Introduction

Douglas R. Parks

Presented here is a collection of 156 Arikara oral traditions that range in content from mythology and other folkloristic genres to historical narratives and cultural descriptions. The collection provides comprehensive coverage of the major genres in the historical and literary tradition of the Arikaras, and through the contributions of numerous narrators illustrates what survived into the late twentieth century of the tradition's fundamental attributes as well as its topical diversity and range of stylistic variation.

The texts are presented in a five-line interlinear form: Line 1 represents a transcription of the original Arikara oral recordings of the narratives. Lines 2 and 3 present, respectively, a morpheme-by-morpheme parsing of the words in line 1, together with English glosses of those morphemes. Line 4 provides literal, word-by-word English translations, while line 5 offers free English translations of the texts.

This text collection was published in a four-volume set entitled Traditional Narratives of the Arikara Indians (Lincoln: University of Nebraska Press, 1991). In volumes 1 and 2 of TNAI the texts are presented in a two-line interlinear format, the first line a transcription of the original Arikara and the second one a literal word-by-word English translation. Free English translations of the texts are presented separately in volumes 3 and 4, where the arrangement of the stories parallels their presentation in the first two volumes. Volume 3, moreover, begins with an extended introduction that serves the entire collection, covering all topics but those that relate to the tape recordings and the presentation of the linguistic texts themselves, the latter of which topics are discussed in volume 1 and here.

First established in the late nineteenth century, this form of presentation—transcriptions of texts in the native language with both literal and free translations—has become the standard in American Indian language study, which for over a century has sought to document as fully as possible individual languages whose communities of speakers continue to diminish. It is, moreover, a format that maximizes the accessibility of the collection to the broadest possible audience, one that includes Arikara people, scholars, and interested readers. To the more general end, the documentation of the Arikara language, this set of volumes constitutes the first of a tripartite description of the language, to be followed later by a dictionary and a grammar, both of which will be based in large measure on this corpus of material. Because these texts represent the only major collection of traditional narratives ever recorded in the Arikara language, and because that oral tradition has become virtually moribund with the passing of this generation of narrators, the collection will serve as the basis for all future study of the Arikara language as well as this people's rich oral tradition.

The Recordings and Transcriptions

The texts were recorded during the period from 1970 to 1988. They are the contributions of eleven narrators, all elderly at the time the recordings were made and representative of the contemporary Arikara-speaking community. (Biographical information on each of the narrators appears in the Introduction to Traditional Narratives of the Arikara Indians, vol. 1.) A list of the narrators together with the date when and location where each story in the collection was recorded, is given in the List of Tape Recordings below.

Most of the recordings were made on either Scotch 290 (0.5 mil- or Scotch 209 (1 mil) magnetic tape, using a Uher 4400 Report L or 4400 Stereo IC tape recorder. A small number of others were recorded on a Sony TC-105 recorder. Early recordings were made with either a Uher M534 dynamic or a Superscope EC12B electret condenser microphone, while later ones were made with a Sennheiser MD402 super cardioid directional microphone. All recordings but one were made at 3¾ ips; the exception was recorded at 7½ ips. All taping sessions were inside a house, either the narrator's or the editor's, generally at a kitchen or dining room table or in a living room chair, where at least partial control overextraneous noise was possible.

Mrs. Ella P. Waters collaborated in the transcription and translation of the stories of all narrators except Lillian Brave (1-61 and 85-156). Working together with her daughter, Angela Plante, Mrs. Brave assisted with her own stories (62-84). In the early period of my work two other individuals, Mrs. Fannie Whiteman and Mrs. Nellie Yellow Bird, also aided me in writing out and translating several stories, while in the last stage Mr. John Fox and his sister Mrs. Melfine Everett assisted with two texts. The procedure followed in all cases was to replay a story in short stretches, which the native speaker then repeated in Arikara and translated into English. After the transcription and translation were written out, each segment was generally played back once again to insure correctness of the transcription. After a handwritten interlinear version of a story was completed, the text was typed. Unfamiliar and questionable forms were then noted and later clarified with either the translator or the narrator. Following these procedures, all the texts were verified for accuracy, and the transcriptions and translations were refined.

