be 3.38. Some long syllables become short when they become pre- fixes, or if part of a stem, when another element is added: 13 £6-* badcfcfdji-nili (< £<5*' ba- dcfcfdji'ni-ll) they (4) just give him things expecting no return) (NT 300:7) diydhi (< diyd'h-i) he who starts walking yidiyo-sfrl (YM 182) {< yi* di^o-sj-l (FH)) she will feed it, force food into it 'ajdjiH (< 'ajdjfr'i) those which someone (4) laid (on it) y 4i 'dsa*' bei-ltqji (< be* yidtqji) potdrum tapper 12 Cp. Hoijer 1945c, pp. 37-8, 13 Cp. ibid., p. 40, 26 NAVAHO GEAMMAB 3.38.-3.41. bitsV naxaido'lti'l (< nixa* yido-lti-l) he will give us his daughter (in mar- riage) (NT 308:17) cao'b^ (< ca*yo'b$) I lost at gambling The changes discussed in 3.34-3.38. probably depend upon two factors, the character of the vowel that takes the change, and that of the following vowel. These rules may well be clues to genetic relationships, especially in determining tones. So far, however, speech and texts are too inconsistent to make any deductions about the tonal effects final. 3.39. Glide Consonants 3.39. Glide sounds sometimes join stems and suffixes (glide syllables affecting conjugations are discussed in 10.47.). The two most common glide sounds are -d- and -g- ; their use and the choice between -d-> -g-, or occasionally -y- is a phase of lingual diversity: -igi- "the one who, that which" has the form -idi* on the western part of the Navaho Reservation (5.30.)* td-' dkd-d-igi that very one 'asinfoi'-y-ic are you making a mistake ? (FH) 'asiniai'-h-ic are you making a mistake ? (DD) na'fa-g-i- (< na-fa-i-) birds; the-particular-ones-that-fly-about (NT 106:26) 3.40-3.44. Vowel or Consonant Loss 3.40. A vowel or consonant may be lost: fd'l be: na* ntsdxdkcrsn (< tar Id be' nar ntsdxdke's ni) indeed you are considered to be in first place for a fact (NT 220:1) '(W deilni (< '4- Id) thus truly they spoke (NT 220:23) 'dkol td- di-nd-ln (< 'dko Id td- di-nd-l ni) even so you (must) go (NT 388:12) 'alnd^dna- (< ^atni^ gdna^) around the middle (NT 412:8) ba- neise-le (< ba- neise-l le) I keep dreaming about them (NT 234:11) di- y sdzdni-gi- (< di- 'asdzdnigi-) these who are women (NT 104:19) ddkwi-gdncq' (< d6kwi-go-ine-' > ) I am not sure how many (NT 276:16) xtf dtfrgo* 6ncq? (< xa'dte-go-gdne'') I'm not sure that . . . (NT 316:15) ni' J (< ni-i-^) yisol after saying ... he blew (NT 42:24) 3.41. Reference has been made to the effect deriving from a com- bination of prefixes of type Ca-'a-CV > Ca'CV or Ca'CV (3.13.). Here the process will be treated as a vowel loss with a possible globalizing effect on the following consonant, and illustrations will be given. A number of prefixes — 'a-beyond, da-plural, wa-down, net-about, #a-up — may occur before 'a-indefinite pronoun, "some, someone, 3.41.-3.43. phonology 27 something.'* If there is no other prefix between the combination Ca'a- and the stem complex, the forms remain stable; that is 'a- some is a paradigmatic prefix. However, if a prefix of type Ci-, such as di-, ni- t yi-, or dji- intervenes, the result is that Ca'a-CV > Ca'CV or Ca'OV : da'alyal "they pi. are eating some meat," but da'djilyal or da'tdilyal (< da'adjilyal) "they (4) are eating some meat." 3.42. Consonants occurring in such a position that may be glot- talized are the alveolars d and n ; y ; and the sibilant voiced affricates dz and dj: y a'fe*cnil (< 'a'ade'cnil) "I shall move some pi. obj. beyond;" dariinvl (< da'aninvl) "they are moving some pi. obj. beyond;" da'tdo'lyal (< da'adjo'-lyal) "something meatlike has been eaten by them (4);" da J t£ztig (< da'adziztig) "they (4) have woven something." In the following example Ca'a- precedes dji- which is contracted to -j-, but 1 attaches itself to n: bajnVd (< ba'adjintfq) "he (4) has lent round obj. to him" (YM 6). In the next examples d > Rafter -j- : 'dj'folzin (< 'a- 'adjidolzin) "he (4) maintains himself, his position;" y ij'tcrlxoc (< 'ayi- 'adjido'lxoc) "he (4) will go to sleep." I have not encountered 'j y though Hoijer has ^dnd^jdvso m l (< 'dnd-'adjidrso-l) "he (4) whistled again;" ^add'jnvyf'h (< 'ada- 'adjiniyf-h) "they (4) are beginning to eat." 14 Actually two prin- ciples are in conflict here — the tendency to glottalize dentals, and the necessity of preserving the prefix positions. Doubtless the conflict has not been settled and there is a choice of forms (cp. NT 264:7,266:21,268:15,282:17). • When the combination Ca'a- occurs before x, following the rule of position, the glottal stop precedes x\ in this case no more readily glottalized consonant follows: da'xe'sriil (< da'axcsriil) "some pi. obj. have been moved by them repeatedly." The globalizing process may also apply to Ci'a-: 'andziz bi-'fiyoji (< 6r' 'adiyoji) "trachoma; that-which-is-indefinitely-botryoidal- in-someone's-eyes;" bi'He'ldld-dv (< &i*' 'ade'dl&'di') "the particular one into whom the sun has shone;" xabi'te'Vf' 9 (< xabi'ade'Vi'') "he has been caught in the act by someone" (YM 102); xabVUo'ha'd (< xabVadzo % ka*d) "he has been slapped with some fabriclike obj.;" Hdifje-cnil (< 'axidi'aye-cnil) "I shall repeatedly move some pi. obj. beyond repeatedly." 3.43. As we shall see when considering prefixes, y is very unstable. It is often lost after preceding sounds, its loss causing two words to become one because neither the noun, postposition, or adverb preceding, nor the word beginning with yi- exists in the resulting form. Note in the following examples that yi- has different values, sometimes being a possessive or objective pronoun, sometimes an aspective prefix : u Ibid., p. 25. 28 NAVAHO GRAMMAR 3.43.-3.47. nttiz bd bitidriil ( -djol), -l-djol, or -l-djol. It is apparent that one change is -d-djol > -djol, an illustration of the fact that hardly a stem can be chosen which is not affected in some form or other by rules of assimilation. If, for instance, the apparent stem is found in the form yo-djol "bunchy substance is moving progressively," or yo'ldjol "he is moving bunchy material progressively," it is easy to determine that -djol is the stem, and -l-djol the stem complex. If, however, the form encountered is yvldjol "we 2 are moving bunchy material pro- gressively," it is impossible to know from this form alone whether the stem complex is -l-djol "cause bunchy material to move pro- gressively," or -l-djol "bunchy material is caused to move pro- gressively." The reason is that -d- of yrd-, the first person dual pro- gressive prefix, combines with -I- of -l-djol to become -1-, hence in this form yvldjol, but -1-, the passive causative can absorb -d- and results also in the form yvldjol. We shall see that the process is not an isolated, but rather a common phenomenon. 3.48. A more complicated type of assimilation occurs with -c- the first person pronoun, because it may absorb the stem initial or the classifier, and it may change the stem initial (3.54-3.133.). 3.49-3.53. Relationship between Vowel and Consonant 3.49. The relationship between vowel and consonant is another important problem. One aspect of this question is the nasality of the vowel in relation to -n, and to a vowel with a high tone. For instance, in compounding, a stem normally ending in -n may lose n and the vowel may be nasalized, or the n may disappear entirely. This may be an effect in the same class with the shortening of noun stems as they appear in prefixed forms (cp. 5.39.), but it may also be a principle of contraction which parallels others, as we shall see (5.38.) : tai-dil, tsi-dil (< isin-dil) bouncing sticks (used in game) tsi-tda'\ tsi-tSa-' (< tsin-tda-') box; wood-bowl tsi-Uiz, tsi-Hiz < tsin-ltiz) crack in log, wood tj-tse-d (< tin-tse*d) ice cream; ice-pounded ^q^d'tld* (< 'a'd-tt-rfa') he crawled into a hole do* ditci*cyi'8xf'dah (< ditcin ci-yi'sxj-dak) I have not died of hunger; hunger-has-not-killed-me (NT 50:7) 3.50. Nasality or n may apparently be lost in compounds, but actually survives as the high tone of a vowel. A class of nouns 30 NAVAHO GRAMMAR 3.50.-3.53. requires a high tone of the possessive: bitcf-h "his nose," instead of bitcf'h. All such nouns may be found in other Athabaskan languages as of the pattern -V-stem, or -n-stem. The important point here is not so much the derivation of the stems, but the fact that a process which operated in differentiating languages is at work in Navaho. We must, therefore, use the concept that nasality, n 9 and high tone are related. In practice, when I encounter a vowel with high tone, I conclude that nasality or n lurks within the vowel, and I try to discover its relation to the complex. This has proved to be a useful concept, but I do not feel justified in concluding that q is a vowel, that n in -an is a consonant, or that -i- of 'i- (< 'a-n-) is a vowel and nothing else. 3.51. Many Navaho speakers use -e-, others use -i-, a habit which may be considered mechanical. In compounds, however, some insist upon -e-, and rightly, I believe, since -i- may be considered "pri- mary" whereas -e- is "secondary or derived" (5.1.). In rapid speech the differentiation is exceedingly difficult to make. In analysis it is impossible to sustain the differentiation conclusively, but there is good evidence that -e- is a combined form e*- is certainly secondary. Since -e- or -e*- results from a combination of -*- plus n> and since we have varying stems such as -ni\ -n\, and -we' (pres.) "say, tell," I should consider -we' as possibly equivalent to -ni plus -w-, as is -n{\ Once more then, I question whether -e is a vowel, or a vowel plus a consonant. 3.52. The same problem comes up in connection with s which, may be "lost" in a high or lengthened vowel (3.98, 10.55, 10.107.), and again when ^-repetitive action appears in one paradigmatic form not only as xi-, but also as xe-> or xa- (10.114c, d, g.) or when xo-plaoe becomes xa- (10.116b.), the two last forms overlapping with those of rm-out of (10.85.). Is the vowel i, e, a, or o ? I conclude that we must have a variable scale by which to judge, a scale that forces us to keep in mind the possibility that one is not "the same" as another, and that the operating definition of the vowel must include the possibility that it may stand for something more, per- haps even a consonant. 3.53. A comparable reconsideration must be conceived for the interrelationship of consonants. They are related in series exempli- fied by the following : y y y, x, s, z, dz, I, I y, g, y> % i f i, di 3.53.-3.55. PHONOLOGY 31 xiic ltuso iwo series — J, c, dj ; and l, I, dl — seem phonetically and phonemically plausible, since they concern sounds, which by definition are related. Even y, g> y, and x have a class name — they are palatals — but the first relationship — y, y, x, s, z, dz, I, I — seems to cut quite radically across the phonetic definition. If y is related to g t y, and x, how can it be related to the alveolar sibilants and the laterals ? If y is related, on the one hand, to the alveolar sibilants and laterals, is it the "same" as the y related only to the palatals ? Apparently it is not, but is it then a phoneme ? We shall have to consider that it may be vowel, or merely the quantitative aspect of a vowel, and this is not included in a strict definition of a "sound." 3.54-3.133. Consonant Combination 3.54. This long, but necessary discussion will now be illustrated, first in connection with the effect of consonant combination. Of first importance in understanding (and therefore of "looking up") stems is the effect of classifiers upon stem initials. The final -d- of the dual first person pronoun has the same effects on the stem initial as d-classifier. They may be tabulated as follows: Table I j b w d t t n d ag. t b ni d t t ri I caus. V lb lm Id It U In I caus. pass. V lb Im Id It U In y(y) 9 U y (y) X z 8 d ag. y, d, dz g A* d, g — d, dz Is I caus. 8 ig IU lx — 8 — I caus. pass. I, dl ig w. h — Iz — dz t8 ts i c dj tc d ag. dz t8 ts dj Ic dj tc I caus. Idz Its Its c _ Idj Uc I caus. pass. Idz Us Its lj — Idj Uc a I I dl tl tl d ag. td Udl I dl tl tt I caus. ltd I I Idl Ul Ul I caus. pass. ltd I _ Idl Itl Ui 3.55. All stem initials have been included in the table so that the effects of assimilation may be compared in various settings. It will be noted that d — either agentive classifier, or final consonant of 32 NAVAHO GRAMMAR 3.55.-3.64. first person dual pronoun — is absorbed by b, d, t, t\ g, k, #, dz, ts, ti, dj, tc, t6, dl, tt, and tl. dz, dj, and dl may be stem initials without ^-influence, but when d meets any of them, there is no change in the affricate, nor does d persist. Since all affricates have either ad or t attack, they behave like d and t, that is, they are unaffected by preceding d. 3.56. Examples of consonants changed by contact with d follow; they include some of affricate initials to illustrate overlapping between two sounds which become an affricate, and the absorption of d by an affricate : 3.57. d- > t: yi'fac (< yi-d-'ac) "we are going prog.;" yi'tic (< yvd-ic) "we are stringing beads ;" yitic (< yi-d-ic) "they are being led (on a string);" nvUrc (< nrd-'i'c) "we 2 are starting to lead them; we-2-attached-by-a-string-are-starting-for;" yvfol (< yvd-ol) "we 2 are floating prog." 3.58. d-m > rrt: yvrtiqs (< yvd-mqs) "we 2 are rolling a sphere;" H'dvnlal (< 'vdvd-mal) "we 2 will gulp it down" 3.59. d-n > n': drriah (< drd-nah) "we 2 shall crawl;" di'rifrh (< di'd-ne'h) "we 2 are starting to crawl;" nvriil (< nvd-nil) "pi. obj. have been laid down;" yvriil (< yi-d-nil) "we 2 are carrying pi. obj.;" yi'n'ih (< yrd-nih) "we 2 are milking;" nineiriih (< ninei- d-nih) "he cust. distributes them;" bi'tb'riij (< bi'ad-yo'-d-nij) "it has been plucked by someone" 3.60. d-y> if (exceptional): (6- 'dxoni-yoi (< 'dxoni'd-yoi) "we are increasing in number" (YM 234); xonvyoi (< xonvd-yoi) "we 2 are brave, good at . . ., we 2 excel;" dini'yog (< dini'd-yog) "we 2 are fluffy" 3.61. d-y> d (exceptional): yvd4(< yi'd-yq) "we 2 are eating it;" yidq,'* (< yi-d-yq*') "it has been eaten" 3.62. d-y > dz: yvdzol (< yi-d-yol) "we 2 are blowing;" yidzol (< yi-d-yol) "it is being blown;" nidzd'd (< ni-d-yfrd) "several are being driven" 3.63. d-y > d (exceptional) : di'df'l (< dvd-yf'l) "we 2 will eat it" 3.64. d-y> g: dvgfrl ( d: no-dfrz (< n dz: yidza*z (< yi-d-za-z) "it has been snowing;" nrdzas (< nvd-zas) "we 2 are sprinkling it in a continuous line;" 'anddzi' (< 'ana-d-zi') "he oust, rakes;" yrdzoh (< yvd-zoh) "we 2 are marking it;" yvdzi'l (< ^t^z£*£)"we2arecomingtoastandstill;" 'adzfrs (< 'a-d-zfrs) "something is being singed;" do'dzoh (< do*-d- zoA) "it is being carried in the mouth" 3.67. d-s (< d-l-z) > Is: di'lzah (< dvd-l-zah) "we will find it gone;" dvlsas (< dvd-l~zas) "we are strewing it in a line;" nvlse'l (< ni-d-l-zfrl) "we 2 are growing up prog.;" srlsi'h (< svd-l-zvh) "we have missed the mark, made a mistake" 3.68. d-dz > dfe : yi'dzf's (< yi'd-dzls) "we 2 are dragging it prog. ;" xadrdzih (< xadvd-dzih) "we 2 will speak out;" dvdzih (< dvd-dzih) "we 2 will be left, will survive" 3.69. d-? > d?: yido'djW (< yido--d-j{l) "he will be blackened;" 'adidi'djah (< 'adidvd-jah) "we 2 will spit;" bi'fodjih (< bi^ad-yo-d jih) "he is being named, called by name;" yrdjoh (< yvd-joh) "we 2 are combing it" 3.70. d-c (< d-£-?) > fc: dvlcih (< drd-l-jih) "we will mow it, cut it (as hair);" yvlcic (< yvd-l-jic) "we are poking it with slender obj. (as stick);" yi'lcf*' (< yvd-l-jf*') "we have blackened it " 3.71. d-dy > dj: yidrdjfl (< yidvd-dji'l) "we 2 shall be black- ened;" si'dje*' (< si'd-dje^) "we pi. exist;" 6a* dvdjd'h "we 2 are giving him wood" 3.72. d-l > I: yi'ldjq (< yvd-l-djq) "we 2 are stamping along;" yi'ldlal (< yvd-l-dlal) "we 2 are ripping it prog.;" yrlgic {< yi'd-l- gic) "blade cutting is being caused by us" 3.73. d4> dl: yrdloh (< yrd-loh) "we 2 are looping, lassoing it;" yi'dWs (< yi'd-16's) "we 2 are leading one along on a rope;" svdlf'' (< si-d-ty') "we 2 have become;" naxadld (< naxa-d-ld) "ceremony; things-are-being-done-in-order" 3.74. d-l > i: yt-Zd-Z (< yrd-l&l) "we 2 are whistling, singing in a high key;" yt'krl (< yi'd-la-l) "we hate him;" yi'Vwl (< yvd-l-wl) "we are sending him on an errand;" yvltas (< yi'd-l-tas) "we are twirling a small obj.;" yvltsos (< yi'd-l-tsos) "we 2 are moving fabriclike obj.;" yvlzi'l (< yvd-l-zi'l) "we 2 are blessing it;" se-lyin (< srd-Z-a:/ m) "we killed him for a fact" (EW 112:2) 34 NAVAHO GRAMMAR 3.75.-3.81. 3.75. d-dl> dl: yi'dty ( rri, d-y > ]), d-y > d, and d-z > d ; all others are quite common. 