The community of Minto lies about 130 miles north-west of Fairbanks, Alaska. This village is known as ‘New Minto’ because it was created, and moved to, in response to repeated flooding of the historical Tanana River site of the village (Menhti), now called ‘Old Minto’ (Map 3). The indigenous name of the site of New Minto is Menok’oget, ‘face gets chapped’, for the sharp winds that blow on the bluff.

The people of Minto are Alaskan Athabascans. Their indigenous culture is characterized by a complex kinship system and seasonal subsistence activities including hunting of moose, caribou, small game and birds, salmon and whitefish fishing, and berry gathering. The people of Minto have been in contact with people of European heritage since the late nineteenth century, when steamboats began to work the Tanana, serving settlers and prospectors during the Alaska gold rush.

Robert Charlie recalls an occasion that proves the importance of composition in Athabascan tradition. He is speaking about his father, Moses Charlie:

Map 3 Map of Alaska, including places and rivers occurring in the text. Note that Tanana has three usages: the Tanana river, the Tanana town and the Tanana language.

As fewer people learn the local language and the traditional customs, many elders are increasingly engaged in preserving their heritage. Robert’s brother Neal Charlie was a traditional chief of the village, and he knew his culture very well. He was of the opinion that singing was a key for preservation; and at a 2005 workshop, he urged researchers to take up the task of music research:

Alaskan Athabascan language and vocal expressions³

Though there are some studies of Athabascan language and music, there was very little, initially in our research, about the interrelation between the two. It was known, however, that the vocal expressions that are studied here were composed. Even though the individual compositions are formulaic to a certain extent, it was not certain whether performances would be useful in the analysis. For this reason, a primary objective in this chapter is to try out how the possible performance templates are constructed, and whether new knowledge can be achieved by using them. Since so little was known about the object of study, the transcription models used are graphs and music notations.

Athabascan languages are spoken in Alaska and also include Navajo and Apache, as well as certain languages of the west coast of North America.⁴ All the eleven Athabascan languages spoken in Alaska are endangered, though two languages, Koyukon and Gwich’in, have more adult speakers than the others. Koyukon is spoken in 15 villages in the western part of Alaska. The Gwich’in people live in and near the Brooks Range and sections of the Cordillera, and also in the valleys of the middle Yukon and lower Mackenzie.⁵ The Gwich’in village Venetie is on the Chandalar River, a tributary to the Yukon. The Tanana people live on the middle stretch of the Tanana River, which flows into the Yukon. The last surviving dialect of Tanana, Minto (Lower Tanana), is spoken in the village of Minto. There are estimated to be 150 speakers in Koyukon and 25 in Lower Tanana,⁶ but these estimates may be over-generous, since all three languages continue to suffer significant loss of elders. Gwich’in (Venetie), Tanana (Minto dialect), and Koyukon are all considered to be part of the Central Alaska-Yukon sub-group of Northern Athabascan.⁷

As elsewhere in the western part of the United States, many speakers of Alaska’s indigenous languages were relocated as children to boarding schools, where the use of their native languages was strongly discouraged. This assimilative practice continued into the 1970s, when bilingual education was introduced as a new approach.⁸ Speakers of Alaska’s Native languages were recruited to work in bilingual programmes, though these programmes still had the goal of maximizing the use and learning of English, and not of promoting bilingualism in students. Time was set aside in some schools, making use of semi-volunteer speakers, to provide support for students who came to school with limited English.

Figure 3 Elders listen to, and comment on, archival recordings at Neal and Geraldine Charlie’s house in Minto. Left to right: Bergman Silas, Sarah Silas, Geraldine Charlie and Neal Charlie. August 2010.

As the Native languages became less used, owing to attrition in the older generations and stigmatization in the majority-language community, the focus in schools has shifted to support for language learning by children whose first language is English. In some areas, speakers of Alaska Native languages have succeeded in gaining educational certification and are able to work with young bilinguals and English speakers alike. In other areas, partly because of the small size of individual language communities, there is minimal representation by speakers in standard certified teacher positions. A number of Athabascan speech communities are of this type.