Isolation versus Discourse Forms

An important feature of the published versions is that the transcribed Arikara words are written as they would be uttered in isolation rather than as they occur in connected discourse. The salient differences in form conditioned by the two contrasting environments are relatively slight—they include the loss of final glottal stops, shifts in stress, and changes in vocalic and resonant voicing and devoicing—and are summarized by three general rules:

(1) A word ending in a glottal stop in its isolation form will, when another word immediately follows it, lose the glottal stop; at the same time stress shifts from the vowel of the penultimate syllable to the final vowel, which is generally lengthened in slow speech. For example:

(2) When stress placement changes, devoicing occurs in a syllable when there is an appropriate new environment or shifts from one syllable to another, as in the following examples:

(3) When a word ending in a voiceless vowel in isolation is immediately followed by another word, the devoicing of that vowel generally does not occur. For example:

A short paragraph (from story 3) presented in parallel isolation (I) and discourse (D) forms further illustrates the differences:

Narrator Usage and Translator Preference

In the transcription process, translators frequently edited taped speech, whether the story was the translator's own or someone else's. Some of these changes were simple corrections of grammatical forms that were obvious errors or misstatements, and such revisions, when clearly warranted, were made when writing texts out or later when checking on specific problems. Other emendations, however, were more idiosyncratic, affecting the style or nature of the narrator's speech, and were in most instances not incorporated into the final versions of the texts. The latter emendations, which were primarily a matter of personal taste or perceived notions of correctness, generally fell into one of five editorial categories: eliminating repetition, uncontracting deictic proclitics, inserting adverbial proclitics, inserting evidential prefixes, and substituting one demonstrative form for another. In all these cases the translator or narrator repeated speech as she thought a passage or word should be said rather than as it actually occurred on tape.

Eliminating repetition. Repetition is a common stylistic device used in Arikara to emphasize or add dramatic impact to a statement or to clarify it. Sometimes the repetition is an identical restatement of a word or sentence; at other times the word order is inverted, and occasionally the repetition is an expansion of a sentence that adds a noun subject or object that was not in the first statement.

In narrative performances there is also another type of repetition—the reiteration of a single word or phrase—that has no function other than as a “space filler” to give the narrator time to pull his thoughts together while deciding what to say next.

When helping with the transcriptions of their own stories as well as those of others, translators frequently eliminated both types of repetition, commenting that, even when the repetition was clearly stylistic, it added nothing to the story and impeded the narrative pace unnecessarily. In editing the texts, however, I have retained all stylistic repetition and omitted only those instances of duplication that are clearly hesitation forms. If there was doubt about the possible stylistic function of a repeated word or phrase, it was retained.

Uncontracting deictic proclitics. Two commonly occurring proclitics are contractions of the independent deictic particles, tiiháʾ ‘this; here’ and tiihéʾ ‘this; here,’ with the clitic niku- ‘the one who/which; the place where.’ The resulting contractions are taaniku- and teeniku-, respectively, both translating as ‘this is the one who/which; this is where.’ For example, taanikutiihiʾ ‘this is where it is’ is a contraction of tiiháʾ and nikutiíhiʾ.

Although most speakers use these contractions in preference to uncontracted phrases, Mrs. Waters was adamant that the uncontracted phrase was proper usage, even though she herself occasionally used contractions, and whenever she repeated a narrative she always separated the deictic particle from the clitic. In text presentation, however, I have followed actual usage.

Inserting adverbial proclitics. When producing verbal forms inflected for one of several modes in Arikara—the indicative, assertive, and absolutive—a speaker almost invariably adds to them one of the two most commonly used adverbial proclitics, we- ‘now’ and noo- ‘then; there,’ particularly if no other clitic occurs. Neither of these adverbial clitics has the semantic force of its English translations. Consequently, when translating verb forms into English, an Arikara speaker frequently does not give an English equivalent for them, and when translating English verbs into Arikara a speaker nearly always adds one of them to the Arikara verb even though the English rendition does not overtly call for it. The proclitics, then, virtually serve as formal elements defining a fully formed word in colloquial usage.

In narrative discourse, as in isolated verb forms, most verbs in the indicative, assertive, and absolutive modes possess one of these clitics, but occasionally they do not. When repeating forms on tape that do not have one of them, translators frequently inserted one. The insertions, however, were not incorporated into the final versions of the texts.