3.76. A study of the chart with prefixed I shows assimilations or changes with y, y, and s, and with the laterals. The general rule is that d-y > dz t d> or y\ l-y > s, and l-y > ly, or Iz. Probably no verb stem has a primary initial s ; stems with ^-initial are equivalent to l-y, or Iz. Examples of changes with ^-classifier follow: 3.77. l-y > s: xasd-' (< xa-l-yd*') "deserted, abandoned place;" naxasd (< naxa-l-yd) "he has caused ceremony to start;" yisqrd (< yi-l-yqrd) "he has her affection; she loves him;" yisil (< yi-l-yil) "he has grabbed it" 3.78. l-z > s\ bil 'i'sa'l (< H'-l-zwl) "he sailed off in it (car);" yi'do'sq/l (< yrdo'-l-zq/l) "she will love him;" niriisq (< nirii-l-zq,) "it grows;" yise'h (< yi-l-zi'h) "he is making it pliable;" yo'sas (< yo'-l-zas) "he is strewing it in a line;" yosih (< yo'-l-zih) "he is causing sharp obj. to move swiftly" 3.79. The causative I unvoices the blade alveolar as it does the alveolar sibilant : l-j > c: yiyvce' y (< yiyv-l-je^) "he has sheared it, he has caused cutting of woolly, grassy material;" neic Ix: 'adilxdx (< 7 adi-l-ydx) "they are biting something;" dzo'lxal (< dzo'-l-yal) "he is twirling a clublike obj.;" 'arivlxi (< 'arii'-l-yf) "I have killed someone, something;" dilxi'' (< di-l-yi'') "it is thawing;" yil-xod (< yi-l-yod) "he caused oscillation;" yilxoj (< yi-l-yoj) "he caused tickling;" yi'ilxd-c (< yi'i-l-yax) "she is putting it to sleep" 3.81. The same principle operates for the laterals preceding laterals as for d, that is, when one sound precedes a stem with the same initial, it assimilates to it ; the articulation is not doubled : 14 > I: xonild*' (< xonil-W) "I have had a ceremony started;" di'lid (< di*-l-lid) "he burned it;" yiyvlqrd (< yiyv-l-lqrd) "he has increased it, caused it to increase" 3.81.-3.84. phonology 35 Of the three classifiers I is the most stable. It sometimes changes 7« fn / hut, apparently the stem complex with Zn-initial is closely related to that with Z-initial. y and y as stem initials seem to be related to the laterals. Since these relationships are doubtless of historical derivation, and since the number of stems showing these changes is small, they will be discussed elsewhere (8.102, 8.103.). A single example of different construction is an exception to the rule that laterals in juxtaposition are not doubled: 'di ciye*l-le'l (< ciye-l do-k'l) "that will be my offering" (NT 116: 13). 3.82-3.97. -c-first personal pronoun 3.82. A common assimilation concerns the alveolar and blade alveolar sibilants, which may have reciprocal effects. The first personal pronoun -c- causes various changes of stem initial. It assimilates the classifiers I and I except in the perfective where its position may make it a test form. The changes with -c-first person are listed in Table II. Since -c- sometimes absorbs the classifier, or causes a change in the consonant cluster, such changes are also included. Table II 2/-prefix y -stem V z 8 1 c -c-1 subj., ag. a ■s- 3-3 pf. 8, y h final h 8 8 hy, ha c ay x, ha a 8 8 a* ha c cj c c c h, Z-eaus. /-stem ly Iz Ix I 11 -c-1 subj., ag. c 5- 3-3 pf . 8 ~h final I cl zl, al I cx a ly az az, a Iz cx sx Ix c 8 Ix cl si I * a may result from a combination of -l-y or -l-z. -c- has the same effect on both. The combination -l-y- > -a is exceptional, being found only with -yfrl "eat," whose stem initial is irregular and may perhaps be y. ** a may result from -l-y- or -l-z- ; the effects of *-pf. with the a or either of these derivatives is the same. 3.83. c-y > s: xonisq, (< xoni-c-y4) "I am aware of, wise about things, careful of it;" bixodesah (< bixode*-c-yah) "I miss it, I find it gone;" dine-sol (< dinex-yol) "I shall drive several" 3.84. c-y > c (exceptional, stem initial not clear): ba- yicdh (< yie-ydh) "I caught up with him;" na-cdh (< na-c-ydh) "I am going about;" bvcd'h (< brc-yd-h) "I am going up (down) along it" (cp. binicyah "I am capable of;" biyah "he is capable of, it suffices" (YMG 25-6) 4 Seicbard 3 6 NAVAHO GRAMMA& 3.85.-3. 9J g 3.85. c-y > ex (regular): dide-exdx (< dide-c-ydx) "I shall nibfy e it;" yiexqh (< yic-yqti) "I am killing them;" 'Miie-sxas (< l aMiex- yas) "I will scratch myself;" 'rcxa' (< 'i'C-ya') "I am shaking flexible obj. ;" a:a*ca:e'A (< ascrc-ye'A) "I am taking ropelike obj. out" 3.86. c-z > «: ye'sis (< yex-zis) "I am singeing it;" yisoh (< t/ic- zoA) "I am marking it;" yisgs (< yic-zQ-s) "I am tearing it (as fabric)" 3.87. c-s (< l-y) > 5: s/rsoZ (< yic-l-yol) "I am rep. blowing on it;" bv 'aso'l (< 'ac-l-yo-l) "I am pumping air into it;" yisas (< yic-sas) "I am sprinkling it in continuous line prog." 3.88. c-s (< l-y) > 5 (exceptional): bi* diye'si'l (< diyex-l-yf'l) "1 shall feed him, force food into him" 3.89. c-s (< l-z) > ^: yiM (< yic-l-zin) "I am blessing it;" yvskl (< yex-l-ze'l) "I am dressing hide;" y azdiye*sih (< 'azdiyex-l-zih) "I shall throw sharp obj. beyond rep.;" ncsoh (< nex-l-zoh) "I am marking it" 3.90. c-/ > c: de-cah (< dec-jah) "I shall spit;" yide m cvl (< yidex- jH) "I shall call him by name;" yico-h (< yic-jd'h) "I am combiing its hair;" ^tcic (< yic-jic) "I am breathing it in" 3.91. c-c (< 2-/) > c: yici'h (< yic-l-ji'h) "I am cutting strands, I am shearing, mowing it;" ^iepe (< yic-l-jgc) "I am throwing hoop- like obj. ;" yicg* (< yic-l-jg-) "I am taming it, breaking colt" 3.92. c-Z-caus. > c: nacnic (< rvac-l-nic) "I am working;" 'dxdcyq, (< 'dxdc-l-y$) "I have sense;" nax'a? (< nax-l-'a') "I am being sent on errand;" xexyal (< xex-l-yal) "I am giggling on belly;" dexyal (< dex-l-yal) "I am eating meat;" 'ddicje-h (< 'ddic-l-jfrh) "I am shaving;" dinicyo' (< dinic-l-yc?) "I run slowly;" 'oc^c (< 'ac-l-jic) "I am dancing;" 'ddicjo-h (< 9 ddic-l~jo-h) "I am brushing myself, combing my hair" 3.93. c-Z-stem initial > cZ: wrf£ (< 7w;-Z/) "I am;" yicte'h (< yic-le-h) "I am becoming;" 'dxlvl (< \ac-lvl) "I am creating, making it;" yicli'l (< yic-U'l) "I am carrying a ropelike obj., a pair of obj.;" naxonclin (< naxonc-lin) "I look like him, I resemble him" 3.94. c-fe > 52: 'dxodideszih (< 'dxodidex-l-zih) "I shall become motionless" (YM 239); 'dkdszis (< 'dkdc-l-zis) "I am putting on belt" (YM 243) 3.95. c-Z-caus. > c: yisbqs (< yic-hbqs) "I am driving it (car, wagon); I am causing it to roll prog.;" xode'cbjrl (< xodex-l-bi'l) "I shall build a hogan;" xadicbin (< xadic-l-bin) "I am filling it; 3.95.-3.99. pflONOLOGY 37 I am causing filling;" xanictca'd (< xanic4-tca*d) "I am carding wool ; I-am-causing-it-to-swell-out M 3.96. c-l (< l-l) > cl: dide'clil (< didex-l-lil) "I shall cause it to smoke, burn;" ndxide'dah (< ndxidcc-l-lah) "I am choosing, selecting them;" yield (< yic-l-h'l) "I am becoming" 3.97. c-lx (< l-y) > ex: 'acxoc (< 'ac-Z-yoc) "I am sleeping;" yiexoj (< yic-l-yoj) "I am tickling him;" 'adiexd-c (< 'adtc-Z-ya'c) "I am biting something;" dicxal (< dic-l-yal) "I am opening my eyes" 3.98-3.111. 5t-perfective 3.98. The formulas for combination of d, I, and Z, and c are essentially the same as Hoijer's, stated in somewhat different terms. However, my analysis of si-perfective differs greatly from his. 15 si- seems to be a persistent prefix of the perfective combined with the completive inflectional -ni- which may be separated by the personal pronouns in the intransitive and active transitive (10.55, 10.107.). The position of these two prefixes in the third person active transitive, where si- is pushed toward the stem by yi-thxrd object, gives rise to the phonetic effect of voicing — si- in this position becomes -z-. This process enters into the matter only when the zero stem is used ; whenever the stem is affected by a classifier (d, I, or I), the prefix of the third person perfective is si- or -«-. It seems clear, therefore, that basically si- or -s- is the stable form, and that -z- is to be explained on the basis of position. 16 si-perfective may therefore be analyzM like other prefixes, noting first the effects of -8- on the stem and stem complex, and explaining -z- as a prefix complex, somewhat unusual, but paralleled by other prefix combinations (10.59.). 3.99. -s- before a stem consonant has an effect comparable to that of -c-first personal pronoun, but the alveolar and blade-alveolar positions are reversed, that, is, the alveolar sibilant -s- assimilates the blade-alveolar to it, or the blade-alveolar sibilants become alveolar sibilants. If -s- precedes a stem -z- or -s- initial, they may merge and only one s results. If s- or -z- precedes a stem with a blade alveolar, -s- changes to -c- and -z- changes to -j-. Such a -c- 16 Hoijer 1945c, pp. 19-20, 43-8. 16 Hoijer considers -z- as "augmented by d, I, or V* (Ph 43). I consider the stem "augmented 11 by the classifier, to use his terminology, as usual, and -z- the exception because of the position of si-. In my terminology -z- occurs before the zero classifier, -s- before all the others, or before all stem complexes. 4* 38 NAVAHO GRAMMAR 3.99.-3.106. assimilates to -c-initial, again leaving only one -c. Just as -c-first person pronoun assimilates to I and I, so does s-perfective, but either can assimilate only one lateral, and if a stem complex is composed of a causative and a lateral stem initial, the second stands, but is unvoiced by s, that is, s~l > s, but s-l-l > si. s-c(< l-j) > c, but 8-l-j > cj\s< l-z or < l-y > 5, but s-l-z > sz. The unstable initials y and y present a few exceptions as they do for -c-first person. Table II indicates changes brought about by si-perfective third person subject with a third person object. Examples are as follows : 3.100. s-y > y (exceptional): de*yd (< desyd) "he started from;" xode-yd (< xodes-yd) "short singing" (cer.); 'ane'yd (< 'ane m s-yq) "maturation; something has matured" (YME 54) 3.101. s-z > z: de-za' (< des-za') "he belched;" nde'zid (< ndes- zid) "time passed;" nvzq, (< nes-zq) "he is wellbred" 3.102. 8-8 (< l-y) > s: xo'sa* (< xo-8-l-ya?) "he missed it, found it gone;" ninisq, (< nini-s-l-yd) "he grew up" 3.103. s-8 (< l-z) > s: yixosa 9 (< yixos-l-za?) "he missed it;" yidesas (< yides-l-zas) "he sifted it;" yisvh (< yis-l-zvh) "he missed target, he made a mistake;" yisil (< yis-l-zil) "he grabbed it" 3.104. s-c (< l-j) > c: naxac&h (< naxas-l-jo-h) "he has swept a place;" yide-cd'd (< yides-l-jo-d) "he has dragged a fabriclike obj.;" nSicoh (< niis-l-joh) "he moistened it" The same change takes place in an entirely different setting: di-gicf' (< drgis-cf') "it must have been twisting" 3.105. s-Z-caus. > s: xa'asde*' (< xa'as-l-de*') "some group moved up out;" kina'sda (< kina'8-l-da) "she has menstruated for the first time;" ca- 'dxo-syq (< J dxos-l-yq) "she took care of me;" 'ddisyaz (< 'ddis-l-yaz) "he scratched himself;" yicyi'j (< yic-l-yi'j) "he is in a crouching position;" nesyal (< ne-8-l-yal) "he threw himself down;" desyis (< des-l-yis) "he dodged (a blow);" tsd'dszi' (< tsffaS'l-zi') "yucca; main part is fibrous;" tarie*szani m (< ta'anes- l-zanv) clan name; naxjah (< nas-l-jah) "he went hunting" 3.106. 3-Z-stem initial > zl: bidanvzlah "they are tied here and there;" na-zlf "it flows about;" na-zlo* "he moved loop, lasso about;" biddne'zldh (stat.) "they are touching (as branches of a shade or corral);" ba* 'ayaxo-zlv' "he suspected him" (YM 133); yizlih "he tasted it" (YM 135); xazty' "things have become;" dd-dlce xa-zld "oblong field" 3.107.-3.111. PHONOLOGY 39 3.107. 5=24=stem initial > si: yisldh (< yis-l-ld) "he has it (light- ning arrow);" citte'esla (< ciUi'as-l-la) "he has a claim on me;" nidjo-sld^ (< nidjo-s-l-W) "he hated you" 3.108. s-l-x > sx: yisxi (< yis-l-yf) "he killed one;" yisxal (< yis-l-xal) "he clubbed it" 3.109. Voicing a consonant as in yiz-3-3 si-perfective is not an isolated phenomenon. A comparable form is djilgai "he (4) is white," in which I is not passive causative but Ji-natural (10.124.), voiced because of its position in the complex : li-dji-, or dji~li- > djil-. The process may be related to that in which certain nouns with voiceless initials take possessive prefixes and voice the stem initial (5.9.). When yi-3 object comes in contact with a xi-prefix the result is yi- (10.114d, e.) or, with a more complex combination of prefixes, yo- or yo'- (10.109.). The voicing of si-, like other processes, is a function of the selection and position of the several prefixes rather than of the particular perfective or person. 3.110. The combination of si-perfective with the first person subject -c- is another point of difference between Hoijer's analysis and mine. I analyze the form se~ of sitf "I exist as an animate obj." as si- < si-pf.-c-l subj.-m-completive. This form illustrates the importance of e as a combined form, the effect of the combination of sibilants si- and -c-, and of position, which differentiates the pronominal prefixes of the active (first subject) and the passive (first agent). In comparison with the formula just given the passive has the form sis- < 5i-pf.-m-c0mpl.-c-l ag. ; c> s by the general rule of sibilant assimilation. The final s or c of si-s-, which may become cic- if there is a blade alveolar in the stem, follows the same rules given for -c-first personal pronoun (3.82-3.97.). 3.111. si-perfective has yet another effect, which really belongs with contraction rather than assimilation, but since one phonetic process interacts with others, all perfectives are exceedingly com- plex in structure. The effect is of the fourth person pronominal prefix dji- whose position is as near initial as possible in the con- jugation (6.19.). The formula of the fourth person si-perfective is dji-4: subj. (or ag.J-s-pf.-m-compL, and it contracts to dzi- in the intransitive. Here we have an assimilative change from dji- to dzi- because of dj plus s. The transitive active form of the zero stem is dziz-, that of the d-, l~, and I- forms is dzis-. The conclusion to be drawn seems to be that -z-, the sonnat, is the "aberrant" form in need of explanation, not -s- which retains its identity in some form and dominates other prefixes with which it comes into contact. The analysis is compatible with that of other 40 NAVAHO GRAMMAR 3.111.-3.116 even more complicated forms, for instance, dzi-ajw&y §fi& Wjv- attitude, in the paradigms of which dji- and dzi- shift about accord- ing to the position each occupies in relation to other prefixes (10.121-10.121 d.). 3.112-3.133. Finals 3.112* Final h of prefixes affects certain consonants that follow in several ways: A voiced consonant may be unvoiced by h. A sound may be changed entirely, as y to s, by the prefix final, either with or without the persistence of A. The changes are summarized in Table II (p. 35). 3.113. Final h of prefixes assimilates to following y and the sibilants in a manner consistent with the rules already given. Prefixes most commonly entering into combination ending in h are: O'oh "fall short of, be less than, just miss being. . . ;" dah- "forth, forward; suspended;" and -oA-second person dual pronoun. dah- is a pre-paradigmatic prefix and does not occur in a position of conjugation, that is, it must be followed by other prefixes which have a position nearer the stem or stem complex. Examples of changes brought about by h are : 3.114. h-y-j>Tefix initial < h: dahilteos (< dah-yiltsos) "you are holding fabriclike obj.;" daho-ltsos (< dah-yo-ltsos) "you 2 are holding a fabriclike obj.;" dahvte 9 (< dah-yvfe') "he started to run forward;" dahe-z'q, (< dah-yiz'q) "he has suspended a round obj." 3.115. My material, from both ^-speakers and others, has final h of a prefix persisting before s: 17 dahsild "ropelike obj., pair of obj. lay upon (a shelf)" (EW 106:9); dahsity "narrow rigid obj. is suspended, lies on top of...; there is a crescent moon;" do* 'ahsoxodo'be'jdah "things are hopeless, there is no hope;" tsin bq- dahsa 9 ^ "apple (fruit) is hanging on tree" (FH). 3.116. Final A of a prefix unvoices a following voiced consonant, usually a sibilant: dacdcyd (< dah-dji-de'yd < dah-j-de m y&) "he (4) has started to go forward;" datsiztj (< ddh-dziz-ti) "he(4) is lying on top;" bo'ocne*r4 (< bVoh-dji-nt'Vq) "he(4) cannot afford it, he falls short of it" (YM 10); 9 dlacdo-le-l (< 9 dlah^-do'U'l) "they(4) will assemble" (EW 106 : 9) ; 18 bq-cde-yd (< bq*h djide-yd) "he(4) passed them" (NT 54:23); bixno'td (< brh djino-fy) "he(4) put his hand into it" (NT 78:17). 17 Cp. Hoijer 1946c, p. 39. 18 Cp. Haile 1938, p. 248, n. 43. 3.117.