Plains Apache language
The Plains Apache language, also known as Kiowa Apache or Na’isha, is a Southern Athabaskan language historically spoken by the Plains Apache people, a small Native American tribe residing primarily in southwestern Oklahoma. As the most divergent member of the Apachean subgroup within the Athabaskan language family, it features distinctive phonological elements such as pitch accent and a highly complex verb morphology, including classificatory verb stems that encode information about the shape, consistency, or movement of objects. The language became extinct as a fluent spoken tongue following the death of its last native speaker, Alfred Chalepah Sr., in 2008, though semi-speakers and community members retain partial knowledge, and documentation efforts support ongoing revitalization initiatives.[1][2][3][4] Plains Apache belongs to the Eyak-Athabaskan language family, which spans from Alaska to the southwestern United States and northern Mexico, with its closest relatives including Jicarilla Apache and Lipan Apache in the eastern Apachean branch. Unlike the more widely spoken Navajo or Western Apache languages, Plains Apache developed in isolation on the Great Plains, leading to unique lexical borrowings and adaptations influenced by prolonged contact with non-Athabaskan neighbors like the Kiowa. Its grammar is characterized by subject-object-verb word order, inalienably possessed nouns requiring specific prefixes, and postpositions to mark grammatical relations, all hallmarks of Athabaskan structure. Verbs, the core of the language, incorporate up to 15 prefixes in a templatic system to indicate tense, aspect (such as imperfective or perfective), subject, object, and classifiers, enabling nuanced expression of actions and events.[1][5][3] The Plains Apache people, whose Athabaskan-speaking ancestors migrated southward from western Canada to the Southern Plains by around A.D. 1400, adopted a nomadic Plains lifestyle centered on buffalo hunting and tipi dwellings by the 16th century. Historically allied with the Kiowa since the late 17th century, they maintained their linguistic distinctiveness despite cultural assimilation, using the language for oral traditions, ceremonies, and daily communication until English dominance in the 20th century accelerated its decline. Documentation began in the mid-20th century with linguists like Harry Hoijer, and recent projects, including NEH-funded work in the 2000s, have archived texts, dictionaries, and recordings to aid tribal revitalization programs. Today, the Apache Tribe of Oklahoma, with headquarters in Anadarko, continues cultural preservation efforts, emphasizing the language's role in identity despite its dormant status.[6][7][2][5]Classification and History
Linguistic Classification
The Plains Apache language, also known as Kiowa Apache, belongs to the Southern Athabaskan (Apachean) subgroup of the Athabaskan language family, which forms part of the Na-Dene phylum.[8][4] This classification positions it among a diverse set of indigenous languages spoken across North America, with Athabaskan languages extending from Alaska and Canada to the southwestern United States and northern Mexico.[8] Within the Southern Athabaskan subgroup, Plains Apache stands as the most divergent member, alongside Navajo, Western Apache, Chiricahua Apache, Mescalero Apache, Jicarilla Apache, and Lipan Apache. Its closest relatives within the eastern Apachean branch are Jicarilla Apache and Lipan Apache.[1][4] This divergence reflects an early split from proto-Apachean, as identified through comparative linguistic analysis.[3] Glottochronological studies estimate that Plains Apache linguistically diverged from other Apachean languages around 1300–1500 CE, aligning with ancestral migrations to the Southern Plains region.[9] These innovations include distinctive phonological shifts and morphological patterns, such as developments in its register tone system featuring high and low tones, which set it apart from closer relatives in the subgroup.[3]Historical Development