Inserting evidential prefixes. In narratives an Arikara speaker must attest to the validity or source of his statements by selecting one of several evidential prefixes for use with verbs. One of the two most common prefixes is wi- ‘quotative,’ which occurs with verbs in the indicative mode. This prefix is never translated into English by Arikara speakers but is explained as indicating that the narrator is telling a story that he or she heard from someone else. The prefix is the equivalent of what in many American Indian languages is translated as '‘it is said.‘ Another equally common prefix is an- ‘evidential proper,’ which occurs with several modal prefixes (usually in the form Ah-) and connotes a statement of some action or state that was not witnessed. It, too, is generally not translated into English by Arikara speakers, although occasionally context will elicit a translation like ‘I guess . . .’

Ideally all verbs in a narrative about something the speaker did not witness should have these prefixes, but in the actual performance of storytelling speakers vary, sometimes dramatically, in their use of them, with no apparent pattern to their occurrence. At one extreme is a story like 136, told by Dan Hopkins, in which only 40 percent of the verbs that would potentially take evidential prefixes actually have them. In other stories of Mr. Hopkins, however, the average number of verbs with evidential prefixes is approximately 65 percent, slightly under the average for most speakers. At the other extreme are narrators like Eleanor Chase, Lillian Brave, and Ella Waters, who use evidentials on nearly every verb, with an average incidence of the prefixes generally exceeding 90 percent.

Yet when repeating taped speech in the process of transcribing texts, translators almost invariably added evidential prefixes to all verbs, whether they had them or not. Several speakers, in fact, remarked on numerous occasions that when stories are told these elements should appear on all verbal forms. I have not included them in the transcriptions, however, unless they occur in the actual delivery of a narrative.

Substituting demonstratives. Demonstratives in Arikara are gerundial constructions based on a small set of verbs of position, shape, and movement that modify or refer to a noun. Depending on their position in a given context, humans and animals are described as sitting, standing, lying, or going or coming, although sometimes even finer distinctions are made. Examples of the most commonly occurring ones are:

    tⁱnaaríčI aruúsaʾ ‘this horse’ [arik ‘to stand’]
    tⁱnaákUx wiináxtš ‘this boy’ [kux ‘to sit’]
    tⁱnaáxA suúnatš ‘this girl’ [xa ‘to lie’]
    nuunaátA wiítA ‘this man’ [at ‘to go’]

Other objects are classified according to their shape and size. Long, slender objects, like an arrow or a snake, for example, are classed as lying, while round objects are sitting. Some objects are arbitrarily classified. Celestial bodies, for example, though round, are categorized as standing.

In late twentieth-century Arikara speech this classification had begun to give way to a simplified system in which most objects were denoted as sitting. Although younger speakers in particular neutralized the older distinctions in most contexts, the shift to the simpler system did not entirely follow generational lines among speakers but also occurred within the speech of many individuals, both older and younger. Some elderly speakers, particularly Mrs. Waters, carefully adhered to the older system, but most others tended to mix the older and newer usages to one degree or another, and speakers also differed in their usages from one story to another. Use of the simplified system predominates in some of Mr. Morsette’s stories, for example, while in others he makes the older distinctions. During the transcriptional process actual usage was preserved in the stories even though occasionally the translator unconsciously changed demonstrative forms, giving a correct older form.

Other editorial conventions

A common feature of Arikara oral delivery is use of the particle he, the equivalent of English “uh,” a hesitation form used like its English counterpart to give the narrator time to think of the next word, phrase, or sentence that he wishes to utter. Some speakers use the particle more frequently than others, and its usage generally increases with the amount of effort necessary for an individual to recall features in a story, particularly one that has not been told in many years. There is no discernable pattern to the use of he, and since it is a meaningless element that only interrupts the narrative flow of a story and is never repeated by translators when transcribing a story, its occurrence has been omitted in the transcriptions.

In the tape recorded versions of the stories, it was also not uncommon for narrators to make false starts in beginning a word and occasionally to make grammatical errors. In published and web versions of the texts, false starts have been omitted and errors or misstatements have been silently corrected, unless there was a compelling reason to comment on a correction in a footnote.