-3.126. PHONOLOGY 41 3.117. Final h of -oA-second dual pronominal prefix has effects comparable to those of final h of daA-forward, suspended, but since it occurs in juxtaposition with the stem or stem complex, its mani- festations are more extended. Generally it unvoices a consonant, but like c and s y it has other assimilative effects: 3.118. -oh-y-stem initial > -o%-(exceptional): to* 'axonahyoi (< to- 'axond-oh-yoi) "many of you;" 'ohyo'l "you 2 are inhaling, taking a breath" (YM 234) 3.119. -oh-y-stem initial > -ohs- or -os- (exceptional): 'ohsq, (< 'a-yi-oh-yq) "you 2 are eating something;" xonosq (< xoni-oh-yd) "you 2 are wise;" dinohsd'd (< dinoh-yo'd) "you 2 are driving a few" (YM 233) 3.120. -oh-y- > -ox-: yoxd-d (< yoh-yd'd) "you 2 are shaking fabriclike obj.;" do-xas (< do-h-yas) "you 2 will claw it, scratch it with nails;" 'o-xeh (< 'o-h-yeh) "you 2 are being married" 3.121. -o^-y-stem initial> -ohs- (exceptional): do % hsfl (< do'h-y('l) "you 2 will eat it" 3.122. -oh-z-stem initial > -os-: didosah (< dido-h-zah) "you 2 will belch;" ndo'sil (< ndo'h-zil) "you 2 will rake them together;" nosi'h (< noh-zi'h) "you 2 want it;" yosi'h (< yo-h-zi'h) "you. 2 are coming to a standing position;" bitdqh xo'soh (< xo'h-%oh) "you 2 draw a line of protection" (EW 110: 12) 3.123. -oh-s- (< l-z) > -0A5-: yoksis (< yoh-l-zfrs) "you 2 are singeing it;" bixodo-hsah (< bixodo-h-l-zah) "you 2 will find it gone;" dohsas (< doh-l-zas) "you 2 are sifting it;" nohse'l (< noh-l-zi'i) "you two 2 are growing up" 3.124. -o^'-stem initial > -oc-: xodido*cah (< xodido'h-jah) "you 2 will spit;" yido'ci'l (< yidd'h-jvl) "you 2 will call him by name;" yoco& (< yoh-joh) "you 2 are combing it" # 3.125. -0&-c (< £-?') > -o/rc-: do-hcih (< do-h-l-jih) "you 2 are mowing it, cutting strands;" yo-hcic (< yo'h-l-jic) "you 2 are poking it (with a stick);" yo-hci'h (< yo'h-l-ji'h) "you 2 are blackening it;" naxohco'h (< naxoh-l-jo'h) "you 2 are sweeping a place;" yohcg'h (< yoh-l-JQ'h) "you 2 are breaking a horse, taming it" 3.126. -o^-Z-pass. caus. > -ol-: dolde-h (< doh-l-dfrh) "you 2 are starting with a group;" yaKoldjo-l (< yah'a-oh-l-dj&l) "bunchy substance is being carried in by you 2;" ba- 'dxolyq> (< ba r 'dxoh- l-yq) "you 2 are taking care of it;" do-lzih (< do*h-l-zih) "sharp obj. will be hurled by you 2 ;" ncrlje'h (< na-h-l-je'h) "you 2 are hunting" 42 NAVAHO GRAMMAR 3.127.-3.133. 3.127. -o^-Z-stem initial > -ol-\ 'oh'h (< 9 6h-le'h) "it is being done by you 2 ;" yolqrl (< yoh-lqsl) "you are increasing, becoming many;" xonoli (< xonoh-li) "you exist;" naxonolin (naxonoh~lin) "you look like, you resemble;" yoloh (< yoh-loh) "you 2 are roping, lassoing it;" yo&rd (< ydh-lo'd) "you 2 are lapping, sipping it" 3.128. -oh-l-y- > -oly-: yo'lyal (< yo-h-l-yal) "you 2 are eating meat, tearing meat from bone;" yinolyi (< yinoh-l-ye) "you 2 are called, have the name. . . ;" J adin&lyil (< J adin6h-l-yil) "you 2 will doze;" xadido'lyoc (< xadido-h-l-yoc) "you will shout" (YM 86) 3.129. -oh-l-z- > -o2z-: 'akaso'lza'z (< 'dkdso-h-l-za-z) "you have your belts on" (YM 243): sodo'hin (< sido'h-l-zin) "you have prayed" (YM 242): ! } dxodido-lzih (< 'dxodido-h-l-zih) "you will calm down, become motionless" (YM 239) 3.130. -oW-caus. > -o2-: yah'oldjd'l (< yah'a-oh-l-djfrl) "you 2 are carrying bunchy substance in;" 'dldi'h (< 'dh-l-di'h) "you are destroying it;" ydlti' (< yah-l-iV) "you 2 are talking" 3.131. -oh-l-y- > -olx-: do'lxal (< doh-l-yal) "you 2 will club it;" dcrlxe-l (< do'h-l-ye'l) "you are calming down;" ndlxes (< nah-l- y&8) "you 2 are turning it around;" nmiolxod (< narioh-l-yod) "you 2 are lame" 3.132. -oA-Z-caus.-Z-stem initial > -oZ-: didolil (< didoh-l-lil) "you will make it burn, smoke;" yoZa/A (< y6h-l4qrh) il yo\i are increasing in number;" yoh'h (< yoh-l~h'h) "you are creating, making it" 3.133. The influence of h as a stem final is helpful in reconstructing stems whose endings are doubtful, h is one of the few consonants that may end a stem (-CVC), but it is sometimes so feebly articulated that doubt is left as to whether the stem is -CV-zero or -CVh. A clue to this problem may be found in the suffixes which are freer than prefixes or stems. Such suffixes as the nominalizers, -r, -£, and -igv, -6* "custom, way;" -e-' "future subordination ;" |" "past, afore- mentioned;" -»•' "after having. . . ;" -ic "interrogative" tend to be attached in a manner that indicates the stem final. If it is zero, the suffix vowel may cluster with the preceding vowel, or the suffix may have the initial y, really a glide consonant. If the stem final is h, the suffix seems to be hV, and if the stem is an open syllable with a nasalized vowel, whether or not it is pronounced, the suffix has the form -nV. If the stem has any other consonant final the suffixes have the forms mentioned. A few examples follow: 'dhi "fog, mist;" tsehpcf' "surely he must have referred to a rock ;" ntm-higr "the one that is wide, large;" cddiyp "my deceased older sister;" 'akone' 9 (< 'aky-e-') "you'll see, it will happen so;" fa* do- 'odinini "don't say 3.133.-3.136. phonology 43 that, don't let him rpaa.V thu»" (NT 136:23) (cp. 'ddini "you speak thus"). The process here referred to is not always consistent, but I believe it may prove a helpful device to differentiate some stem finals and suffix initials which may be of aid in historical reconstruction. For example, I sometimes hear xa'dff for xa'dt'i "whatever." AB, despite the fact that he is an n-speaker, protested that xa'qtf was "wrong," yet we find xa'dfrne*' "whatever it may be in future." At the very least, the forms pose a problem which, when properly worked out, may yield useful results about stem structure. 3.134—3.135. Tone Change and Assimilation 3.134. The relation of nasality and n to tone has been indicated (3.49-3.51.). Tone change is not limited to the nasals, but, as we have seen, may be concerned with si-perfective (3.98.). It is a question whether the survival of consonants in a high tone belongs under assimilation or contraction, but it is mentioned here to account for some forms already discussed and others to be encountered later (10.117-10.118 g.). 3.135. Although I do not agree with Hoijer that "inherent tone" has been determined, it is obvious that certain elements have a more dominating effect than others. It seems, however, that such dominance can be accurately expressed only in the relationship of one prefix to another ; it is difficult to see how it can be absolute. For instance, yi-3 object has a low tone which dominates many other prefixes in its vicinity, and yi-S passive subject dominates in a different way. Each derives its strength from its origin and, possibly, from its position at the front of the verb paradigm. Another example is dji-4k subject, whose tone is less dominating than dji-4 agent. Neither is "more or less" dominating, but each has great power to affect other prefixes in its vicinity (cp. 10.55, 10.90a.). 3.136-3.140. Interrelation of Phonetic Processes 3.136. If I seem to reiterate the importance of interrelationships, it is because they obscure, as well as clarify Navaho grammar. One vowel is related to others and influences them retroactively ('a- ? a- > HH-, Vo-, etc.) (3.30, 10.76 b.); vowels and consonants are related ('a-n- > '{-, y q>- > -an, or -a*n\ n> - or r; -a'a-di- > -a'ti-); some consonants change in contact with others (-l-z > -s-; -l-j > -c-, -h-j- > -c-); alveolar sibilants become blade alveolars and the reverse. Tone, which is thought to belong to vowels, nevertheless is 44 NAVAHO GRAMMAR 3.136.-3.139. affected by consonants (n, s, c). As the prefix paradigms show (10-10.124.), length may stand for a vowel or a consonant, or both since several syllables of the primary form CV may persist only in a lengthened vowel. And further, many processes of assimilation and contraction operate at the same time (dji~ > -/- and may be affected by preceding Ca'a- so that the result is CVj'C; yi-si-ni- > yiz-\ and yi-si-ni-l- > yis- ; or yiz-with a blade alveolar > yij-)- 3.137. Contraction, involving shortening of the theoretical form CV to C, lengthening, change of tone, lengthening and change of tone; voicing, unvoicing, and glottalization of consonants, distorts any attempt at a realistic theory of the syllable. Probably Navaho in particular, and Athabaskan in general, may require the revision of many so far accepted definitions, (cp. 10,35-10.46.). 3.138. The greatest difficulty, in the light of what seem to be conflicting results, is in arrangement. If a stem with initial z is found in the form s, or a prefix combination Va- appears as Ve-, 'i'i-, or Vo-, how is either to be found by the reader of Navaho ? If a prefix si-harm is present as -cr- or yo'-, how is it to be indicated so as to be understood ? The compilation of any lexicon and the presentation of any grammar present many problems, but those posed by Navaho phonetic interrelationships have not been squarely faced. In this work they have been indicated by different arrangements. I have tried to remember that the reader has before him only one form ; that he should be able to find the form in the grammar and lexicon as it appears in the context he is reading. By constant repetition — of form, of variation, of interrelationship — and by numerous cross references I have tried to indicate the relationship between form and vocabulary. 3.139. As we shall see when analyzing, there may be various reasons for any one form — overlapping is a hazard which may be handled only if relationships are constantly kept in mind. Over- lapping of form is due to the fact that a few consonants and vowels have been overworked, distinguished by the to us unfamiliar pro- cesses of variation of quantity, tone, nasality, glottalization, and the like. Furthermore, there are what almost seem to be infinite sub- divisions of an idea. For instance, one "starts a motion from a point," but he also "starts forward." The first idea is understand- able as an inceptive, the second as a double inceptive, but Navaho does not stop with these. It also adjusts a start so that it may be progressive, continuative, or completive, and such a start may be interrupted by pausing or by "getting stuck." All these ideas con- cern a prefix di~ which is relatively simple compared with prefixes of form ni- or yi-. 3. 140. PHONOLOGY 45 3-140. And not only do we have splitting of ideas differentiated by intricate processes, but we must also deal with phonetic diversity in the population. Diversity involves not only the history of Atha- baskan, but also the extreme tolerance of adaptation characteristic of the Navaho, a cultural, perhaps a psychological development. Besides the task of unraveling the numerous relationships of forms as they stand, we are obliged to discover the limitations to which any one form is restricted. They may be ascertained by continual comparison, but the comparison must be held down to similars, especially in the same series, that is, in the paradigms. ^Although the conditions posed by the language as it now exists may seem difficult, they have not by any means resulted in chaos, nor is their disentanglement hopeless. It is likely that the rules and formulas here presented will be greatly revised and simplified. If so, such simplification must take note of the elements that go into the shortened forms, instead of combining unrelated elements and meanings in classes too simplistic for significance. 4.-4.36. THE WORD 4. In their discussions of other Athabaskan languages Sapir and Hoijer assume that the noun is the primary form, and that verbs derive from it. They are careful, however, to note that the question is a large and intricate one which cannot be settled by any one language alone. 1 Athabaskan languages, such as Mattole, Hupa, Kato, and others, seem to be much simpler in form than Navaho and would, therefore, point to Sapir's conclusion which is accepted by Hoijer for Navaho without question. Navaho seems to be a great melange of various Athabaskan elements — nominal, verbal, adverbial, or independent. It is still too early to determine the original form of the word because of the free- dom with which these elements combine. Nevertheless there are suggestive clues, and it should be remembered that Navaho is far removed from primitive or original Athabaskan. When he reconstructed the history of Athabaskan Sapir derived the basic verbs of motion and state from a few nominal stems. 2 Certainly he was justified in this procedure since the basic verbs of motion with ■ their corresponding static forms are found in all languages for which we have examples. Related nouns, however, are even with our presentday extensive vocabulary, relatively few, and the process of derivation, from verb to noun, or the reverse, is not by any means clear. On the other hand, an interrelation between the so-called parts of speech is quite obvious. Since it is by no means limited to noun and verb, but includes elements such as post- positions, which are as basic and "primitive" as nouns and verbs in all the languages, and since postpositions become prefixes, tense- aspect elements, as well as nominal and adverbial prefixes, all these relationships will be discussed here. Many of them pose far-reaching questions as to what a noun, verb, or adverb is, not to speak of the reasons for the particular forms in Navaho. 4,2. The data seem to me to point to a verbal origin as more basic than the nominal. My conclusion is derived from the close phonetic and morphological relationship between the different parts of speech. 1 Sapir 1923; Li 1930a, p. 62; Goddard 1910, pp. 107ff.; 1912, pp. 19ff. 2 Sapir 1923. 46 4.3.-4.6. ts:e word 47 4.3. Grammatically noun, pronoun, verb, adjective, and postposi- tion may be differentiated. Each is treated in a distinctive way, but there is much overlapping. 4.4. The theory that the noun is the original base probably origin- ates in the large number of monosyllabic stems, many of which are phonetically similar to verb stems. Such monosyllabic stems may be used independently, whereas the simplest verb is a compound of a stem with at least one prefix. I know of only a few Navaho verb forms without a prefix. However, the primary meaning of nouns, pronouns and post- positions, and other elements seems to be verbal; so common is the verbal meaning of the nouns, pronouns, and locatives that a great deal of idiomatic communication may be carried on without any verbs whatsoever. The translations "it is a garment" fe-'), "it is a flint'* (bfrc), "it is my mother" (rimy) seem much better than "garment," "flint," "my mother." Similarly, "it is I, I am the one" (ci) y "it is mine" (ci' y ) "it is for my benefit" (cd), "it is with, by means of it" (6e*), "it is over him" (biki), are better renditions of Navaho than "I," "mine," "for me," "with it," "over him," respectively. 4.5. Possessed nouns, that is, noun stems which seldom occur without a possessive prefix, are an outstanding feature of Atha- baskan, although the absoluteness of the possessive requirement has perhaps been overstressed for Navaho. Some nouns, particularly those referring to body parts and kinship terms, usually have the possessive prefix, but such nouns occasionally occur without it. Perhaps to be explained by poetic license is the independence of body part nouns mentioned in songs; they occur, however, after a series of similar nouns with the possessive prefix. 3 Poetic license is not the sole explanation, however, as the follow- ing examples indicate: ke didilye "Moccasins-are-laid-in-the-fire" (place name) (NT 32:14); kehi ridzo'fe*zi' 9 "after putting on his moccasins" (NT 34:10); keh$- gone' 'ado'lni*' "he reached into the place where the moccasins had been" (NT 32:22); ke bvh djinil "he(4) shook it (dust) into his moccasins" (EW 196:24). he "moc- casins" in these examples is to be compared with -ke'' "foot, foot- gear, moccasin, shoe," interpreted as a possessed noun, with ke- "foot," listed as a "nominal prefix" (5.48.), -ke* 7 "track, footprint," a possessed noun, and -ke^ "following, behind, next to, back of," a postposition (7.79). 4.6. Another characteristic of the possessed noun suggests its verbal quality. All the possessive prefixes have the same form as the 3 Haile 1943, pp. 71, 73ff. 48 NAVAHO GRAMMAR 4.6.-4. 10. objective prefixes of the transitive verb in the active voice, and of the subjective prefixes in the passive voice, with the exception of a?a-fourth personal possessive. The objective-subjective prefix of the verbal series is xo- which combines with the aspective prefix yi- continuative to become xa-. The so-called possessive xa-fourth person may, therefore, be viewed as the subject of the passive verb form — xa-mq, "his(4) mother" would then be analyzed as xo-yi-mq > xa-mq "he(4) is being mothered." 