Old ideas about bilingualism die hard. Many teachers and parents believe that students will not speak English well, and hence will not be successful, if they also speak their Native language (or any other language besides English). Bialystok refers to a ‘folk wisdom of childhood bilingualism’, which causes fears that acquiring more than one language in childhood could cause linguistic confusion or even general cognitive difficulty.⁹ In her studies, she finds a cognitive benefit to childhood bilingualism, beyond the obvious one of knowing more ways of talking about things. Balanced bilinguals, people with equal proficiency in two or more languages, seemed to be better at metalinguistic tasks. Whether this would be considered an important benefit by teachers and parents in Alaskan villages depends on how much they value metalinguistic ability – the ability to observe and analyse patterns in language or, basically, to understand grammar. Bialystok makes it clear that situations where bilingualism is not balanced may not provide any such benefit, even as regards awareness of language.¹⁰ In many Athabascan families in Alaska’s interior, some knowledge of and exposure to the heritage language can be heavily outweighed by constant and insistent exposure to English.

For young Athabascans in Alaska’s interior, the result is a varied landscape of language opportunities, only a few of which might result in strong proficiency in the structure and pragmatics of the heritage language, and only one or two of which could be available for a particular child. In the experience of the authors, the young people who become most successful in learning their heritage language do so outside the setting of formal education. They achieve this by actively apprenticing themselves to elders who, in turn, take on the responsibility of providing the input the learners need by keeping to their native language when they are in their apprentice’s company.

There is another context that seems to result in proficiency, at least for people with a strong passive knowledge of a related language from their childhood. Teaching a language one does not really speak, with constant reference to and study with an elder, seems to bring some people to a high level of proficiency. Again, it is time with the language and an ambition to participate in it that makes the difference – although for teachers, attaining literacy in the language may also facilitate metalinguistic awareness and improve their ability to notice patterns.

It should not surprise anyone that a person can be hired to teach a language they do not really speak. It is not uncommon in US schools for non-native speakers to teach (at some level) a language they learnt in college. In the context of Alaska Native languages, a desire to provide local language content in class produces the most able and willing volunteer, who may well be a certified teacher in some other field or a dedicated learner of a highly endangered language.

Northern Athabascans and music

Northern Athabascans today are in contact with a multitude of musical styles.¹¹ Among those that have been incorporated over a long period of time are Christian hymns that were given texts in Athabascan languages in the early decades of the twentieth century and are still being used by the elderly. Besides, there is old-time-style fiddling, especially strong in the Gwich’in community, and another strong component is country music. This is often performed in English, and also in Athabascan and other Alaska Native languages. The transmission of oral traditions in Athabascan communities in Alaska – including speaking and writing in the heritage language – is sometimes supported by explicit instruction in school programmes. Child-directed classroom songs with Athabascan lyrics, set to common Euro-American melodies (mainly nursery rhymes), are used in language classes with students of all ages. There are projects that have produced learning materials in the form of traditional songbooks, with cultural information sometimes combined with recordings of performances.¹² These materials are mainly accessed by those involved in documentation, but sometimes language and music learners find them useful too.

Music and dance are sometimes included in after-school programmes. In the Minto village, for example, students have participated for decades in an active after-school programme led by two generations of fluent elders who were also song-leaders and song-makers. While indigenous vocal expressions are rarely part of the formal curriculum of the Minto School, dancing and singing practice has been nearly as prominent as basketball as an extra-curricular activity. The Minto Dancers have participated in the University of Alaska’s Festival of Native Arts for many years, demonstrating the power of their local music tradition.

Tanana (Minto dialect) and Gwich’in (Venetie dialect) have developed differently with respect to music and dance, following contact with white cultures. The Gwich’in encountered Europeans (French and English speakers) in the nineteenth century and developed their own Western musical styles, adding violin and guitars to their traditional vocal and drum tradition.¹³ Their older traditional music is less practised today than that of the Tanana, which has retained a strong native musical tradition until the current generation. The recently active generation of elders in Minto has now been nearly completely depleted, as the majority of the group has passed away, and the remaining elders are either ill or inactive in cultural revitalization. Younger people, generally speaking, are not acquiring the language as children, although they are studying the musical tradition.