The Plains Apache people, also known as the Kiowa Apache, originated as an Athabascan-speaking group near the eastern Rocky Mountains and continued their southward migration in the 17th century, ranging between the Platte River and eastern New Mexico by acquiring Spanish horses for trade.[10] By the 18th century, they had reached the Southern Plains, including present-day southwestern Oklahoma and the Texas Panhandle, where they allied closely with the Kiowa for protection against larger hostile tribes, adopting a nomadic lifestyle centered on buffalo hunting.[10] This migration marked their distinct adaptation to Plains culture, distinguishing them from other Apache groups in the Southwest while retaining their linguistic roots.[10] Historical documentation of the Plains Apache language emerged in the 19th century through ethnolinguistic records collected by explorers and anthropologists encountering the group during their nomadic phase. Early mentions appear in explorer accounts, such as Lewis and Clark's 1805 journals noting the tribe in the Black Hills region.[10] More systematic linguistic data was gathered in the late 1800s by anthropologists like James Mooney, whose fieldwork among the Kiowa and their allies included vocabulary and cultural notes on the Plains Apache during Smithsonian expeditions. These records captured the language's use in daily communication, rituals, and intertribal sign language interactions on the Plains. The U.S. government's confinement of the Plains Apache to reservations in western Oklahoma following the 1867 Treaty of Medicine Lodge fundamentally shifted their nomadic lifestyle to sedentary reservation life, confining them to the Kiowa-Comanche-Apache Reservation until land allotment in 1901.[7] This transition, enforced through military containment and federal policies, disrupted traditional practices and accelerated language loss as English became dominant in interactions with settlers and officials.[11] Assimilation efforts, including mandatory boarding schools from the 1870s onward, prohibited native language use, further eroding fluency among younger generations and tying linguistic decline to broader cultural suppression.[11] Key documentation milestones in the 20th century included extensive fieldwork by anthropologist William E. Bittle in the 1960s, who, along with his students, compiled field notes, oral interviews, and texts on Plains Apache culture, kinship, and language from Oklahoma speakers.[12] Bittle's work preserved grammatical structures, ethnobotanical terms, and narratives, providing a critical archive amid declining speakers.[12] The language reached extinction in 2008 with the death of Alfred Chalepah Sr., the last fluent speaker, marking the end of natural transmission after over two centuries of socio-historical pressures.Phonology
Consonants
The Plains Apache language possesses a large consonant inventory aligning with patterns observed in other Southern Athabaskan languages, featuring contrasts in aspiration, glottalization, and laterality among stops, affricates, fricatives, nasals, and approximants.[13] These consonants are articulated at bilabial, alveolar, postalveolar, palatal, velar, and glottal places of articulation, with ejective and aspirated series distinguishing obstruents.[13] The inventory includes a prenasalized alveolar stop /ⁿd/, which occurs exclusively in prefixes.[13] Notable allophonic variations affect the glottal fricative /h/, which alternates freely with the velar fricative in preconsonantal position before /a/; /h/ does not occur before /o/.[14] Labialization appears in contexts involving the approximant /w/, where it influences adjacent velars, resulting in secondary articulation effects such as [kʷ] in certain syllable onsets.[13] Sibilant harmony, a key phonological process, involves long-distance assimilation between alveolar (/s, z, ts, tsʰ, ts'/) and postalveolar (/ʃ, ʒ, tʃ, tʃʰ, tʃ'/) series, often triggered by stem sibilants on prefixal ones in a right-to-left manner, sometimes yielding intermediate phonetic realizations.[14] The following table presents the consonant phonemes, organized by place and manner of articulation, with IPA symbols and approximate orthographic representations based on practical systems used in Athabaskan linguistic descriptions (e.g., Bittle's transcription).[13]| Place/Manner | Bilabial | Alveolar | Postalveolar | Palatal | Velar | Glottal |