4.7. The prefixes used with the postpositions and sometimes called "objective prefixes" are formally the same as the possessive prefixes ; they are so uniform that it seems unnecessary to different- iate them in two series. The only reason for doing so seems to be the fact that in Indo-European languages the preposition, to which the Navaho postposition corresponds, takes an object. We shall see that the phonetic and semantic relationship between some postpositions and verb stems is so close as to suggest that postpositions also are primarily verbal. 4.8. xwe' < xo-e* "with, by means of him(4)" and xol "with, accompanying him(4)" suggest that xo- is the primary pronominal prefix, whatever it may be called. Since -e* "with instrumental," -i'h "into," and -I "with accompaniment" are postpositions phoneti- cally different from most, the theory that the postposition generally derives from the verbal form of the type yi-ta' "it is between. . . , it is placed between," seems justified. If so, the compound of "noun" and "postposition" would actually be a verb form, differing from the usual verb form merely in its simplicity rather than in its quality or function. It would thus be interpreted as a passive, "they are betweened," and with a noun, "I am mothered." Such an inter- pretation seems to account for much more than the classifications hitherto made, although of course the divisions, noun, postposition, and verb are justified as a matter of convenience. 4.9. A small class of verbs seems further to corroborate the con- clusion. Such verbs consist of a stem with a "possessive" prefix, the only change in the paradigm being the "possessive," better con- sidered as a passive subject (cp. 9.2.): si-dzvl I am strong rd-dzi'l you are strong bi-dzid he is strong xa-dzi'l he (4) is strong nxi-dzi-l we, you 2 are strong 4.10. The stem -trn "road, trail, path," apparently nominal, is treated as a verb in that it takes verbal prefixes : 4.10.-4.13. THE WORD 49 'a-ti*n road, trail tcidi bi-ti-n highway, auto road * bikd^ ^a-ti'ti highway; on-it road tsita* dibi ^aH~ti*n mountain sheep trail; rocks-between sheep their- trail-leads-beyond (FW 57:6) 'andbi-ti-n his trail back (YM 207) td* 'e'e-ti-n trail leads just (so far) (NT 168:3) t^4-ti-n exit, doorway, the way out : "* *■- e J e "fc * > o yas bi- na'a-ti-n path in snow; snow in-it here-and-there-something-leads yd^abi-ti-n his trail is lost (YM 207) The examples show that what appears to be a conjugation (the stem -tvn is distinctive) has many of the ordinary verbal prefixes, but a possessive seems also to be thematic. They may be extended variations of the possessed verbs of 4.9., but since -ti m n may be a stem of motion, perhaps it has more forms that those of possessed verbs which may be static. The point here is that the complexes have nominal (possessive) and verbal (conjugated) forms. 4.11. Apparently -Ice-' "track" is comparable in the following: nabi-Jcd-* his tracks were (visible) here and there (NT 130:23) 'abi-kfr'' his tracks led off naffl-ki*' track him nasil-ki'' I tracked him xode--M*' there was a footprint (EW 90: 11) The last three examples are verbal with no nominal traits, yet I have not found any other verbal form of the stem -M'\ 4.12. Another example seems to be a possessive noun used as a verb — bit nd-bi-ycl "with them another offering" — the possessive of ~ye*l "offering" is bi-ye*l, nd*- "again" takes the inflective prefix (-nd-) which raises the tone of 6i-possessive (10.94aff.). The same process operates in the forms 'andbitvn "his trail back," and 'and^d-ti-n "another trail." 4.13. Despite these unusual forms some nouns are differentiated from verbs by their form: They may be monosyllabic stems, they may have possessive prefixes, they are often independent. Normally they precede the verb, if there is one. Verbs also may be determined by their form, since they require prefixes. However, if the noun immediately precedes a verb with a simple prefix, particularly yi- progressive or yi-continuative of the third person, the noun may become the subject of the verbal complex and behave like a prefix. Actually the last sound of the noun, usually a vowel, contracts with 2/i-prefix. The former may absorb the latter so as to leave no trace of the verbal prefix, or like a verbal prefix, the vowel of the noun may be modified by lengthening, change of tone, or both. In such cases the noun functions as a prefix rather than as an independent word: 50 NAVAHO GRAMMAR *. 13.-4. 17. 'aze'-bfrj (l) Old Age lay (NT 128:13) x&jQle-lgo 'afe (< x6j6ni-yi-le-l) may it be beneficial, satisfactory xa'o-lydiati^ (< xa'o-lyd-yistM^) Things-pulled-out (name) was heard (NT 144:16) In the last two examples a verb is the nominal subject. 4.14. Obviously then there are at least three ways of interpreting the noun — as an independent word, with primarily nominal signif- icance, as an independent word having verbal significance, and as a verbal prefix. The distinction may be indicated by writing the noun separately when its function is independent, or as a part of the verb complex when it combines intimately with the verbal prefixes. We shall see that postpositions and adverbial elements may be treated the same way (10.30-10.31.). 4.15. The analysis of the noun will show that verbal forms without any modification whatsoever are often nouns (4.17, 5.98.). They may have possessive prefixes exactly as have the monosyllabic indepen- dent or possessed nouns : bibe-'altdi'didloh his buckle: his with-it toward-each-other-something-is- looped bibe-'etsxia, bibe^tskis his with-it something-is-jerked cibe-'eldg* my gun: my with-it something-is-caused-to-explode 4.16. The nominalizing suffixes -v "the particular one which" and -* "the one that" are free and may be suffixed to any form — verb or particle — to form a noun. Since many nouns, some even mono- syllabic, end in -i, -r, or -i, such forms must have a verbal derivation (5.23-5.30.). A further development of the same idea is the tendency of the stem with a low vowel to change to a rising tone ; the resulting form lacks the nominalizing suffix which is preserved in the tone as in: tid-h (Ktiah-i) Lefty, the-one-who-is-left-handed 'aban, 'abati, 'abcmi buckskin, soft worked hide 4.17. In these cases there is little modification of the verb to form a noun. On the other hand, certain nouns are verbal forms in every respect. They do not take the possessive prefix, but differences in person are indicated by conjugation, as in do* yic'vnv "my mother- in-law; the-particular-one-whom-I-do-not-see" (man-speaking); but do* yo-'i'ni' "his mother-in-law; the-particular-one-he-does-not- see;" 'actiohi "my weaving; something-I-am- weaving," but 'atlohi 4.17.-4.23. THE WORD 51 "her weaving; something-she-is- weaving." Consequently nouns can- not be understood and properly modified without conjugation. 4.18. Although many ideas which in English are adjectives are expressed in Navaho by static verbs, nevertheless a class of verbs may properly be called adjectives (9.). They are absolute in form; they stand immediately before the verb complex, and are therefore written as independent. Although not conjugated some of these adjectives have a verbal characteristic which also distinguishes postpositions — they have static and progressive forms : 'acte' (stat.) calm, soothing, tranquil, composed 'ae£6- (prog.) changing to calm, tranquil, composed; quieting down xaete* (stat.) normal, regular, usual, orderly, neat, ready xa&tfr (prog.) changing (from out of order) to normal, usual, orderly *adt' (stat.) wellbred, having breeding 'adi (prog.) becoming worthy, deserving, honorable 4.19. Comparable with this class of words is -ye- J a static verb in the form xo-ye ,y "weakening, futile, feeble," but ye-' is absolute or an adjective in bil ye,'* 'dxo'la- "he was intimidated; with-him futility was-made-thus. ' ' 4.20. Numerals seem to belong to this class of word. They, like nouns and postpositions, may contract with stem prefixes: df'sh}, df-'skq, (< df ,y -yiskq,) "four nights." In the following noun the numeral is compounded^with the noun and the possessive-nominal- izing frame bi-. . .-i* : bila'tfi-'v (< bila'-td-'-v) "fork; the-particular- one- which-is-three-f ingered. ' ' 4.21. Interestingly enough, the nominal stem may be conjugated. A prefix conjugation of a passive static verb is used with a stem identical with the noun (10.108.) : . . . yiltdah he has a hat like . . . , he is hatted like . . . cac yinistsi-' I have a head like a bear n&icdja-'' yilke-H owl claw; the-one-that-has-feet-like-an-owl (plant name) 4.22. The postposition, an important element, has been frequently mentioned as related to noun and verb. Some postpositions resemble verbs in having static and progressive forms : W-' (< hi-i-') (stat.) completely within it bi*h (< bi-i'h) (prog.) into it; moving into it M-ta? (stat.) between them bi-tah (prog.) among them 4.23. Some elements (stems) with forms identical with the post- positions are conjugated: 'anictah I am among, in the midst of y atah lie is in the midst of y ana* Uad da'ani'tah we are now in the midst of war; these enemies now we-are-amongst 5 Beicbard 52 NAVAH0 GRAMMAR 4.23-4.27. Compare bvnicyah "I am able to do it, I measure up to it" and bvyah "he is able to do it, he measures up to it, it is proportionate to it" with the more frequently used ciyah "alongside me" and bvyah "alongside him, it fits, it is enough." However, these two postposi- tions are the only ones I have found to be conjugated as stems. 4.24. Although it is convenient to speak of postpositions, actually they are not sharply differentiated from adjectives with independent forms or particles. Comparable with ye*' is tah, an element referring to time. That this is an independent form is exemplified by the negative, tah do* nndxdka'h-dah "still they are not returning;" and tah da m U% "later perhaps;" tah nfp 9 "time had passed," and other examples. Suffixes may be used with tah, in which respect it behaves like an adverb: tah-a*' (< tah-e*') "wait; later-future" tah-cq' "how about waiting, staying?" (WE). In the form 'dtcth td* kwe'e "wait right here" the analysis seems to be 'a-there remote (demonstrative adverb here used of time) and tah "time passes, there is an interval." 'd-tah-i-go "in a little while" shows tah with prefixed 'a-remote time and suffixed -i "that which" and -go the subordinating element, literally "future-time-that- which-is-being." Other examples of varied forms are : 4.25. to' xa\* 'e-lyodigi "he merely ran to the edge; the-one-who- merely-ran-off-to-a-place-on-the-edge-of-a-place dariiltia^go* in several directions (they went off) {-Uq,' "radiating from") -led-' "on" in words like naxokd-' dine "earth people" is verbal because the compound prefix naxo- is conjugated (10.116.). Compare also xo'tah "village, town; place-where-they-are-amongst;" #o-place is a verbal prefix. -dq* 3 is an enclitic which usually has temporal significance. Kasdq,-' and fta sidd*' (NT 44:27) are said to have the same meaning; the second form seems to be verbal (si-pf.). Both should be compared with 'vdq*' (< 'a-beyond-s/i-pf.) "at that time" and with td&e-dq,-' (< fc#-out-Vbeyond-[na-]) "doorway, entrance, yard." The spacial significance of -da*' is not clear in the last example. 4.26. A postposition may be suffixed to a noun : td-ta" between the waters xoyan-di at home k\-h (< kin-i'h) into town 4.27. A postposition may be suffixed to a possessive pronoun which serves as the object of the postposition: 4.27.-4.30. the WORD 58 bi-tei* toward him, in his direction M-yd through it bi-kd'* on it 4.28. A postposition may be suffixed to a locative (adverbial) element : 'a'-di at a nearby place y d--di at a remote place ko-di here, at a place near speaker kwi- ( -r, and ~i include the complete meaning of the verb with prefixes or postpositional con- struction, so frames consisting of prefix and suffix such as the negative do* . . .-dah "not . . .," fa- do- . . .-i "don't be the one to . . .," and others, qualify the words between them and indicate their close relationship. The combination of do* and -dah has several 4.35.-4.36. THE word 55 values: when it is a word, do'dah means "no, never;" when it occurs rn-at in an interrogative sentence do % may appear as dox, showing that it is a word ; when it is the first element of a frame, therefore, it is written independently. There may be a single word or several between the elements of a frame, and they are written according to the system already described. The last element of the frame, how- ever, -dah or -£, for instance, is not a free, but a bound form and it is suffixed to the last element of the complex it joins; do* yd 9 die* -dah "it is not good, it is bad, evil," do* la' xo-dza*-dah "it is by no means finished ; one-thing-has-not-happened." 4.36. To summarize, there is a base or radical, hereafter called the stem, which may be nominal, verbal or postpositional in character — its primary function seems to be verbal. A radical which is ostensibly nominal may be conjugated like a verb. Verbal complexes may be used as nouns with no modification whatsoever. Nominalizing suffixes may be used with any stem, adverbial element, or interrogative, that is, with any "part of speech," to form a noun. Postpositional or enclitic elements are closely related to verbs: They may be static or progressive. Verb stems and postpositions sometimes have the same forms which may be conjugated. Generally in such cases progressive-continuative stems have the progressive form of the postposition, perfective stems have the static form. Postpositions affixed to nouns, pronouns, or adverbial elements often have verbal meaning. A class of independent words may be designated as adjectives if defined according to meaning, but since they qualify the verb, they should probably be called adverbs. They differ from verbs mainly in not being conjugated; some have progressive and static forms. Formally noun, pronoun, interrogative and adverbial elements are in the same class insofar as they are related to the postposition — the postposition may be suffixed to any of them and the result is the same kind of complex. The noun, pronoun, adverbial element, with or without a postposition or enclitic, may become a verbal prefix. The utterance may be anything from a monosyllable, consisting of two sounds (usually consonant-vowel) to a complicated "sen- tence," composed of all parts of speech combined in complex ways. After this long discussion the question still remains as to what the parts of speech are. It is convenient to classify nouns, verbs, post- positions, and enclitics or particles, but I conclude that theoretically all are in a single class. The evidence that the elements are essen- tially verbal has much weight. 6-5.114. THE NOUN 5-5.2. Possession 5. Possession is a significant part of the Navaho noun, since many monosyllabic, that is, basic nouns change their phonetic form with the possessive prefixes. Such monosyllabic nouns, seldom used with- out a possessive, form one large class. Two types of possession are distinguished. Inalienable possession is indicated by prefixing a possessive pronoun (6.16.). Such nouns are primarily body parts and kinship terms. 5.1. Alienable possession is indicated by prefixing the possessive pronoun to the indefinite form of the noun: 'atd 9 "someone's wing, wing belonging to something," ce'efa* (< cVata?) "my wing, the wing that I use." There is reason to conclude that the forms of indefinite possession are nominal forms of the verb, that is, participles. This means that a stem like -fo' is interpreted as "it is winged," and J a-(a* "something is winged." If this interpretation is correct the relation between 'a-indefinite pronominal prefix, subject, object, or agent, and 'a-possessive is very close, if indeed they are not identical. The phonetic change from ci'afa' to ce'efa' is a common one; i-a> e'e in many settings, and the rule holds for all possessive pro- nouns except xo-his, her(4), because their vowel is i — ci-my, m-your, bi-, yi-his, her, its; nixi-, nxi-ouv, your (dual). There is, however, no way of explaining xo-a > xwe'e- for fourth person alienable poss- ession on this basis. If, however, the stem with indefinite possessive 'a- be regarded as a participle, the interpretation xo-e m 'a-> xwe'e- is comparable, -e* is a postposition meaning "with instrumental;" xo-e' > ocwe' "with him(4), and -e* is often shortened to -e- in this position. The verbal interpretation is substantiated by analogy with xwe* 'idin, often xwe'edin "he (4)has none; with-him(4) there-is-none," and anvc x$4 "he(4) has some; with-him(4) there-are." Thus the alienable pos- sessive would mean literally "with . . . something is . . .ed," or more specifically xwe'efd' "with him(4) something is winged," and all phonetic changes would be accounted for. Other examples of alien- able and inalienable possession with indefinite pronouns or fourth person possessives are: 'a-zis "pouch, sack, flexible container;" xwe'ezis "his(4) sack, pouch;" 'alt%' "bow;" xwe'elti* "his(4) bow;" *awd-\ 'aywt' "baby;" xwe'ewe-' "his(4) baby." 66 5.2.