In several instances narrators also expressed a wish to rearrange the order of one or two paragraphs in a story in order to correct the sequence of presentation. These rearrangements (e.g., story 5, p. 46; story 153, p. 1273, TNAI ) have been made without editorial comment. Occasionally, too, a narrator wished to add some detail to a story that he or she neglected to mention in the recording. When the addition was just a single word, it was silently inserted in such instances, but when a narrator dictated a sentence or paragraph to be added to a story, its insertion has been specified in a footnote.

Variant usages among narrators

Careful examination of these texts will reveal occasional differences in usages among narrators, many of which are indicated in footnotes. Some differences are clearly the remnants of older dialectal splits within the Arikara community, while others seem to be simply idiosysncratic, although they, too, may be older dialectal forms. An example of the latter is the word for ‘ball,’ which for some speakers is xáwos and for others is xáwes. A similar type of variation is illustrated by the compound meaning ‘smokehole,’ for which some speakers like Mrs. Waters use the form suuxaakAhíniʾ (composed of isuux- ‘nose’ + haaka- ‘mouth, oriface’ + -hiniʾ LOC), while others like Mrs. Brave reverse the order of the constituent nouns and use the form haakeesuuxíniʾ.

A more obvious remnant of a former dialectal difference within the Arikara speech community is preserved in several common words based on the northern Caddoan stem *yaak ‘wood.’ These words exhibit a variation between stem initial h and n, both reflexes of *y, with some speakers using underlying naak- and others haak-. Thus, for Mrs. Waters and many other speakers the form for ‘saddle’ is NAhnaaničitawíʾuʾ, whereas for Mrs. Brave and others it is haahnaaničitawíʾuʾ. For these speakers there is the same variation in words like ‘cupboard,’ which is nakataáRIt and haakataáRIt, and in ‘pipe,’ which is NAhnaaWIškáhtš and haahnaaWIškáhtš, respectively. (For many younger speakers whose families have apparently used NAhnaaWIškáhtš for ‘pipe,’ the form has been reduced to naaWIškáhtš.)

Another example of variant usage is illustrated by differences in vowel length in a word like ‘womankind,’ which for some speakers is sapaahnoóčI and for others is sapAhnoóčI. These two forms suggest competing or different underlying forms for the stem meaning ‘woman,’ one with a long aa in the final syllable and the other with a short a, namely, sapaak and sapak. Such examples, however, are not common.

An illustration of a common difference in speaker usage occurs in the combination of two verbal prefixes, witi- ‘reflexive; reciprocal’ + wi- ‘quotative,’ preceding the third person indicative modal prefix ti-. For older speakers this sequence reduces to wiwi-, whereas for younger speakers the quotative never occurs with witi-. In the word noowiwituuniéwat ‘then he told about himself’ (story 85, p. 911, TNAI ), for example, Mrs. Waters uses the contracted form wiwi, preceding the indicative form tuuniéwat, whereas in the form noowititiisiNIstarékUt ‘then he pulled his nose off’ (story 132, p. 1141, TNAI ), Mr. Howling Wolf typically omits the use of wi- with witi-. Some speakers alternate between use of the contracted form of the two prefixes and use of only the reflexive.

Several differences in pronunciation among speakers created by recent, on-going phonological innovations in Arikara are noted in the section on orthography below.

Orthography

A systematic phonetic orthography is used for presenting the texts. It is the alphabet that was established in 1975 for Arikara language instruction on the Fort Berthold Reservation and that has been used subsequently in all printed teaching materials.² The orthography differs from a phonemic writing system only by its indication of vocalic and consonantal devoicing.

There are twelve consonants in Arikara. The stops and one affricate (p, t, k, č) are voiceless and unaspirated. All of the fricatives are voiceless, too. Three of them (s, š, h) are like their English counterparts. The fourth one, a velar fricative x, has two phonetic variants. One, [x̯], is prevelar in its articulation, giving the sound a more fronted, softer pronunciation that occurs whenever it follows the vowels u and uu. When any other vowel precedes x, this sound has a mid-velar variant [x] that is more strongly articulated.