|---|---|---|---|---|---|---|
| Nasals | m (/m/) | n (/n/) | ||||
| Plosives (tenuis) | p (/p/) | t (/t/) | k (/k/) | ʔ (/ʔ/) | ||
| Plosives (aspirated) | tʰ (/tʰ/) | kʰ (/kʰ/) | ||||
| Plosives (ejective) | t' (/t'/) | k' (/k'/) | ||||
| Prenasalized | ⁿd (/ⁿd/) | |||||
| Affricates (alveolar) | ts (/ts/) | |||||
| Affricates (aspirated alveolar) | tsʰ (/tsʰ/) | |||||
| Affricates (ejective alveolar) | ts' (/ts'/) | |||||
| Affricates (lateral) | tɬ (/tɬ/) | |||||
| Affricates (aspirated lateral) | tɬʰ (/tɬʰ/) | |||||
| Affricates (ejective lateral) | tɬ' (/tɬ'/) | |||||
| Affricates (postalveolar) | tʃ (/tʃ/) | |||||
| Affricates (aspirated postalveolar) | tʃʰ (/tʃʰ/) | |||||
| Affricates (ejective postalveolar) | tʃ' (/tʃ'/) | |||||
| Fricatives (voiceless) | s (/s/), ɬ (/ɬ/) | ʃ (/ʃ/) | x (/x/) | h (/h/) | ||
| Fricatives (voiced) | z (/z/) | ʒ (/ʒ/) | ɣ (/ɣ/) | |||
| Approximants | w (/w/) | l (/l/) | j (/j/) |
Vowels
The Plains Apache language features a vowel system with four oral vowel qualities: high front /i/, open-mid front /ɛ/, low central /a/, and mid back /o/. Each of these vowels occurs in short and long forms, where length is phonemic and distinguished by duration, as in /i/ versus /iː/. Vowel length plays a key role in lexical distinctions and morphological processes.[15] Nasal vowels are also present, formed by nasalizing the oral vowels and typically represented with an ogonek diacritic in linguistic descriptions, such as /ǫ/ for the nasalized mid back vowel. Nasalization is phonemic and can occur on both short and long vowels, affecting meaning; for example, the word /kóó/ (with long oral /oː/) means 'water', while /nǫ̀ǫ̀/ (with long nasal /ǫː/ and low tone) means 'earth'.[15] Phonetic realizations of these vowels vary by environment. The open-mid front /ɛ/ often raises to or [ɪ] in closed syllables or before certain consonants, contributing to allophonic variation without altering phonemic contrasts. Vowels serve as tone-bearing units, with length influencing tone realization, though tone itself is a suprasegmental feature.[15]Syllable Structure
The syllable structure in Plains Apache adheres to a maximal template of CCVːC, permitting complex onsets and long vowels followed by a coda consonant. This structure accommodates the language's consonant inventory, including affricates and fricatives, while vowels can be short or long. For instance, the word for "pants" is realized as /stłèːh/, exemplifying a cluster onset with /stł/ transitioning into a long vowel and final coda /h/.[16] Onset clusters are phonologically constrained, occurring primarily as an obstruent (such as a stop or fricative) followed by a sonorant (like /n/ or /l/), or a fricative combined with the lateral affricate /ł/. These limitations ensure that initial consonant sequences remain pronounceable within the language's articulatory patterns, avoiding more complex combinations seen in some unrelated languages. Codas, in contrast, are simpler, limited to a single consonant that is often a glide (/w/, /j/) or nasal (/m/, /n/), which facilitates smooth transitions in connected speech.[16] Derivational processes like reduplication and infixation influence syllable count by adding or modifying syllables to convey plurality, diminutives, or iterative actions. Reduplication typically copies the initial syllable or CV unit, increasing the word's prosodic weight, while infixes insert segmental material medially, often expanding monosyllabic roots into disyllabic forms. These patterns integrate seamlessly with the core syllable template, preserving the CCVːC maximum even as word length varies.[16]Tone