-5.5. THE NOUN 57 5.2. The phonetic stem changes discussed below are based upon these principles of possession ; they are the same for alienable and inalienable possession. For convenience both types will be referred to as nominal prefixing, the verbal significance suggested above being theoretical, probably historical. 5.3-5.19. Monosyllabic Nouns 5.3. The simplest form of the noun is monosyllabic; many such nouns are related to verb stems. It is impossible to predict which of the stems may be closest in form, but the static or perfective corresponds most frequently. Among the examples of 5.4. only two have a form other than static or perfective, six have a form common to the perfective and some other aspect (see 5.6-5.7. for the method of listing nouns with phonetic stem changes). 5.4. Noun HI that which comes free, twig, branch, fuzz, leaf tqj a flip, peck td {•to') water yas, zob snow -yol breath Uid hump, ridge, prominence xi*l (-y6*l) load, pack xoc (-yoc) thorn, cactus, splinter sdi (-z&i) sand, gravel, what has crumbled, slid sd'* abandoned place, evidence of previous occupation sq) old age sin (-yi-n) song tfy bog ci-^ (-ji-*) saliva, foam djd-d leg djddi antelope tci'l falling snow, falling cotton of cottonwood, what flutters in the air lid smoke dle-c white clay ti&' night tioh grass Verb stem diHl (stat.) it has long soft hairs -tqj (pf.) flip, peck di-to' (stat., pf.) it is watery -zo8 (prog., mom., pf.) sprinkle in a line, strew powdery material -yol (pf.) sob, weep -Uid (pf.) be humped, ridged -yi-ly -l-xe-l (prog.) move load, load moves -yoc (prog. , mom. , pf . ) mass be- comes thorny -sdi- (pres., pf.) pulverize, granu- late -5d*' (pf .) be disturbed at someone's absence -8$ (pf.) grow mature -yin (stat.) be holy, -sin (pres., inc., pf.) bless -tig (stat.) be boggy •j^ (pf.) spit di-djd-d (stat.) be fleet -tci'l (pf.) snow falls, it is snowing •lid (pres., pf.) cause fire to smoke -dle-c (pres.) smear, rub clay on -tU''' (inc., pf.) it is night di-tioh (stat., pf.) be grasslike, spiny, 5.5. Many monosyllabic nouns are independent: 9 6- 9 "clothes, shirt, garment;" bqrh "border, edge, rim;" kg' "fire;" Hal "notch. 58 NAVAHO GEAMMAR 5.5.-5.10. inner angle;" Uos "cloud;" JSq-j "body odor;" nil "ax, stone head;" ny** "storage pit, cache;" gic "cane." 5.6. These nouns and others, some of which do not often appear as unpossessed forms, have the same phonetic structure in independent and possessed forms : -da' "man's sister's son;" -da %y "lip, bill, beak;" -de*' "horn, antler, spoon;" -doh "muscle; -fa-' "father;" -tiah "pocket, crevice, fold;" -nv\ -nj,-' "face;" -god "knee, stump;" -kq? "male, husband;" -Jfce*' "foot, moccasin, shoe." 6.7. Nouns which occur seldom, if ever, as unpossessed forms are written with a hyphen preceding the stem, as -de-' "horn, antler" — 'adc' "someone's horn, spoon," is either a possessed or participial form. 5.8. Prefixing causes a phonetic change in some nominal stems; generally the possessed form is more like the verbal stem than the independent form. As we shall see, some changes affect consonants, some affect vowels, some affect both. 5.9. The initial surd of an independent noun becomes voiced when the noun has a position other than initial : Independent noun Prefixed form 'a-yS-l someone's pack, load bi-yoc its thorn, his cactus \i-za-d someone's speech bi-zdi its sand bi-zq-s his wart bi-zi-l its steam bi-ztf his star, its star 'o-j^'* someone's saliva bi-le-j its soil, his soil bi-lvj his urine bi-lj-' his horse, pet 5,10. The final surd of an independent form may be voiced in a compound form. This change may occur in connection with the voiced initial or the lengthened vowel; all three changes may be simultaneous (cp. 5.9, 5.13-5.16.): xfrl pack, loud xoc thorn, cactus sa-d speech, langu; s&i sand 8$-8 wart si-l steam *£>' star ci' saliva, foam le-j dust, soil, dirt lij urine *f horse, pet Independent form Prefixed form bfrc te*c xi-l 818 flint, metal, knife soot load, pack belt bi-bi'j bi-fS-j bi-yi-l bi-zi-z his flint, knife, metal its soot his load his belt djd'c tUl tffrl vagina fire-making apparatus rope, cord, string J a-dj6'j bi-tiH bi-tid-l someone's vagina his firedrill its cord, string, rope 5.11.-5.14. THE NOtTN 59 5.11. The long vowel of some nouns of the type CV-' becomes short in possessed forms : Independent form ya-* louse yo^ bead, necklace #a-' arrow Possessed form 'a-ya' someone's louse bi-yo' his necklace bi-Ua? his arrow, weapon 5.12. Some nouns of the same type have two compounding forms : 'daa*' jar, pot, dish tSa-* basket, bowl be'esa-' her jar; t6 y d8a > water jar (cer.) bi-tsa>'\ bi-t4a > his bowl, basket 5.13. The vowel of a nominal stem is sometimes lengthened with the possessive prefix. Some nouns of this sort have two possessed or compound forms, one with the short vowel, one with the lengthened vowel. One informant (AB) differentiates the two forms in meaning; the unmodified stem with prefixed possessive denotes simple pos- session. If the vowel is long and the final consonant voiced, the form means possessed as a part of a whole, functioning in a particular way." It is doubtful whether most speakers today, especially the younger ones, make this differentiation : Independent noun Possessed noun in relation to the whole HI twig, branch, that which is bVi'l its branch, foliage detachable sis belt bi-zi'Z his belt 5.14. The following illustrate merely the contrast between voice- less and voiced final consonants, and vowel quantity: Independent noun ttd'l rope, string, lariat Compound form non-functional 'aya-sis-tld-l rope of An- gora wool tsi-ttd-l hairstring dja--tl6'l earstring Compound form functional ''aza'-ttd-l rein, mouth- bowstring root;its-base- -tcei maternal grand- father, grandchild bi-tcei his maternal grandfather, grand- child rope bike--tld-l string biM-tid-l hobble, shoe- string; its-foot-string cd bi-ttd'l sunray; sun its rope bi-tsi'-tid'l his hair- string tcoc-ttd-l cinch, sur- cingle; belly -rope bi-tcei- his maternal grandfather, grand - child in relation to him 60 KAVAHO GBAMMAH 5.15.-5.16. 5.15. Some stems for which these differentiations are made have n as a final consonant : Independent noun Compound form non-functional tain song stick bisin bi-tain his song 1 its stick tatn bone tcin dirt, filth tcxin expended body substance subject to sorcery M-tSin his bone Compound form functional bi-yi*n his, its song re- lated to a song-group bi-tsi'Ti its handle, framework 'alta-tai^n arrowshaft, arrow-handle bi-tsz-n his bone, part of skeleton bidjct'-tSi-n bone in mastoid region, its- earbone bi-tci-n body filth bi-tcxi"n his body sub- stance subject to sor- cery 5.16. If the stem vowel is nasalized and the final consonant is a glottal stop, the simple possessed form is unchanged, in the com- pounded form related to the whole the nasalization becomes n and the glottal stop is lost; this is a process related to one occurring with perfective stems : Independent noun dd'dtq^ cornhusk, cig- arette paper tcq'* human excre- ment (vulgar) t6i'yif food (gen.) xosti' man, husband 'oadzfy woman, wife Compound form non-functional 'ate£*' base bitcq^ his excrement bitte-tf baxo8t%' his prayerstick her husband be'esdty his wife Compound form functional da'dfa-n cornhusk, fod- der ' dlatsi'Ti someone's wrist ; someone's- hand-base y ak6't8i'fi someone's ankle ; someone's- foot-base ''atca'Ti manure (polite term t&i'yd'fi food ready to eat Me-td'n prayerstick baxasti'n her husband in relation to her btfesdzd-n his wife in relation to him The following verbal forms should be compared with the stem changes just given; 'add sil4 "webbeam; in-front-of-something-long- obj.-lies;" 'add sitd-n "upper loompole;" 'dh sitd'n "lower loom- pole." 1 West of Ream's Canyon Main is commonly used for either meaning of "his song;" at Ganado it is considered ungrammatical, "wrong." 5.17.-5.20. THE NOTTN 61 5.17. A few nouns of type C^ change to type CV in possessed forms: Independent noun Possessed form t6 water bi-to' its water, spring ta6 stone, rock bi-tae* its stone, rock 5.18. Some nouns require the possessive with a high tone, a remnant of an older n-element, either a nasalized vowel or n : -t&s forehead -Uify' rib -tog collarbone -ji* name -ndi older brother -ji uttered breath -ni\ ~nf mind -tci*' nose -ni*\ -n{^ inside of nostril -tcx^ muzzle -Mai 9 mother's older sister -la* finger, hand -Mid protuberance, eminence -tiah angle, corner, side of face 5.19. The nouns are so arranged in the dictionary and word lists that the exceptions may be easily noted. If no possessive form is given in parentheses after the word, the possessive is regular, that is, there is no change. Wherever an exception occurs, the possessive form is in parentheses immediately following the noun, for example, W' (-V *) "bead;" sis (-zvz) "belt;" to (-to 7 ) "water." If the noun demands a high tone of the prefix it is written -stem, for example, -la? "finger, hand," 'dW "someone' sfinger, hand," cila' "my finger, hand," etc. 5.20-5.113. Compounding 5.20. The discussion of monosyllabic nouns includes changes in form and meaning due to prefixing. Nouns may be the result of composition of two or more apparently nominal stems. Since one stem follows another, there may be phonetic changes of the stem final. The final glottal stop of a stem is often lost in a compound: Noun Compound noun 'ayo*' someone's tooth 'a-yo'-ctiah inside of someone's cheek V/tev" someone's head, hair 'a-tei-ya' mane someone's tongue arrow 'a-tso'l&tah tip of someone's tongue *a-tta--tsi *n arrowshaf t 'adja-' someone's outer ear Ma--b6*c iron weapon point Ua'-yi'l quiver 'adja -'ti-j earwax dja'-tid'l earstring of beads 62 NAVAHO GRAMMAR 5 ^l.-5.24. 5.21. Nouns of more than one syllable are the result of compound- ing which is not limited to nominal stems, but may include other elements. A noun may be composed of a base with a suffix, a process which gives rise to many bisyllabic nouns. A noun may be a verb form, in which case it is analyzed as a verb complex — stem with prefixes; it may or may not have a nominalizing suffix. A large series of nominal prefixes, related to, but not identical with the monosyllabic stems, combine with nominal or verbal stems to form compound nouns. Whether the nouns are considered as nominal or verbal forms, affixes must be determined, and since it is good practice to start the analysis of the Navaho word from the end — the stem is final or near-final— the suffixes are presented first. Some of these suffixes sometimes seem to be bases, but they are included here because the examples are too few to make the category convincing. 5.22-5.37. Nominal Suffixes 5.22. -e- concerning, customs, manner, way: M6-c Hail Chant; conceming-hail t6-e* Water Chant; concerning-water na'kai-e' Mexican ways, customs; Mexican-concerning bektgd-na-e* American (white) ways 5.23. -v the particular one. This suffix is free; it may be added to any Navaho word to indicate "the particular one that " It is sometimes suffixed to a stem, becoming an inseparable part of the bisyllabic noun : *&-di- group of females mq'i' ma y i • coyote tqj-i- turkey; the-particular-one-that-pecks gd-g-i- crow • kq'-i* group of males y6H- god, gods tsa'-i- group of mature females, female sex tSi J -i* flea tM-i* small animal, Mexican hairless dog (referring to breed) tcah-i- crybaby, the-particular-one-that-cries •tcei' maternal grandfather and daughter's child in relation to each other (cp. 5.14.): ttiah-i- Man-with-a-hat (personal name of first man who wore a hat) do'tfij-i- turquoise; the-particular-one-that-is-blue 5.24. If the stem ends in a vowel, the suffix ~v may be shortened, combining with the vowel to form a short vowel cluster that makes it seem to be a part of the stem : 5.24.-5.30. THE noun 63 *k&i spouses, wives, husbands -Uei clan relatives, clan relationship xastoi old man, men, the wise old men, those in authority -tcei maternal grandfather, daughter's child (used in contrast with -tcei* when relationship to each other is not indicated) 5.25. -i the one that is. . . , the one who. . . , the place where .... This is also a free suffix, used exactly as is -i m ; it is less particular in its meaning : 'dh-i fog, mist 'dcf-i older sister -dtf-i man's mother's brother, man's sister's son older than speaker -de'j-i younger sister dine'-i (< dine*-i) tribe, group, people ltod-i sheep with coarse curly hair -tsil-i younger brother -j&& father -ttffi woman's daughter na'azis-i pocket gopher di'tcil-i abalone; the-one-that-is-iridescent 5.26. If the stem ends in a low vowel, -i may combine with it to form a short vowel cluster with rising tone. The suffix thus becomes a part of the stem : -nai older brother r H-ndi livelihood tcdi (< tcah-i), tcayi crybaby (FH) 5.27. The suffix -i may be incorporated into the stem, retaining its identity in the length and tone of the resulting vowel : tid'h ( ga-ni, ga'tl Mr. Arm (personal name) 5.28. Two nominalizing suffixes may be used in the same com- pound, that is, two nominalized forms may be compounded. Ex- amples of this kind illustrate how intimate the suffixes become as a part of the complex they create : na J azi8-l-to'-i Gopher Spring na'acd'-i'-to'-i Reptile Spring tsid-i' -to' -i Bird Spring 5.29. -igi the very one who, the very one that, the place where: belasd-na bitse^ x6l6n-igi pear, the-apple-that-has-a-tail bijfr' x6l6n-igi beer, the-one-that-has-foam 5.30. -igi' 9 -idi* the more remote one, the separated one which, who. This suffix is probably not carefully differentiated from -igi 64 NAVAHO GRAMMAR 5.30.-5.34. by the ordinary Navaho speaker. Both are used to form a substan- tive clause, as well as to denote nominal specification. Like other nominal suffixes, -igv is free and may be affixed to any word: tsi-tid-l-igi- old-fashioned navy blue hairstring; the-one-that-is-a-hair- string tsd'dszi' ntwh-igi- Yucca glauca; the-yucca-that-is-large xata r l-igi- that which he sings yiihzin-igi* that which he wants 5.31. -o*, -yd' dear, beloved; biJtis-6* his beloved brothers ciy&-y6* my dear husband bd'dcxe'-yd' my dear brothers-in-law 5.32. -ni 9 deceased, past, the late . . . , the one who used to be (cp. 11.36.): bij&i-nV his late father gici'-ni* the late Cane (User); the-particular-one-who- (had) -cane- deceased t8€k'tsoh6-ni > the late Big Belly (NT 368:27) xastin na^ta^ni^-nV Old Man Chief deceased; mister the-particular-one- who-was-chief-deceased (NT 370:17) bVi'^ lUciH-ni* the late Red Coat; the-one-whose-coat-was-red-deceased (NT 370:27) nigola'8-nV the late Nicholas (NT 372:1) * ncfkai sani-ni' the late Old Mexican (personal name) 5.33. -nv belonging to a place, people of the place. This suffix is used in clan names : bitah-ni' Folded-arms-people (clan name); place-wheie-it-is-in-its-fold (NT 92:13,14) mq?i' de-cgij-ni- Coyote Gap people (clan name) io*oa;e-d/t*-tit-People-of-the-place-where-waters-flow-together(clanname) td dit6i''-ni> Bitter-water-people (NT 92:13) tdxd'-ni' People-of-water's-edge-place (clan name) to-tsoh-ni- People of -big-water-place (clan name) kiya-^d'-ni* People-of-the-masonry-houses (clan name) xactlic-ni* People-of-the-mud-place (clan name) xonayd'h-ni' People-of-the-place-where-he-walked-about (clan name) ndxddld'-ni' people exist at a place 5.34. -M, -Mi plural of persons: *d-k& (< y d-d-k4) wives cinai'ki my older brothers ts4l-k6 young men, youths (EW 144:24 "young relatives") bitsiW-ki (< bitsili-kd) his younger brothers t&i-kii girls, daughters (general, not necessarily related) tso'-M (< taoi-M) daughter's children bila-kii his sisters, her brothers, siblings of the opposite sex 5.35.