Glottal stop is another occluded sound. It occurs in only two environments: intervocalically, where it functions phonetically to separate two vowels; and in word final position before a preceding vowel, which it protects from devoicing. The symbol for glottal stop is a raised comma, ʾ.

There are three resonants, of which two, w and n are like their English equivalents, while r differs. The latter is a voiced apico-alveolar tap that occurs only intervocalically in words. In other phonological environments (i.e., initially or preconsonantally) r becomes n.

One notable contrast among Arikara vowels is length. There are five short vowels (i, e, a, o, u) and five long ones, which are written doubled (aa, ee, ii, oo, uu). The primary difference between the short and long varieties is duration of articulation. A long vowel is prolonged when uttered, so that its pronunciation lasts about twice as long as its short equivalent. The short vowels i and u vary in articulation between high [i] and lower-high [ɪ] variants, while their long counterparts ii and uu are always high. Similarly, the articulation of short e ranges between higher-mid [e] and lower-mid [ɛ] tongue height positions, but tongue height for its long counterpart ee is always higher-mid. The articulation of the back vowel o, in contrast, is always higher-mid in both its short and long varieties. Short a has low [a] and mid [ə] variants, too, but long aa is always articulated low.

A pervasive phonetic feature of Arikara is devoicing of certain vowels and resonant consonants, creating whispered sounds. This process occurs with three vowels (i, a, u), which are written with capital letters (I, A, U ) when they are voiceless. Whenever a voiceless vowel is preceded by a resonant (n, r, w ), that consonant is devoiced, too, and is indicated by a capital letter (N, R, W ). Vocalic and resonant devoicing are generally predictable phonological processes; thus the voiceless sounds indicated by capital letters are not phonemic. They are nevertheless indicated in the orthography because the rules for devoicing are sufficiently complex that all of their occurrences are not easily reconstructed, and their phonetic specification as whispered facilitates proper Arikara pronunciation.

Also represented in the texts are two anomalous vowels that are restricted to certain characteristic forms. One is ą, a nasalized low mid vowel, that occurs only in the word hąąʾ ‘yes,’ used by females. (The male form is heeʾ.) The other vowel is æ, a low front unrounded vowel, that appears long in the exclamatives hææ ‘hey, what’ and čææ ‘hey, what!’

Stress is marked by an accent over the vowel; on long vowels the accent is placed over the second member of the pair. Although there are exceptions to the rule, stress generally occurs on the penultimate syllable of Arikara words.

Phonetic processes

An orthographic convention used here is the indication of the two vowels i and u when they are elided in certain contexts. Whenever a prefix ending with an unstressed i preceded by t is joined to a following morpheme with initial n, the i is elided and n becomes syllabic. It is, however, retained in written Arikara as a small raised ⁱ, as in tⁱnaákUx ‘this one’ and atⁱnáʾ ‘my mother.’ Similarly, u is frequently elided in the same phonological context as i. Unlike i, however, which is rarely articulated in this environment, u is only optionally dropped, generally in fast speech. Thus one hears, for example, both tunaahé and t^unaahé ‘it is good,’ depending on speech tempo.

In certain contexts, voiceless vowels condition regular phonetic changes that are not indicated in the orthography. One change occurs when an unstressed i is followed by a sequence of h and a non-front vowel (i.e., a, u, o). There the i devoices and the positional features of the syllabic peak represented by it shift to the following spirant h, palatalizing it. For example, NIhúhtš is pronounced [Nh^yúhtš]. Similarly, tIhuúkAt ‘he came inside’ is pronounced [th^yuúkAt], and tIhoowiisákUx ‘he is sitting on the bank’ is [th^yoowiisákUx].

A similar process occurs with voiceless U. When unstressed u precedes a sequence of either h or x and a non-back vowel (i.e., i, e, a), the u devoices and its positional features then shift to the following segment, the spirant h or x, where they become a labialized offglide. The form naakUheešá ‘when morning came,’ for example, is pronounced [naakh^weešá] and nakuhkUxáxI ‘for them to run’ is [nakuhk^wáxI].

Still another similar phonetic process involves voiceless A. When unstressed a is followed by x and the vowel i, the a devoices and its positional features shift to the following segment, x, where they become a retroflexed offglide. The word kAxiítš ‘bag,’ for example, is pronounced [kx^riítš], and kaakAxíhtoʾ ‘you sure are hateful’ is phonetically [kaakx^ríhtoʾ].