The Plains Apache language employs a register tone system consisting of high (H) and low (L) tones, which primarily distinguish lexical meaning on short vowels. Short vowels thus bear either a high register tone or a low register tone, functioning as the basic tonal contrast in the language. This two-level system is characteristic of many Southern Athabaskan languages, where tone serves as a suprasegmental feature to differentiate words that would otherwise be homophonous based on segmental content alone. On long vowels, the tone system expands to include four distinct contours: rising (LH), falling (HL), high-level (HH), and low-level (LL). These contours arise from the combination of the underlying H and L registers across the two morae of the long vowel, allowing for greater phonological complexity in tone-bearing units. For example, the word /tʼǫ́ǫ́š/ 'bark' features a rising (LH) contour on the long nasal vowel /ǫ́ǫ́/, where the tone starts low and rises to high, illustrating how contours can mark semantic distinctions. This contour system on long vowels is a key feature that contributes to the language's prosodic richness, with tones realized acoustically through variations in fundamental frequency (F0). Tone sandhi processes in Plains Apache include assimilation rules that affect tonal realization in compounds and across morpheme boundaries. In particular, adjacent tones may assimilate, such as when a high tone from one element spreads to a following low-toned vowel in compound formations, resulting in a leveled high contour or partial fusion to maintain prosodic harmony. These rules prevent tonal clashes and ensure smooth phonetic transitions in connected speech, similar to patterns observed in related Athabaskan languages but adapted to Plains Apache's specific register and contour inventory. Historically, the tonal system of Plains Apache derives from Proto-Athabaskan, which lacked phonemic tone and instead featured glottalized consonants that initiated tonogenesis through the reanalysis of laryngeal features onto preceding vowels. In the Southern Athabaskan branch, including Plains Apache, this led to the development of register tones, with innovations such as the elaboration of contour tones on long vowels representing a departure from simpler level systems in Northern Athabaskan languages. Plains Apache's increased contour complexity likely emerged as an internal development post-migration to the Plains, enhancing lexical differentiation in a language with a relatively small vowel inventory.[17]Orthography
Writing System
The Plains Apache language utilizes a practical orthography developed by linguist William E. Bittle in the mid-1950s during extensive fieldwork with fluent speakers in southwestern Oklahoma, as detailed in his 1963 publication.[18] This system emerged as part of Bittle's comprehensive grammatical study, aiming to facilitate documentation, analysis, and potential teaching of the language amid declining speaker numbers. Bittle's orthography is based on the International Phonetic Alphabet (IPA) but modified for greater readability and ease of use in non-linguistic contexts, employing the Roman alphabet supplemented by digraphs and special characters. Common digraphs include "ch" to represent the affricate sound /tʃ/ and "ł" for the voiceless alveolar lateral fricative /ɬ/. This approach prioritizes accessibility while preserving phonetic accuracy, making it suitable for transcribing texts and eliciting data from speakers. In comparison to orthographies for other Athabaskan languages, such as those for Navajo or Western Apache, Plains Apache's system shares the Roman base but features unique adaptations, particularly for marking the language's four contrastive tones, which distinguish it from tonal representations in more standardized Southwestern Athabaskan scripts. These adaptations reflect Plains Apache's phonological divergence within the family, emphasizing low and rising tones alongside high and falling ones.Representation of Sounds and Prosody