-5.36. the noun 65 5.35. -kfr youth: tsel-kq* youth, young man t6i-k%- young girl, maiden 5.35a. -lie*, -He seems to be the nominal suffix for place, compared with xo-, the prefix denoting place in verbal forms. It should be compared with the verb -fte'd (pf.) in xo*-Ue m d "there is empty space." It is debatable whether this is a nominal prefix or suffix, a verb, or a postposition: 'and-fc'e* someone's eye socket; someone's-eye-place ''at-Ue- in the same dwelling with; reciprocal (pronoun)-place Vo2-#e- it's there so let it remain; it has no special place but it will do there td-Ue* stream bed, channel; water-place ni--ke*h-& yucca mask of Night Chant; that-which-is-face-place yisd&a-Ue* lair, den; safety-some-place kg* ni-Ue'y xoni-Ue fireplace; fire space-in-line tsas-tte bed : main-part-peculiar-to-place tcdc-tte arroyo, broad -bottomed wash; above-rim-peculiar- to-place Compare %e*toh "wristguard, bow guard," and Ue'(d*n "prayer- stick, place-feathered," in which #e* seems to be a verbal prefix; 'allci dadcfatte "terraced gardens; one-above-the-other in-front there- are-spaces," in which -Afe seems to be verbal, or a postposition. In xaltcv'-Ue (< xo-yi-l-tci') "ground is red, place-reddened-place," and xo-tiv-yan "ghost hogan, place-empty-house" #e* is used with xo-. It is not unusual to find the same idea expressed more than once in Navaho words, and these seem to be verbal forms nominalized by -#e. Compare also xo'Ue'd "ruin, empty space, space." From these verbal examples I have concluded that -Ate* is the static continuative, -tte'd the static perfective verb stem. 5.36. -c-, the only element of its kind, connects two nouns and means that one "belongs to, is peculiar to the other." For example, n&i-c-til "nasal mucus," but ne'idil "nosebleed." "ne'ecdil would sound far-fetched because blood may be found anywhere but nasal mucus belongs especially to the nostrils" (AB). The element is relatively free as is illustrated by the elaborate compound, be-'eUe- 'eltcihi-c-to* "fountain pen ink, liquid for water or oil color" (cp.4.3L). It is interesting to compare this element with the possessive. The form y facial perspiration; someone's-face-peculiar-to- water *6lc\ -c-td > nasal mucus; someone Vnose-peculiar-to-water *dld-c-t6*j bark of tree td-c-to\ kd-c-to* perspiration 66 NAVAHO GRAMMAK 5.36.-5.39. na£o-s-tse* (a < c before ts) tobacco pipe; tobacco-peculiar-to-stone 'odjd-c-tcoh leg hair; someone's-leg-peculiar-to-eoarse-hair 'adjfr-s-tsi-n skeletal portion of trunk; thorax-framework bi-ni--c-t8i' his cheek; his-face-peculiar-to-flesh (NT 64:1) The freedom of -c- is shown twice in the example H-gQ-s-tei'-s-g&j "cartilage of his shoulderblade; his-shoulder-peculiar-to-framework- peculiar-to-cartilage". 5.37. Attention may be called to Hoijer's interpretation of -c- (although so far as I know he does not translate it). He says, "There is one example of a tri-consonantal cluster of type 3 [having a point of syllabic division between the second and third consonant of the cluster] differing from the above [final clusters beginning with glottal stop and having a second element s, j, I, or h] in form." 2 Hoijer goes on to cite the example xas-tvns-t&M "Wide Man" (per- sonal name), and concludes, "Here the cluster ns ends the syllable." On the basis of his own theory of syllabification, I think he should divide the word as xas-ti'n-s-til-i and consider -s- (< -c-) syllabic, for it is what he calls a "prefinal" prefix in its own right, as explained above. To me it is phonetically in the same class as syllabic s or z of the third person transitive (3-3) si-perfective. The examples given above show, however, that it is distinct from s-perfective (cp. 3.98.). 5.38-5.70. Nominal Prefixes 5.38. The nominal prefixes used in compounding nouns or as verbal prefixes are closely related to independent nouns. Some are shortened nominal stems, lacking the final consonant, as tsi*- (-tsi*') "head, head hair;" tso m - (-teo m ') "tongue;" dja'- (-dja-') "outer ear;" za,-- (-za m d) "word, speech, language;" dja- (-djd*d) "leg;" le m - (le-j) "soil, dust, dirt." Others are rarely used independently, although several examples show that they may be: di-hfrnV (-de* 9 ) "antlers of the one men- tioned." Compare de nneinrlgd' "when the antlers have been put back" (NT 322 : 14, 17). In the first example de is independent ; in the second, it is doubtful whether de is a word or a prefix. Forms of the prefix type may be used with postpositions in which respect they are like stems: naya* s^si "One-that-has-a-wart under-the-eye" (place name). 5.39. The nominal prefix often has the form CV. The verbal char- acter of the noun is demonstrated by the fact that if the nominal prefix with a high tone is followed by another prefix, the latter is high because of the inflective prefix (nd~) (10.25, 10.93.). In the list of nominal prefixes the independent or final stem is written in parentheses : 2 Hoijer 1945c, p. 25. 5.40.-5.45. the NOUN 67 5.40. da- (-M-') rim, edge: -dd-gi rim of orifice (g is a glide consonant) ; the-one-that-is-the-rim dd-ya* person's beard and mustache; mouth-hair •dd-yi* throat ; edge-inside ~dd-ziz uvula; edge-flap -dd-djd'j opening between mouth and nose dd-ydtddsi tapering mustache (of animal or person) dd-ydndzbqsi One-who-has-a-handlebar-mustache (personal name) 5.41. dd'd- corn, plant: dd'd-tq^ cornhusk, corn leaf, cigarette paper d$d-ta*n cornhusks, fodder dd'd-kaz cornstalk dd'd-Ke cornpatch, cornfield; corn-place da d-ydli- rattlepod (plant) dd'd tca-n cornsmut; corn-excrement dd'd-tSi* pith of corncob dd*d-lgai 'akd-n white cornmeal 5.42. dd'- y da- (cp dd'h "moving in front of" 7.40.) in front, fitted in opening : 'a-dd-dildjoli herbs for sealing prayersticks (cer.): some-bushy-sub- stance-that-is-placed-in-opening dd*~dinizi*ni* door guard; the-particular-one-who-stands-in-front dd--diitt&i door guards (as snake, lightning) (cer.) ; house furnishings kept near doorway dd'-didjah bars across corral opening dd'-ditj, dd-Mit{hi wooden door, gate in one piece dd'-fostlvn* dam; something-has-been-piled-in-front bi-dd'-ne^zld slightly overlapping, just touching (as branches of a shade) (FH) 5.43. td- pertaining to water td- y dgi8 washing, cleansing, dipping, bath td~bq*hd People-of-water's-edge (clan name) td-bq*hgi shore, beach ; place-at-water's-edge td-zil sac on internal organ : water-collects td-tci* sweathouse 5.44. ib- (< td-w&ter-i'h into) deep water, in deep water: ti'-lj-* water horse (myth.) tfr-xo'lted'di* water monster (myth.); the-particular-one-that-grabs-in- deep-water 6.45. n&-> -nd- (-nd*') eye, small seed, grain, essential part: 'a-nd-diz someone's eyelash; eye-twists (on-itself) ^a-nd-t&j someone's eyebrow; eye-blackened y a-nd-gai someone's cornea; eye-whitened 'a-nd-Jce* someone's eye socket; eye-place 'a-nd-Me'Cto' someone's tears ; eye-place-peculiar-to-water 'a-nd-ziz someone's eyelid; eye-flaps 'a-nd-tdi'Ti someone's brow ridge ; eye-bone 6 Reiohard 68 NAVAHO GRAMMAR 5.45.-5.51. J a-nd-ji-n someone's pupil ; eye-black 'a-nd-tlah corner of someone's eye; eye-corner nd-ya- s$-zi One-that-has-wart-under-the-eye (name of Joseph City, Arizona) 5.46. nffe- { L nv\ -n^') pertaining to inside of nostril: n&i-dil nosebleed ni'i-til nasal mucus ; nose-slimy n&i-cdja-^ screechowl; nose-peculiar-to-outer-eajr -ni^i-ya^ hair of nostrils 5.47. -gqs- {-ga-n) pertaining to arm: -gq'-doh arm muscle -gq'-yos shoulder joint -gq*-ziz sleeve; arm-flaps -gty-stsi-n shoulderblade ; arm-peculiar-to-foundation -gq*-8tsi*ttf middle part of shoulderblade; arm-peculiar-to-main-part- between -gq--lo- i lower arm; arm-appendage 5.48. M- (-he*') foot, footgear, pertaining to foot: hi bi'h dji'nil (dust) he shook into his shoe; shoe into-it he(4)-moved pl.-obj. Here hi must be interpreted as an independent noun. hi-tal heel of shoe, moccasin ; foot-move-swiftly ki-ni'' instep; foot-middle ki-kal sole of foot ki-ldtah toe ; foot-tip ki-ltci moccasin uppers;- foot-red ki-'abani buckskin footwear -hi-tid-l shoestring -hi-ke footprints; foot-place •hi-cga-n toenail; foot-peculiar-to-shield xa-ki-li-j his(4) foot dust (cer.) (EW 196:244) hi nine-zi boots; the-footwear-that-is-long 5.49. fte-- see 5.35a. 5.50. -yd- (-ya\ -ya*') pertaining to body hair, fuzz, fur, wool: 'otze** dd-ydi (< ddydisjin) Sloan's liniment; medicine-blackened- mustache (named for picture on box) dd-yd-tioai tapering mustache ; mustache-narrow dd-yd-ndzbqsi handlebar mustache : the-mustache-that-curves dd-yd sika-d mere tuft of beard, small goatee isi--yd mane; head-hair 5.51. -yd- (-yo tJ ) pertaining to teeth: ^a-yd-tah fold between teeth and lips ; tooth-fold bi-yd-cgi-j He-lacks-a-tooth (personal name) : his-tooth-peculiar-to-gap ^-yd-ckal place where tooth is missing; tooth-notch 'a-yo-ctlah inside of cheek; tooth-peculiar-to-angle 5.52.-5.59. THE NOUN 60 5.52. yd'-, ywi'- smooth, shiny body covering; insect, worm: y6'~ne*ct&jdi' cicada (gen.) yd'-ldjini black ant yd'-ldtci-' ant (gen.); red ant (spec.) yd' ait&ili bedbug yd- siUidi- worm, caterpillar that humps in moving, measuring worm; the-particular-worm-that-humps 5.53. za- (cp. -za y "belched gas") in mouth, passingthrough mouth: za-'azis packet of garment; mouth-pouch za-xodi"yoh suction of gas Oza-dzo'l-tiin . . . was punched in the mouth 5.54. za*- (swd, -za*d) mouth, voice, word, speech, language: za--riU herbal medicine; pl-obj.-are-placed-in-mouth •za'-bq-h lip; mouth-edge za'-lani western mockingbird; words-many bi-za'-dil sacred blood, blood of animal ritualistically killed; its-mouth- blood 'a-za'-l&tah final meaning of words; speech-tip 5.55. za-, zd*- pertaining to inside of mouth: z&di, zdyi oesophagus; that-which-is-inside-of -mouth zd-bq-h 'dFi'h lipstick; mouth-inside-border is-made-thus \i-zd'-fVi bridle, bit, rein; that-which-is-strung-inside-mouth -zd'-yo-j hard palate; inside-of-mouth-botryoidal 5.56. -zi- pertaining to neck, throat (outside): zi n&yfy 1 surface of chest and neck to lower jaw and ear zi ndz'di separate shirt collar; neck-stiff-obj.-projects-in-arc 5.57. -zfr- (-ze t9 ) breath; part around outside of neck: zi- de-tiH necktie; zi' (Mdyi'h collar worn in ceremony zi- deyini cape zi- di'ldoi scarf zi' na'zfoH necktie; it-lies-strung-around-neck zi' n&ztiH necktie; it-lies-strung-against-neck zi- sidoi, zi' de'Sdoi shirt collar; that-part-of-neck-which-is-warm (Cp. bizi-* xazli-' "he died; his-breath space-became" NT 368:23) 5.58. -tsd- main part of: -tsd-gah part of body between lower ribs and hips -tsd-tiid tendon of Achilles ; main- tendon tsd'd&zi* yucca; something-main-fibrous tsd-sUe bed ; main-part-peculiar-to-place 5.59. t&il- % : tsil-k^ youth, period from early adolescence to middleage tail-tea*'' buzzing beetle 6* 70 NAVAHO GRAMMAR 5.60.-5.67. 5.60. tsi'-, -tsi*- (-tsi' 9 ) pertaining to head, hair, or both: 'a-tai'-yastd'n someone's neck tendon; long-object-lies-under-peculiar - to-the-head 'a-te^*-yq^• , someone's brain; head-nerve- tissue *a-tsi-~ziz seal]), part of foreigner used in War Ceremony; head-flap 'a-tsi'-tM-n skull; head -bone '*a-t8%--y6ti- top vertebra, atlas; the particular-one-where-head-pivots tsi-'tld'l hairstring, hair-cord 5.61. -tso-- (-tso-') tongue: 'a-tao'-yaz tongue papillae; tongue-papilliated ^a-tso^ldtah tip of tongue 5.62. tM-- foundation, essential part: tSi^-tiin skeleton; essential-bone tsi'-zia quiver; essential-pouch dd'd-tH' pith of corncob; corn-essential-part 5.63. cd, cq,- {c4*') pertaining to sun: cd ndi-n sunstreamer cq didi'n sunbeam cq-l&tah end of sunbeam ; sun- tip cd-da'&'h south; sun-round-obj.-is-starting-to-move cd bUtd'l sunray 5.64. dja-- (-djcr 9 ) outer ear, ear lobe: dja*-fini* dock (plant) dja*-n4-z mule : ear-long dja--cjini Black Ears (dancers) Odjci'-kal ... is deaf; . . . -is-ear-covered 5.65. djd~, dzd- (djd'd) leg: -djd-nil fringe of robe; separate-obj.-lie-against-leg -dzd-ziz fringe of robe; limber-obj.-lie-against-leg djd-I6$'j odor of genitals : leg-acrid -dzd-stis shin -dzd-tiid leg sinew 5.66. tea- shade, darkness: tca-xa'oh shade be* tca-xac'ohi umbrella; that-with-which-place-is-shaded tca-xalxe'l, tccfaxalxe-l, tca-xa*lxe*l darkness; shade-dims-place 6.67. ted- above rim ( ?): tca'ol pinyon tree tcd-cUe arroyo, deep sharp wash; place- where-arroyo-begins-to-merge- with-plain tcd~cdjic diaphragm tcd-tfoc syphilis (AB) 5.68.-5.72. THE NOTTN 71 5.68. Id-, -Id (-la 9 ) digit, finger, hand: -ld-ydji small finger -Id-tsi'n wrist; hand-attachment (YM 123) Id-tsini bracelet (YM 123) -la-tsoh thumb ; finger-large -Id tsosiiidi index finger; the-seventh-fingcr 5.69. le- } le-- (le-j) soil, dust, dirt : le'-H'rii-l cemetery; soil-in-which-pl. -are-laid le'-yV in ground te'-s'^n pit-baked bread le--ya neyani One-nurtured-under-ground (myth.) le'Uoh yellow ocher ; soil-yellow le'-tia*'' bowl, earth-bowlshaped le*-djin coal; earth -blackened le- H-ci-j ho poked it into the ground (NT 22:24) le-'aznilg-go- to where pieces of meat were buried (NT 22:28) 5.70. rfa*- (-tta'') bottom, buttocks: tla-~kal skirt; buttocks-covered th'-dji J V*' pants; bottom-toward garment bi-tla'-dja-'' bias binding, rickrack braid 5.71-5.113. Composition of Nouns 5.71. The difficulty of interpreting the Navaho elements, demon- strated by the word and nominal affixes, is further shown by noun composition. Nouns, verbs, and postpositions overlap and inter- change in so many ways that it is impossible to draw a line between any two categories. One rule seems to stand out, namely, that a nominal stem stands first in a compound, whereas a verbal stem usually has at least one prefix ; it may be a noun. There is, however, an exception — the verb stem has an initial position with a nominal- izing suffix: yol-i "ability, capacity, might;" bal-i "shawl, the-one- that-is-curtainlike ;" tioz-i "Angora goat ;" tlog-i "Sia Indian ;" tcahi, tcayi, or tc&'h "crybaby." Such forms are doubtless shortened forms of the participle with nominalizing suffix, but the class is large enough to indicate that a verb stem, like a nominal stem, has some capacity for independence. 5.72. The participial form is very free; it is formed by prefixing 'a- "some, someone, something" to any of the principal parts of the verb. It is nominal as well as verbal: 'atioh "weaving, something being woven (pres.)"; 'asiiQ "weaving, something is woven (pf.);" 'o'fis (<'a-i/i-prog.) "something is roasting (prog.);" 'ates "some- thing is frying (inc.) ;" 'azfe "something is roasted, fried (pf.) ;" 'atcah "there is crying" (cp. tcah "crying"). 72 NAVAHO GRAMMAR 5.73.-5.80. 5.73. The participle may be given a nominal form with a suffix: 'atioh-i ' 'weaver, the-one-who-weaves-something;" 9 aMah-i "miller, the-one- who -grinds-something. ' ' 5.74. Although the assignment of a word to a verbal or nominal category is sometimes arbitrary, it is clear that once a form has been nominalized — as a participle, or by a nominalizing suffix — it retains the nominal character and thenceforth behaves like a noun in a compound or utterance. Moreover, the suffix binds all that precedes it into a unit. 5.75. In the following examples the analysis is the one that seems preponderant; others might be possible. 5.76. A noun may be composed of two nominal stems, the second of which cannot be independent ; the resulting form is noun-noun : td-zia waterbag, glass jar, bottle tsi-dd-' Rock Rim (place name) tcoc-ttd-l cinch; belly-rope (cp. 'abati tid-l buckskin or goatskin lariat) ld-djic glove; hand-case -dd-ya-' person's beard and mustache; edge-body-hair dfrl-da-' cranebill (cer.) cd-ldtah end of sunbeam ; sun-tip 5.77. A noun may be composed of more than one stem, either of which may be independent : dibi dfy* sheep food (plant name) ; sheep corn gah tcidi rabbitskin robe ; cottontail robe JSa- bfrc (< #a-' bi-c) arrowhead; arrow flint tai bq'h Rock Border (place name) ts6 ktf yellow ocher ; rock fire isi' tid-l (< tai* y tid*l) haircord 5.78. A noun and a possessed noun may form a compound noun: tqji- bi-b6j6'' turkey comb; turkey its-comb tainil bi-tai'ti ax handle; ax its-handle tUn bi-tta marrow ; bone its-fat ty' bi-yfrl saddle; horse its-pack 5.79. Two or more nouns may combine with a possessed noun to form a compound noun; the position of the possessed noun may change. Position probably indicates emphasis: tai l\~* biyfrl Saddle Rock; rock horse its-load tai bini' t6 Cliff-face- water (place name); rock its-face water 5.80. A noun may result from combining a noun or nominal prefix and postposition; 5.80.