After a voiceless consonant and preceding h plus any vowel but i, a devoices and then the voiceless A drops out, creating an aspirated consonant. Illustrations of this process are tAheéšaʾ ‘morning came,’ pronounced [t^heéšaʾ] and tAhuúkaʾ ‘he came inside,’ pronounced [t^huúkaʾ]. In the presentation of the texts here, this voiceless A is frequently written even though it is not pronounced.

Phonetic Innovations

There are several innovations in Arikara speech that are illustrated sporadically in the texts. One is the loss of word final t when it is preceded by a devoiced vowel. The word tikoótIt ‘he killed it,’ for example, is often pronounced tikoótI. For many younger speakers the loss of t in this environment is complete, but for most older speakers the sound is usually retained in careful speech, as it is throughout most of these texts.

Two other phonetic changes illustrated in the texts are vocalic shifts that occur when vowels become devoiced. One is the raising of voiceless [A] to [I], generally before h and t. For example, tatuuxátkoʾ ‘I heard it’ in discourse form is tatuuxItkó when another word follows it in a phrase. Similarly, wetAtwóʾ ‘I am going’ is often pronounced wetItwóʾ, and NAtkuwaákA ‘for me to say’ generally is NltkuwaákA. The other shift is lowering of voiceless U to A, as in wekuwituutAhuúnuʾ ‘he continued on’ (see p. 63, TNAI ), which is formed from the stem ut ... uhuun-uu.

Text Presentation

Arikara is a polysynthetic language in which verbs, the most elaborate class of words, are composed of long strings of morphemes signifying an array of grammatical and semantic categories. Person, number, possession, benefaction, mode, tense, and evidence are all marked by prefixes, and they in turn are usually preceded by one or more deictic and adverbial clitics. Verb stems, which occur after these prefixes, are always followed by suffixes denoting aspect.

Noun incorporation is another characteristic of Arikara verb morphology. Many nouns that are grammatical objects of transitive verbs or subjects of intransitive verbs are stripped of any occurring nominal suffix and inserted into the verbal complex following the prefixes and preceding the verb stem. If an incorporated noun is plural, it is immediately followed by the pluralizing morpheme -raan-.

Further characterizing Arikara is a complex set of phonological rules that fuse prefixes and suffixes together with one another and with verb stems, producing fully formed words whose underlying form (or morphemic constituency) is not always immediately apparent to a nonspeaker.

The paragraphing of the interlinear texts matches the paragraphing of their free translations appearing in volumes 3 and 4 of Traditional Narratives of the Arikara Indians. Because paragraph boundaries coincide only partially with pauses in narrative performance and with such morphological criteria as particles marking their beginning or end-and thereby defining them as structural units-pause and morphological elements have not been sufficient to stand alone as criteria demarcating paragraph breaks. Hence paragraph units are based on semantic or content criteria as well, and indicate a variety of transitions, such as actor or dialogue switches, topical changes, and temporal or scene changes.

Punctuation of the Arikara line of text differs markedly from that of the English line. For the latter, standard English punctuation is used, except for two special conventions: (1) all quoted material, spoken or thought, is preceded by a colon; and (2) subjunctive form verbs expressing something wished or unrealized take, depending on context, a first or second person subject, just as if they were being spoken or thought by the person who is the topic. These verbal forms are not actually quotes, but since they function like them grammatically, they are set off in the translations by single quotation marks. Thus, for example, a sentence like the English ‘I want to go to town’ translates literally in Arikara as “I want ‘that I would go to town.’” Or an English sentence like ‘Scalped Man was holding the crowd back so that they could not shoot Raven’ is literally in Arikara “Scalped Man was holding the crowd back ‘that you (pl) should not shoot Raven.’”