The orthography of Plains Apache, as standardized in key linguistic descriptions, employs a Latin alphabet adapted to represent the language's complex consonant inventory, including ejectives and affricates, through digraphs and diacritics. Ejectives are denoted by an apostrophe immediately following the base consonant, such as t' for the glottalized alveolar stop /tʼ/ and ts' for the ejective alveolar affricate /tsʼ/. Affricates are typically written as combinations like tl for the voiceless lateral affricate /tɬ/, dz for the voiced alveolar affricate /dz/, and ch for the voiceless postalveolar affricate /tʃ/. These conventions ensure distinct representation of the language's rich series of stops and affricates, which exhibit contrasts in aspiration, voicing, and glottalization. Vowels in Plains Apache orthography are based on four basic oral vowels—a, e, i, o—each of which can be short or long and oral or nasalized (the latter marked with an ogonek diacritic, e.g., ą for nasalized /a/). Vowel length is indicated by doubling the vowel, as in aa for the long low central vowel /aː/. Nasalization is marked with an ogonek (˛) beneath the vowel, producing forms like ą for /ã/ or ąą for /ãː/. For example, the word for "earth" is written nǫ̀ǫ̀, where the double nasalized o with low tone distinguishes it phonologically. These markers allow precise encoding of vowel quality and modification without altering the base graphemes. Prosody, particularly tone, is a critical feature in Plains Apache, which features a register tone system with high and low levels, often realized as level or contour tones on long vowels. High tone is represented by an acute accent (´) on the vowel, as in á for high /á/ or áá for long high /aː́/. Low tone uses a grave accent (`), such as à or àà for low /à/ or long low /aː̀/. In some contexts, mid-level tones or unmarked low tones may appear without diacritics, though primary descriptions emphasize accents for contrast. For instance, the word for "water" is written kóó, corresponding to /kʰóː/ with high tone on the long mid vowel. Tones on long vowels are placed on the syllable bearing the primary pitch, typically the first for rising or level high tones.[5] The following tables summarize the primary mappings between International Phonetic Alphabet (IPA) symbols and Plains Apache orthographic representations for consonants and vowels, based on established conventions. These highlight key contrasts without exhaustive enumeration of allophones.Consonants
| Place/Manner | Unaspirated | Aspirated | Ejective | Voiced | Fricative |
|---|---|---|---|---|---|
| Bilabial | m /m/ | ||||
| Alveolar stop | t /t/ | th /tʰ/ | t' /tʼ/ | d /d/ | s /s/ |
| Alveolar affricate | ts /ts/ | tsh /tsʰ/ | ts' /tsʼ/ | dz /dz/ | |
| Postalveolar affricate | ch /tʃ/ | chh /tʃʰ/ | ch' /tʃʼ/ | j /dʒ/ | sh /ʃ/ |
| Lateral affricate | tl /tɬ/ | tlh /tɬʰ/ | tl' /tɬʼ/ | ł /ɬ/ | |
| Velar stop | k /k/ | kh /kʰ/ | k' /kʼ/ | g /g/ | |
| Velar fricative | x /x/ | ||||
| Glottal | ' /ʔ/ | h /h/ | |||
| Other | n /n/ | ||||
| l /l/ |
Vowels and Prosody
| Oral Vowel | Short High | Short Low | Long High | Long Low | Nasalized Short High | Nasalized Short Low | Nasalized Long High | Nasalized Long Low |
|---|---|---|---|---|---|---|---|---|
| High front | í /i/ | i /ì/ | íí /iː́/ | ìì /iː̀/ | į́ /ĩ/ | į /ĩ̀/ | į́į́ /ĩː́/ | į̀į̀ /ĩː̀/ |
| Mid front | é /e/ | e /è/ | éé /eː́/ | èè /eː̀/ | ę́ /ẽ/ | ę /ẽ̀/ | ę́ę́ /ẽː́/ | ę̀ę̀ /ẽː̀/ |
| Low central | á /a/ | a /à/ | áá /aː́/ | àà /aː̀/ | ą́ /ã/ | ą /ã̀/ | ą́ą́ /ãː́/ | ą̀ą̀ /ãː̀/ |
| Mid back | ó /o/ | o /ò/ | óó /oː́/ | òò /oː̀/ | ǫ́ /õ/ | ǫ /õ̀/ | ǫ́ǫ́ /õː́/ | ǫ̀ǫ̀ /õː̀/ |
Morphology
Nouns
Nouns in Plains Apache display limited inflectional morphology, focusing primarily on possession and derivation, with number often handled through contextual or verbal means rather than dedicated nominal affixes. The language features three main types of nouns based on structural complexity: primary nouns consisting of a root optionally prefixed for possession, compound nouns formed from two or more roots, and derived nouns originating from verbs or phrases. Primary nouns are the simplest, such as séé 'dirt', and can take possessive prefixes to indicate ownership. Possession is marked by optional pronominal prefixes attached directly to the noun stem, reflecting the possessor's person and number. The possessive paradigm includes ši- for first person singular ('my'), di- for second person singular ('your'), bi- or mi- for third person singular ('his/her/its'), go(o)- for fourth person ('someone's, obviative'), ʔi- for indefinite ('someone's'), and dàxi- or da- for dual forms. These prefixes align with the objective pronominal series used in verbs. For example, ši-č'èèčéé means 'my wife', while bi-čìƛ'á̧á̧ means 'his brother'.