-5.86. THE NOTTNT 73 -dd-yi* throat; edge-within yd-'q-c heaven; sky-opposite -go-dai (< -god-dai) thigh; knee-up-from kin-tah town ; masonry -houses-among tsi-ya- cave; rock-underneath tsi-ytf canyon; rocks- within le'-yi* subsoil, underground; soil-within 5.81. This process may be carried further, so that another noun may be added, the result being a new noun : tsd-tah kin (Sanlldefonso); House-among-rocks ; rocks-among masonry- houses tse-na- 16 Senatoa Spring; Water-around-rock tsd-Kiz-tdh-i Rock-crevice- water ; that- which-is-rock-crevice- water (NT 342:27) 5.82. The noun may be composed of a nominal stem and two post- positions : tsi-yV-gi Place-in-box-cany on; rocks- within-place. Here the noun and postposition is the "noun;" the postposition -gi "in place" is added, the entire compound being the place name. 5.83. The compound noun may be composed of noun-postposition and verb : tsiya- ni-tci^ Red Cave; rock-under red-paused %s6yi* xaydji Small Canyon; rocks- within place-is-small xacM'dji na-td-h war captain, army officer; warrior-side he-exerts authority t8&a*lty Rock-ridge-in- water; rock-into-water-ridged t86na,'dji-n Bennett Peak; Blackened-around-rock 5.84. A second verb may be added to the same structure: tsdiwdji'n-y&ji Little Bennett Peak; Blackened -around-rock -small 5.85. In the following, noun-postposition and verb are so inti- mately related by the phonetic change of the verb initial that the verb cannot be independent : tse-ya-'tindin-i' rock crystal, crystal; some -particular-rock -through -which - light - shines - beyond (AB). Compare with this an alternant form, tse-yd diriidi'n-v the-particu- lar-rock-through-which-light-beams-beyond (YME 22). In both forms the compound is framed by tse and -r, the nominalizer. 5.86. A postposition may be nominalized, or serve as a noun : be--di utensil, implement; that-which-is-"withed" ; that-which-is- an-instrument hi-Jcd'* its top; credit; what-is-on-it bi-yi* bottom side; its-in-it 74 NAVAHO GBAMMAE 5.87.-5.91. 5.87. A postposition with a noun may form a noun : bikd^ *ati*n highway; on-it road biUi xatd-l patient in ceremony; over-him there- was-singing 'aycf-tSi-n someone's lower jaw; something-under-attached-bone 6.88. Some nouns result from a compound of noun and verb stem, as if the noun were the subject of the verbal complex; they are schematized as noun-verb stem : ma'i'-tsoh wolf; coyote-large tsd-gi'j mountain pass ; rock-gap -djd-dl-l (< dj&*d-dH) thigh; leg-big tcici-nd-z Chiracahua; scoundrel-tall tco*~yin menstruation; genitals-dangerous le*~jin coal; soil-black tloh-tcin onion; grass-odorous kin-ldni Flagstaff, Arizona ; masonry-houses-many 5.89. Nouns or nominal prefixes may be compounded with a verb stem, which may have a form unlike any of the principal parts: Va-n "hole" (from -'# [stat.] "be open"); -td-rt instead of ~t4 "long slender obj. lies;" -ywn instead of -yq, "eat;" -djvn instead of -djin "be blackened;" -gvj instead of -gij "be forked, have a gap." Such forms seem to be analogous to the possessed noun with lengthened vowel to show a more intimate relation than does the stem with short vowel — for example, Uin "bone" and bi-Ui-n "its bone as part of skeleton" (cp. 5.13-5.16.): 'd4d-gi*j crotch between fingers; someone's-finger-gapped ts6-gi-j Rock Gap (place name) dzil-ji-n Black Mountain (cp. dziljvn and dzil lijin "the mountain is black") 'an&-ji"n pupil and iris; eye-black tti-yd-n prepared food, food ready to eat (cp. t6i*y4 "food" [gen.]) 5.90. The following noun is composed of noun verb-noun : 16 ltc4'6-ko'h Red-water-canyon (place name) ; water the-red-one-canyon (NT 348:7) 6.91. Many nouns are formed by combining a noun and an independent verb : 'abq'h fitV border string of blanket on loom; some-border it-is-strung-up *ano* xo'dzoh boundary line; stranger place-is-marked 'aze*' ditci*' chili, pepper; medicine it-is-bitter bidd*' xcfazti* Rim-strung-out, Grand Canyon mq'i- Usoh yellow fox; coyote is-yellow dc ndstSa^ (< dfe*' ndatia*') ram; horn curved-bowllike) na'dfr'* 'datfrh crackerjack; corn it-is-roasted kin ya- J d tower; masonry from-underneath-projects (the postposition -ya* has become a verbal prefix) y i-nd yd'dc$ good health; life it-is-good lain dirli' bullroarer; stick it-moans 5.92.-5.98. THE N0TXN 75 5.92. A noun may consist of a noun and several verb stems: tsi-tci'-ta* Red-rock-shelter (place name); rock-red-shelters 5.93. A noun and verb with nominalizer may form a noun: "anv" bildjol-i powderpuff ; face-that- which-is-moved-fluffy-against-it 'dsa*' nimaz-i round utensil, receptacle ; pot the-one-which-is-spherieal V*' na-tsg-d-i' sweater ; garment the-particular-one-that-stretches-around- a-restricted-area yo-' 'aydll sleighbell; bead the-one-that-rattles >>#■' 'asUidi camel; the-one-whose-back-is-humped tsi'istH paper bread; something -that-is-baked-against-stone t6il ncfattyi- grape, raisin; the-particiilar-plant-that-mtertwines-here- and -there tdil xa-€aH- marshhawk; plants the-particular-one-that-flies up-out-of t&il litsohigi- orange; plant the-one-that-is-yellow 5.94. Two nouns may be combined with a verb and a nominalizing suffix to form another noun : belaad-na bitse^ xdlonigi- pear; apple the-one-that-has-a-tail 5.95. A noun, possessive with postposition and verb — practically a sentence — may be a noun either with or without a nominalizing suffix : isi bi*' de-sgdi White-in-rock-enclosure (place name) ; rock within-it it-is- white "atsi ba* na#aw^Mbutcher;meaton-account-of-it the-one-who-exchanges tloh na-ldjo'U hayfork; that-with-which-hay-is-moved-about 5.96. A noun, a verb, a verb with nominalizer and a postposition may form a noun : tsi yiUd-n d&de-atlin-i-gi concrete dam; rock it-is-ground place-where it- is-piled-in-front 5.97. The following seem to be si-static verbs (third person), which I have found in no other forms. They are treated as possessed nouns, as if the stems were as indicated here, the possessive pronoun being the only modification : yi-&te\ yi-stei lunch (si-ste* "my lunch") yi-sga\ yi--gq' covering, skin, foliage yi-stU, yi-atU' legging, stocking (cp. Haihi-ltle^ "he had buckskin leg- gings" NT 160:3) 5.98. A verb may be used as a noun, often without a nominalizer: H"di* thunder; something-that-moans-rep. HH m '$ sunset; round-obj.-has-moved-beyond *6ltaH pupil, student; one-who-reads dahi-stty loom; tied-so-as-to-hang-down dahdini-lya'j fried bread, sopaipilla; prolonged-bubbling-on nane-lje-' warp; it-has-been-stretched-evenly-and-tightly 76 NAVAHO GBAMMAB 5.99.-5.106. 5.99. A verb with a preceding locative may form a noun: 'ada xi-l\ waterfall 'aya- sit$ lower loompole; underneath-rigid--obj.-Hes y dlah xa*zl\ celebration; crowd -is-in-place 'd&ah *ale-h council meeting, conference, assembly; crowd becomes 'q- xa'Vi'l residue after filtering; extraneous it-has-been-caused-to-float- out y q- xd-'frl skimmed liquid; extraneous it-has-floated-up-out t<$d-h djiyd-n, td6-t6i"yd'n watermelon; in-vain one-eats 5.100. The locative preceding the verb may consist of a possessive and postposition: bd xaz'4 rule, code; for-its-benefit things-are be' xaz'q, legal requirement; with-it things-are bil na-'a-c man's male cross cousin; with-him they-two-go-about bi* xo'dzq hollowness (as of a pipe) ; inside-it place -is-hollow be-'etsxis whip, switch; with-it-something-jerks bitte'estcj writing, picture; over-it-something-has-been-caused-to-bear- down 5.101. The complex formed of possessive-postposition and verb may be bound together by a suffixed nominalizer : f aUi dahiriili saddleblanket ; those-which-hang-down-over-something 'alMfedisl candy; those-which-are-twisted-one-against-the-other 6e* 'adiHgci yeast; that-by-means-of-which-something-ferments 6e* y ddifo-di towel, handkerchief; that-by-means-of-which-self-is-wiped 5.102. A noun may be formed from a verb and a noun: *alt86 xasti-n First Man (deity) diyin dine 1 & Holy People; holy group naxokd'' dine*& earth people; here-and-there-in-place people (NT 16:16) 5.103. A noun may be composed of an adjective, noun and verb : 'acdla' 'attS-' xatdd five-night chant (this may be interpreted also as verb verb verb) 5.104. A noun may be composed of an adjective, a noun, and a possessed noun : id** tqji- bikin Three -turkey -house (place name) three turkeys their-house 5.105. Two verbs may form a noun; either or both may be nominalized : narii'd xatsoh Large Span, Big Bridge (place name) : something-projects- across place-is-large biUi 'addni biUestiH tablecloth ; that-which-is-eaten-off-of the-one-that- covers-it 5.106. A verb, either nominalized or not, to which a postposition has been suffixed, may form a noun : J addni-gi dining-room; place-in-which-something-is-eaten HHdke-d-gi motion picture theater; place-in-which-something-is-moved- smoothly-beyond 5.107.-5.112. THE NOTTN 77 5.107. The examples and many other words that hare nominal force, especially the long descriptive complexes that contain all the fundamental grammatical elements, free as well as bound, show that each is an utterance. The nouns are, therefore, syntactic as well as morphological. The tendency to create descriptive terms, marked in all Atha- baskan languages, is especially well developed in Navaho. It is the device that gives the language its large and subtle vocabulary. Any speaker may devise a name for a new object or a circumlocution for a well Iftiown one, and is likely to be understood. Consequently, there are often three or four names for an ordinary object, all of which are correct ; it may be that no two are compounded on exactly the same scheme. 5.108. Place names are often built on the elements to, -to\ td- "water;" kin "masonry house;" tse "rock." 5.109. The prefixes Vindefinite pronoun, and 'a-beyond, into indefinite space, so often form verbal nouns that only those which cannot be analyzed will be listed in the dictionary; if they can be analyzed they are to be put under the verb stem. Many verbal nouns have a prefix zwi-about, here and there; they will be listed under na- and under the stem as well. 5.110. Another prefix #o-place, things, especially "supernatural things, things not explained or understood," is used as extensively, particularly to describe abstractions or ritualistic ideas. It is often combined with wa-about, becoming naxo- or naxa- to describe things referring to the earth or universe. 5.111. As is to be expected, the postposition -e m "with instru- mental" is a part of many nouns of the type "that-by-means-of- which-it. . ." Often the forms are passive: be-'eldg' "gun; that-by- means-of-which-explosion-is-caused," Frequently too be* is used with a nominalizing suffix in which case be' . . . -i serves as a frame to denote an instrument: be-'atiod "small pump; that-by-means-of- which-there-is-sucking-through;" be t9 eldg- be* 'andocdltahi "trigger; that-by-means-of-which-gun-is-caused-to-spring-back-in-place." 5.112. The fourth person serves, among its other functions, to generalize statements, often in nominalized forms, and in this sense is equivalent to the impersonal "one" : tdi'h djiyd-n, tdi't&i-y&'n watermelon; in-vain-one-eats-some t&* xoltad'di' Water Monster; the-particular-one-that-grabs-in-deep-water t6 xadji-lkwdi Place-where-one-splashes-water-out-with-hand 78 NAVAHO GBAMMAB 5.113.-5.114. 5.113. Forms of the verbs 'a-. . .-'i'l "do, make; 55 and 'a-. . . -U'l ''make, construct, create" have become stereotypes for descriptive nouns : H-yU*ini-, HyMHni, H-Vini one-who-makes-rep. bd'h H'yiVini baker; bread one-who-makes td dilxil H-yiVini brewer; water dark one-who-makes 'ayo*' H-yiVini dentist; teeth one-who-rep.-does-thus-to Vase-' *&V\* (< 'drfi) hospital; place-where-medicine-is-made-rep. t&i'yd-n bi-h 'dVini kitchen; food in-it where-it-is-prepared bU *£eTini baking powder ; with-it something-which-is-made-thus 5.114. Borrowed Nouns 5.114. The preceding array of possibilities for word coining may be a reason that Navaho draws comparatively few words from lan- guages outside the Athabaskan family. Particularly interesting are the reinterpretation of the palatals and the modification of accent of words borrowed from Spanish and English. Generally the accented vowel has a high or falling tone in Navaho. The following list is suggestive, but not exhaustive: 'al6*z (< arroz Sp.) rice '<5Za* ( -yi~) apparently identical with many aspective prefixes with which they contract, their isolation is a matter of confusion and doubt. The fourth person is formally easy to identify because its form (dji-) is outstanding, and because its position is far forward, that is, as nearly initial as possible in the stem complex. Moreover, it dominates a great many other prefixes, either absorbing them or contracting in such a way as usually to retain its identity. The difficulties of the fourth person are in usage, for the assignment of the fourth, instead of the third, personal role to a person spoken to or about is as subtle as the usage of the familiar and polite forms of the second person in the European languages. Nevertheless the fourth person may be explained just as are the first and second, in terms of itself: 6.7. Out of respect the fourth may be used consistently of or to an individual even if there is only one third person. 6.8. If there is only one "third" person it will usually be of the form mentioned without qualification, or as "3." If however a second third person is introduced, the one mentioned first may be referred to as "3," but the second one will be designated as "4," and these distinctions will be preserved throughout by a person who speaks consistently grammatical Navaho. 82 NAVAHO GRAMMAR 6.9.-6.13. 6.9. The fourth person may be used instead of the second person by siblings of the opposite sex, or by those who want to indicate extreme respect in their relationships. In some cases a change from second to fourth person may indicate a warning that trickery of some sort is planned or detected, that sorcery is to be performed, or that a person is speaking sarcastically. 6.10. First and second person dual possessives and objects have the same form; first and second person duals differ in all other pronominal forms. 6.11. To summarize, six persons, first, second, and four third persons must be differentiated in the singular; of these the four "third persons" are the same in the dual. First and second duals are distinctive as subjects. Plurals are formed by using plural stems with dual prefixes, or by prefixing da-plural to the duals. 6.12. Hoijer and Young-Morgan include xo- the prefix of "place'* with pronouns in their paradigms. 2 xo- is not by any means re- stricted to the third person as would be expected if it were the subject, for although xo- may be the subject, it may also mean "in place" and may be used with any person meaning "... moves in place," Moreover it sometimes seems to be a subject or object meaning "things, supernatural things ..." and may occur in any person. In fact, it is often thematic. Therefore, instead of including xo- in the regular paradigms, I treat it as a prefix with its own conjugations; often, of course, they may be defective. Incidentally, xo- is one of the very few prefixes with o vowel and therefore a pattern of unusual contractual changes when combined with other prefixes. 6.13. Independent Pronouns 1 cih I Dl nxih we two PI danxih we pi. 2 nih you D2 nxih you two P2 danxih you pi. 3 yih he, she, it D3 yih they two P3 da-yih they pi. {$)bih he, she, it D(3) bih they two P(3) da-bih they pi. 4 xdh he, she D4 xdh they two P4 da*x6h they pi. i Hh someone, Di Hh some two Pi daHh some pi. something Although the final h of this series is often dropped, its occurrence in compounds seems to justify the interpretation of h as the stem final. Compare, for instance, nixih-igv "we who are; the-ones-who- are-we" (NT 66:21); cih 'frdi "I am the very one who . . . ;" xoh-$- "he(4) aforementioned" (NT 36: 17); with cic "is it I, am I the one 2 Ibid., pp. 2, 77 ff. 6.13.-6.16. THE raottOTW 83 who ..." (NT 138:9); ci k$- "I here" (NT 34:4); Had tor ni bbHintv "now you guess it" (NT 58: 10) (cp. YMG 4). The meaning seems to be rather of the type "it is . . . ," than simply, "I, you, he." The independent pronoun is often used without a verb. If an utterance includes a verb the independent pronoun is not often used, unless it be for emphasis. Person is indicated by the pronominal prefix of the verb. 6.14. The independent pronouns may be nominalized by suffixing -*' to the above stems. By preposing id' "just, really, absolutely" such a complex would be emphatic: td m cihi "I myself;" td- xohi "he(4) himself, they(4) two themselves" (NT 142:19); fa* nxihi "we ourselves, you yourselves." 