Punctuation in the Arikara line is limited to the period, which marks the end of a sentence. The use of only this one symbol is motivated in part because there is no written Arikara literary tradition and in part because Arikara morphology renders superfluous some English punctuation conventions or because certain Arikara orthographic conventions could be visually confusing when combined with similar symbols in English punctuation. Such English forms as the exclamation point at the end of an imperative sentence or the question mark at the end of an interrogative sentence, for example, are unnecessary in Arikara since those grammatical forms are morphologically marked by particles and special punctuation for them would be redundant. To avoid confusion with the raised comma representing the glottal stop, quotation marks, single or double, are not used in Arikara. Similarly, the first word in a sentence is not capitalized, since capital letters are used to symbolize voiceless vowels and resonants, which not infrequently begin a sentence and so are already capitalized.

The Translations

The English translations appearing under Arikara words are literal, but they are not morpheme-by-morpheme glosses. They have been cast into well-formed English so that they can be read and understood without confusion, yet they are sufficiently literal to convey their full grammatical and semantic content.

One grammatical form has, however, been incorporated into the translations. In Arikara verbs there are three distinctions made in the category of number: singular, dual, and plural. In most contexts the distinction between singular and nonsingular subjects and objects can easily be inferred from context, but the difference between dual and plural is not always obvious from context alone. To avoid number ambiguity in English translations, the dual is always indicated by its abbreviation in parentheses, (du ), immediately after the personal pronoun(s) in the translation, while the plural is unmarked unless its specification is needed for clarity. When plural number occurs, its abbreviation, (pl ), appears after the personal pronoun.

In the interlinear format, each Arikara word is usually matched by an English translation immediately under it. There are, however, a small number of particles that are not easily given independent translations but which are rendered more precisely into English if they are written as a unit with the following or preceding verb and the entire unit assigned an English gloss. In such phrase units the Arikara particle and verb are separated by one space (whereas all other Arikara words are usually separated by two or more spaces) and the English translation underneath the phrase is cited as a single running gloss. The following are five of the more common particles that occur in such phrase units:

    číkuʾ   ‘indefinite object; indefinite time’
        číkuʾ nookanawitiihunuuwá   ‘there there was nothing going around’
        číkuʾ niikohnuutAxítIt   ‘whatever happened’
    tákuʾ   ‘indefinite person’
        tákuʾ kanawituxkaákUx   ‘no one was inside [the room]’
    na   ‘and’ (combined with a third person subject indicative verb
      to form a mild imperative, as in giving instructions)
        na kaakikaskáWI   ‘don’t look back; you should not look back’
        na nikutaáʾ   ‘this is where you should come’
    saxtš  ‘a little’
        saxtš neewitoókUt  ‘as soon as it is noon’
    waáwi  ‘maybe’
        číkuʾ nii waáwi kohnuúta   ‘whaever he might have done’

In addition to these particles which are treated as phrase units, there is also a double verb construction that is written as a unit. It consists of a verb in the indicative mode followed by one in the infinitive mode, and translates into English as ‘to be able to . . .’:

        kaakíʾ naakuriiwaníka  ‘it cannot miss you’
        nikutíʾ nakuutaánu  ‘that is what he is able to do’

Analyzed Texts

There are five lines in the presentation of each analyzed text. The first line, which is printed in boldface, is the same as the one appearing in the interlinear version; it consists of each word transcribed in its isolation form in a broad phonetic orthography. The second line gives the morphemic constituency of the words above it. Morphemes in this line are given in their underlying, or morphophonemic, form. Two morphemic boundaries are indicated here: plus (+) juncture marks inflectional boundaries between affixes and stems as well as the boundaries between clitics and other prefixes; dash (-) juncture separates derivational morphemes within stems or elements within compounded clitics or prefixes that function as semantic units. Line three presents morpheme-by-morpheme glosses of line two. These glosses are a mixture of grammatical abbreviations and lexical translations. Finally, line four repeats the literal translations of each word as they appear in the interlinear versions of the texts.

A list of abbreviations used in the presentation of the analyzed texts is given at the beginning of each volume. Grammatical terminology is the same as that used in previous publications on Arikara and Pawnee (Parks 1976, 1979).

Oral Recordings

When comparing a sound recorded version of a story with its written version presented in these volumes, the listener must bear in mind the differences between discourse and isolation forms discussed above, since the written words are presented as isolation forms that sometimes differ slightly from the way they are pronounced in actual discourse. One must remember as well the editorial principles that have guided the written presentation and that create other minor differences from the spoken text.

List of Tape Recordings