[15] The language distinguishes inalienable from alienable possession in its morphological expression. Inalienable nouns, such as body parts and certain kin terms, are obligatorily possessed via direct prefixation and cannot stand alone, as in bi-dààh 'his lips' where bi- indicates third person possession of the body part dààh 'lips'. Alienable nouns, including most objects and abstract concepts, may occur unpossessed or use postpositional constructions to express temporary or non-inherent possession, such as combining the noun with a relational postposition equivalent to 'with' or 'of'. This distinction highlights the language's sensitivity to the inherent nature of possessive relationships. Number marking on nouns is not systematically inflected with dedicated suffixes for all categories; instead, plurality or distributivity is typically conveyed through verb agreement, contextual inference, or, in limited cases, reduplication of the noun stem for emphasis on multiplicity. For instance, certain nouns may employ partial reduplication to signal plural or distributive senses, though this is lexical and not productive across the board, with verbs bearing the primary burden of number specification (e.g., plural object prefixes like go-).[15] Derivation of nouns from verbs occurs via nominalization, often integrating a verb stem with a nominal element or through compounding to create descriptive terms. This process allows for lexical expansion, particularly for culturally specific or descriptive concepts. An illustrative example is táłbàyé 'crane', formed by combining tá 'feathers' with łíbàyé 'it is white'. Such derivations rely on the rich verbal system, where noun-verb compounds draw on thematic classifiers from verbs to specify shape or handling properties.[15]Verbs
The verbs of Plains Apache exhibit a highly elaborate templatic morphology typical of Southern Athabaskan languages, incorporating up to 14 distinct prefix and suffix positions that encode polypersonal agreement for subjects, direct objects, and indirect objects, alongside deictic elements, classifiers, and aspectual markers. This structure enables a single verb to integrate multiple syntactic arguments and semantic nuances, such as location or manner, into a compact form.[19] Central to this system is the position-based prefix template, where elements are ordered from outermost (e.g., postpositional or deictic prefixes in positions 12–13) to innermost (e.g., subject pronouns in position 2 and classifiers in position 1), followed by the verb stem and occasional suffixes for mode or iteration. Direct object markers occupy positions 8–9, while indirect objects appear in position 11, allowing for intricate agreement patterns like third-person singular yi- for objects or first-person singular ši- for subjects. No single verb utilizes all positions simultaneously, but the template's flexibility supports rich inflectional paradigms.[15] The classifier system, positioned immediately before the stem, plays a crucial role in verb semantics by indicating the shape, animacy, or plurality of the direct object or postpositional theme, thereby creating classificatory verb constructions. The four primary classifiers are the zero morpheme (ø-) for default transitive or active intransitive forms; ł- for round-shaped, granular, or plural animate objects (e.g., berries, people); d- for slender, flexible items (e.g., rope, belt); and l- for plural or sheet-like/area objects (e.g., plural clothing). These classifiers not only adjust valence but also trigger stem variations to classify handled entities, distinguishing, for instance, "handle round object" from "handle slender object."[3] An illustrative paradigm of polypersonal agreement and templatic layering appears in the perfective form šìyédíʔą̀ą̀ 'he gives it to me,' parsed as follows:- ši- (first-person singular indirect object prefix, position 11)
- yi- (third-person singular direct object prefix, position 8)
- dí- (perfective mode prefix, position 6)
- ∅- (zero classifier, position 1)
- -ʔą̀ą̀ (verb stem for 'give')