6.15. Possessive Pronouns 6.15. There are three series of possessive pronouns — the absolute possessives, the possessive prefixes, and the emphatic possessives. Absolute Possessives 1 d- ' it is mine Dl nxi-' it is ours PI danxi-' it is ours 2 ni-' it is yours D2 nxi-' it is yours P2 danxi-' it is yours 3 yi- J it is his, hers, its D3 yi-' it is theirs P3 dayi-' it is theirs (3) bi-' it is his, hers, its D(3) bi-' it is theirs P(3) dabi-\ da-bi-' it is theirs 4 x6-' it is his(4) D4 x6-' it is theirs (4) P4 daxo-\ da-xo-' it is hers(4) theirs(4) pi. i H-' it is someone's Di '<■' it belongs to Pi daH-' it belongs some two * to some of them pi. Examples of the absolute possessives are: Had ntsoi ndo'h'l "now your grandson will become your own again; now your-daughter's- child your-own will-become -back" (NT 26:22); td' bv' nlfgo "what is possessed; just his-own-being." 6.16. Possessive Pronominal Prefixes 1 ci- my Dl nxi- our, belonging PI danxi- our, belonging to two to pi. 2 ni- your D2 nxi- your, belong- P2 danxi- our belonging ing to two to pi. 3 yi- his, her, its D3 yi- their, belong- P3 dayi-, dai-> dei- their, ing to two belonging to pi. (3)6i- his, her, its D(3)6i- their, belong- P(3) dabi- their, belong- ing to two ing to pi. 4 xa- his, hers(4) D4 xa- their(4), belong- P4 daxa- their, belonging ing to two(4) to pi. i 'a- someone's, Di 'a- belonging to Pi da'a- belonging to something's some two some pi. These prefixes are used with nouns to indicate possession: ci-mq, "my mother;" ni-kql "your husband;" bi-nd*' "his eye;" danxi- Iceyah "our country," etc. They are also used with postpositions — 7 Eeichard 84 NAVAHO GRAMMAR 6.16,-6,18. the "possessive pronoun" serves as the "object" of the postposition: ci-l "with (accompanying)-me;" ni-ya "forcefully away from you;" bi-kd'' "on it;" xa-tii "over him(4)." The use of the two forms bi- and yi- will be explained in connec- tion with the same forms as subject (of passive voice) and object pronouns. They differentiate subject from object, possessed subject from possessed object, and are closely related to the position of nouns or other nouns in the sentence (6.21-6.26.). 6.17. Emphatic Possessives 6.17. The emphatic possessives are compounds of fd m "really, absolutely, ' ' the independent pronoun , and the corresponding possessive prefixed to the noun; td* is sometimes omitted. 1 td- cih ci- my own. . . 2 td' nih ni- your own . . . 3 id' yih yi- his, her, its own . . . (3) id- bih bi- his, her, its own . . . 4 id- xdh xa- his, her(4) own . . . i id* Hh 'a- someone's own . . . Dl,2 fa- nxi nxi- our own, your own PI, 2 id- danxi nxi- our, your own P3 td' dayih yi- their own P(3) id' dabih bi- their own P4 id' daxo xa- their(4) own Pi td' daHh 'a- their (some or others) own Examples of emphatic possessives are: td' bih bi-tso-sJeid bikd^gi "in place on his own thigh;" ci se-'eywe-' (< ci-'aywfr') "it is my baby" (NT 36:27); ce'eywfr' ne cicdnHtf "I got it (baby) back; my- baby truly I for-me it-was-put-down" (NT 38:2). 6.18-6. 18a. Subjective and Agentive Pronominal Prefixes 6.18. The subjective prefixes are the same for the intransitive and transitive active voice. Those of the agent of the passive voice differ in only a few forms, but since these small differences have important effects on adjoining prefixes, both series are given below. The differ- ences are due to position and contraction depending upon the place of the prefix in the complex, even if, as in the third person, the pronoun is zero. Subjective prefixes Agentive prefixes 1 -c- I -c- by me 2 -n-, ni- you -n- by you 3 — he, she, it •yi~ by him, her, it 3 yi- he, she, it (passive only) 4 d/ji- he, she, the one dji- by him, her, the one i 'a- someone, something 'a-, 'adi- by someone, by something Dl -i'd- we two -i-d- by us two D2 -oh* you two -oh- by you two 6.18.-6.20. THE PRONOUN 85 Since the aspective prefix or prefixes intervene between da-plural and the objective-subjective, or subjective-agentive prefixes, changes may occur in various combinations, and the plural is omitted here. The combined prefix forms will be found in the prefix paradigms (10-10.124.). 6.18a. All these pronominal prefixes, except dji- and 'a- stand immediately before the verb complex, some of them may contract with the classifier or the stem initial, dji- and 'a- have a position as near initial as possible, depending upon other prefixes in the complex. 6,19-6.28. Objective Pronominal Prefixes 6.19. The pronominal object or objects of a verb are prefixed to other elements of the verb complex. The object of a verb in the active voice precedes the subject and other aspective prefixes. The objective pronouns are the same in form as the possessive prefixes, except the fourth, which as an object is xo-, as a possessive xa- (4.6-4.8., 5.1.). 6.20. The pronominal prefixes that denote objects of verbs in the active voice are subjects of verbs in the passive voice. It is important to note that the third person passive has a subject yi-, which though lacking in the active voice, is the object, and therefore comes within the rule. For convenience these prefixes are repeated here: Object of Verb in Active Voice and Subject of Passive 1 ci- me (passive, "I") 2 ni-, n- you 3 yi- him, her, it (passive, "he, she, it") (3) bi- him, her, it (passive, "he, she, it") 4 xo- him, her (passive, "he, she") i *a- some, someone, something Dl,2 nxi-y nixi- us two, you two (passive, "we two, you two**) PI, 2 danxi-, danixi- us plural, you plural (passive, "we plural, you plural") The objective pronouns are properly a part of the verb complex. Examples are given here to show their position and function: 'a-ci-yvlxan-i'* "after throwing me away she . . . ; beyond-me-she- threw-after" (NT 52: 18); nd-ci-drlt{ "he picked me up; up-me-he- paused-starting-to-move-live-obj." (NT 52:25); nd-si-'Altj (si < ci-) "he carried me back ; back-me-he-moved-live-obj .-to-end" (NT 52:25); xaHcq? da'tid'-ci-j-do'dlil "let's see what they(4) will do to me again; whatever by-them(4)-pl.-again-I-will-be-done-do" (NT 64:20); k&rte' nvdo'M'l (ni'- < ni-yi-) "in here you will spend the night; in-here night-will-pass-(over)-you" (NT 42:20); xd'dicq? ni-ctcf "where were you born V 9 (NT 52: 16). 7* 86 NAVAHO GRAMMAR 6-21.-6.24. 6.21. Let us now take up the differentiation of the third persons, bi- and yi-> the rules for which apply to possessive and pronominal verb prefixes. Since bi- is a third person object with any subject except the third, it is not modified when referred to. When, how- ever, it appears related to another third person, it is referred to as (3). yi- is the third personal possessive or object as distinguished from the third subject. 6.22. If a simple possessive relationship between two nouns is expressed, the possessive prefix is bi-: 'awe'' bamq (< bi-md) "the baby's mother, it is the baby's mother;" xastvn be'esdzd-' "the man's wife, she is the man's wife;" tcidi bidjd'd "the car wheel, it is an automobile wheel;" tse bikd' 7 "it is on the rock; rock on-it." 6.23. If, however, a noun is possessed and followed by a verb in the third person, or any other relationship between two third persons is established, care must be taken to distinguish the subject and object. These depend upon the position of the nouns and other related words in the sentence as well as upon the verb form. Contrasting forms will bring out some of the differences. It will be noted that some of the sentences become ridiculous when the rules are not observed. The informants say that many Navaho make such mistakes ; possibly the pronominal system is as difficult for them as for us. 6,24. Subject-Object *awi*' yijtci (intr.) the baby was born *aw6'' yictci (< yis-l-tcty she gave birth to a baby y awd^ bam4 yizyas the baby scratched its mother; baby its-mother it-scratched-her bam4 'awi-' yizyas its mother scratched the baby; its-mother baby she-scratched-it *aw6-* bamtj, ndidi-lxa-l the baby hit its mother with a stick; baby its-mother it-caused-it-8tick-to-move-(against-her) rm^V dibd yiyi'8x\ (< yiyi-s-lxj) the coyote killed the sheep ; coyote sheep it-killed-it dini 'awfr* yiyi-si'h the man is standing the baby up; man baby he-is-causing-it-to-stand dini djddi yiyi-lted-h the man sees the antelope; man antelope he- sees -it djddi din6 yiyi-ltse-h the antelope sees the man djan tarn yi'fczlo 9 John cheated Tom djan tarn yidi't'o'loh John will cheat Tom tia m '4 yi' yiyi'giz he cleaned out the basket (EW 104:25) 6.25.-6.26. the pronoun 87 6.25. Object- Subject or Subject-Agent 'awi-' 'asdzfy' bictc{ (< bis-l-tc{) the woman bore the baby; baby- woman she-bore-it 'awi- y bam4 ndbidi-lxa-l the baby's mother beat it; baby its-mother she-caused-stick-to-move-(against)-it dibi mq'i' bi-sx\ (< bi-8-l-x{) the coyote killed the sheep; sheep coyote it-killed-it dini 'awe*-' bi-8j,-h the baby is standing the man up 'aw^*' dini bi-8j,-h the man is standing the baby up djddi dine bi-ltse'-h the man sees the antelope dini djddi bi'ltsi-h the antelope sees the man 'acki-td bi-lx6-h the boy is drowning; boy water it-is-killing-him From these examples it seems that in the active voice the yi- form of the object is used if the nominal subject precedes the object, but that the bi- form is used if the nominal object precedes the sub- ject. FoBowing the rule that the object of a verb in the active voice becomes the subject of the verb in the passive, and the subject of the active becomes the agent of the passive (8.22.), the bi- form is used if the order of the nouns is subject-agent-passive verb. 6.26. The following examples are more complicated ; they involve the bi- and yi~ forms with postpositions and their relation to other parts of the sentence. Forms with yi~ i aw6'' > bam4 yil nli the baby loves its mother; baby its-mother with-her it-is H* y yihd*' dahne-zdd he is sitting on the horse; horse on-it he-is-sitting- on (cp. NT 26:25) yibaH6-8ta? he has counted past the limit; its-capacity he-has-counted- beyond bidjd'd yq,'h niind y 4 he put it back on the wheel; its-wheel on-it he-put-it-back tcidi bidjd-d yq,'h neind^ he put the wheel back on the car; car its- wheel on-it he-put-it- back ytyh yi'a'h he is pawning round obj.; into-it he-is-moving-round-obj.- to-goal ye' xwtfeztty ( td-8td > the list is exhausted; beyond-its-capacity it-has-been counted xaya- xayi'lkd^ni^ biya* niyinikd after he had set it down in front of him(4) he set it down in front of him(self) (WE) 6.27. Compared with the usage of yi- and bi- the fourth person is simple. Like the first, second, and indefinite, it takes bi~ as thethird person object because the subject and object are different personal forms : bamq *«w^*' dzizyas the mother scratched the baby; its-mother baby she(4)-scratched-(it) 'a«^-' xam# dzizyas the baby scratched its mother; baby its(4) -mother it(4)-scratched-(her) *am$ 'awfr* dzizyas a mother scratched a baby; some-mother baby she(4)-scratched-(it) "amd xwe'ewd-' dzizyas this mother scratched her baby; some-mother her(4)-baby she(4)-scratched-(it) be'esdzd'n nte'-' xwe'esdzd'n 'ddji-la- she who had been his wife was made his(4)wife; his-wife used-to-be his(4)-wife was-made-by-him(4) (NT 30:18) ba-jniyd (< 6a- djiniyd) he(4) came up to him xalcq* *&dji-la- (the woman) made him her(4) husband; her(4)-husband he-was-made-by-her(4) tsd'dszV be' be'ts*iztf$ (< bVadziztlfy) with yucca fiber he(4) tied it; yucca with-it he(4)-tied-it The examples cited with a translation as good as I can make illustrate the extreme complexity of the third and fourth personal pronouns. It is likely that I have not stated all reasons for each usage. Another problem is the use of bi- or yi-(in the third person) as a verbal theme. The simplest verb forms and those occurring most often take yi- as a definite third person object if the form is of the type "he . . . it" (3-3); in the other persons the object is implied rather than expressed. 6.28. Other verbs seem to require an expressed object, in which case the object is thematic. The formula for such verbs will be written 0-. . .-stem. The following are examples of such stems (progressive only) : 0-. . .- y ah tempt, lure . . . on, deceive, fool . . . O- . . . -dil catch up with . . . moving O- . . . -tyl take ... a long stiff obj. away O- . . . -toil release . . . , let . . . out of one's grasp O-'o-theme . . . -tto-l tie ... to 'a-beyond-O- . . . -l^ac cause two persons to move, despatch two persons (EW 94:13,15) 6.28.-6.31. THE PRONOUN 89 >a-0- . . .-l-'yl keep hidden from . . . (EW 94:6) 'd-thtw-O-. . .-l-'j-l harm . . . ; cause doing thus to . . . 0'a*-di-. . .-loh cheat . . . O-di- . . ,-si'l make . . . stand up; cause ... to be righted O-di- . . . -16s lead . . . away on a rope O-di- . . . -dial break ... off with force na-0^-( thither from there near speaker 'a-di at (the place) there 'a-do- on from there near speaker *a-dji* to a point near speaker a*- there near second person a--d$-' thither from there near you a- -di in place there near you a- -do- on from there near you a'-dji" 1 to a point near you d- there, near third person, remote d'-d^-' thither from (over) there d'-di in place (over) there d • -do ■ on from over there &--dji y to a point over there xa- (gen.) who, which, what, where xdi (xa-i) who, what in general xa-d^ what thither from definite point xa-di where is it (at, in place) xa-do- on from where xa-gd* on to where xa-djV up to what point 7.2. Interrogatives xa'- what near you, what in immediate time (past, future) xai* what, which one near you xa-Hh who, which of some xa-'d^' 7 thither from where near you xa*-di where is it in place near you xcb'-do* on from where near you xa--go* on to where near you xa'-dji' up to where near you xa- what in remote future; who, which of possibilities xdi' who, which of all xd'ih who, which of some xd--d(y thither from where remote xd *di whereat of all places xd'-dd' on from where remote xd - -go • on to where remote xd--dji y up to where remote 92 7.2.-7.5. BOUND FORMS 93 The following examples, though few, seem to indicate that some of the demonstratives may be free forms : V niznizi-d he(4) scraped it (hot earth) aside; there he(4)-moved- scattered-substance-to-end (NT 80:12) 'd- over there (NT 188:21; 190:15) 'Hep *d- yilyod he ran to them over there (NT 64:6) 7.3-7.10. Adverbial Elements 7.3. Some adverbial elements may have the same initial position as the adverbial demonstratives, and some may also have a position corresponding with that of a postposition or enclitic. Such elements may refer to time as well as space : 7.3a. nah- aside, toward the side (cp. -nah 7.54.) : nah-go- td-di da-tSi nijde-ltd-lgo after taking about three steps aside; aside-toward three-times perhaps after-he(4)-had-stepped (NT 34 :10) nah-dji* dind-h move off to one side (YMG 52) 'ak&'dQ** nah-d^ ^d-d^ '»*(££•' dahi-zf stand over there; next frorn-the- side from-over-there from-here stand yas nah-go- 'ayi-zge*d he shoveled the snow aside (YMG 32) to- nah-dji* 'ayi-ld6-l he merely threw them (moccasins) aside (NT 34 : 7) 7.4. ?wr-, nqr- sidewise, leaning, across. This element may have initial or final position, and is to be compared with na*nv "across, in horizontal position;" and with na-ni-go "breadth, width, crossing, to the side, being crosswise, sidewise." Compare also teeriq; and xaria* both of which are said to have the same meaning "across" : na-H-go' he fell sidewise (NT 336:3) na-H-k$'Z long rigid obj. toppled over, fell sidewise (YMG 32) xadjilgic na- dahizdi-l forked lightning flashed across (NT 208:3) ti-s ncffiti* Row-of-cottonwoods-across (place name) tozia nq- nina'a-h turn the bottle on its side (WM) tsi rwjt* sa'dni Leaning Rock, Tilting Rock (WM) nanVd it projects across (YM 11) Uo8 dilxil nand'd djini yodahgo a dark cloud was lying across up above they say (NT 224:19) na- daaitQ-go (gun) lying across (his knees) (NT 384:22) xa*g6-na- Id tsina- ndi-kah Id ? where truly can we cross ? where-toward- across truly across we-pl.-shall-go-to-end truly (NT 200:4) *4i bitd^-h go-na- do- na'ddd-da in front of this (woodpile) one is not supposed to go; this in-front-of-it toward-across one-does-not-go- across (NT 324:9) yondnd^^ ko-na- Id the crossing from the other side is here: from-the- other-side here-across truly (NT 200: 18) *alna--go crisscross, being-across-one-another (NT 80:16) 7.5. yd- (yo-) off from actor, off into unlimited space, probably out of sight (cp. yfr'o- \