Gwich’in consulted for this study say that they have not heard much traditional singing, but there are more young speakers of Gwich’in than of Tanana. Accordingly, data for Tanana include recordings made in 2005–12, while the Gwich’in data was recorded in 1972 and archived at the University of Alaska Library as a part of a ‘Songs and Legends’ collection.

The above summary may suggest that interior Athabascan language, culture, and musical traditions are in danger of disappearing with nearly only archived materials left behind. This is not really the case, though it is not clear yet just how successful current language and culture revitalization efforts will prove. A series of ambitious projects conducted by the Doyon Foundation, which represents most of the Athabascan languages in Alaska, has resulted in increased activity in language learning, teaching, and materials creation. Other political developments, especially among Alaska Natives interested in reforming the educational system, may also contribute to more active transmission of language and culture in these communities.

Traditional Northern Athabascan music is predominantly vocal, with or without rhythm accompaniment on drums or on other instruments. This music plays an important role in the festivities called potlatch, including, but not limited to, funeral and memorial potlatches. On these occasions people from different villages, or even from different language areas, meet and take part in the singing and dancing. This is one situation in which vocal expressions are learnt and spread.

A memorial potlatch will normally include a dratakh ch’elik, ‘mourning song’, recently made in honour of the person who passed away.¹⁴ A dratakh ch’elik performed by the potlatch host is referred to as the khwtitl ch’elik, ‘potlatch song’, and has a special status. Certain older memorial dratakh ch’elik that are remembered will be performed as well, some dating back to the early 1900s. The potlatch will then continue with dancing and the performance of ch’edzes ch’elik, ‘dance songs’, to drum accompaniment. In many places, like Minto or other Tanana River villages such as Tanacross, the potlatches provide a context where these kinds of vocal expressions are required. This function has resulted in their being actively used and – not least importantly – newly composed.

The Minto-Nenana dialect and music

The Tanana Athabascan language of Alaska is today represented by the Minto-Nenana dialect, spoken only by people from Minto (Menhti), a village 200 kilometres from Fairbanks, Alaska. Minto is a Northern Athabascan language, closely related to Koyukon and Ahtna, among others. The Minto people are Athabascan Indians. Their traditional economy is based on hunting game, fishing, and gathering berries. Vocal expressions remain an important part of the cultural life of Minto, and the elders are actively engaged in composition and documentation.

Athabascan languages, including Minto, are typologically unusual in structure and known for the complexity of their verb morphology. As polysynthetic languages, they are characterized by long words that may contain multiple word roots and affixes that would occur as separate words in more isolating languages, such as English or Chinese. Many are also tonal languages, with different levels of density of tone. Minto has a sparse and variable low tone from historical vowel laryngealization, and a high-rising tone associated with negative utterances and with certain morphemes. All other pitches in conversational language are provided by intonational patterns that are similar to those observed in many languages of the world: unit-final lowering, pitch and duration changes related to emphasis, and intentional use of pauses.

Linguistic tone in the Athabascan languages of Alaska varies in its realization. Minto Tanana has relatively few low-toned syllables and some high tones, while Gwich’in syllables are more stable in both low and high tones. Some tonal syllables are heard in dialects of Koyukon, but the language as a whole is not classed as tonal. In both Minto Tanana and Koyukon, intonation provides most of the pitch patterns in speech.

Minto has a stress pattern in spoken language that makes word roots and certain vowels more prominent: in musical settings, the rhythm of words is sometimes subordinated to the musical rhythm. However, this depends on the genre. In those we will examine in this chapter, most words carry the rhythm you would hear if they were spoken. However, certain important words are elongated and performed with a special voice quality, a throbbing glottal pulse that keeps very clear time with the beat of the rhythm.

The intonational system of Minto Tanana is characterized by final lowering at intonational phrase boundaries, often but not always with a measurable pause. Intonational low tones are lower than low lexical tones,¹⁵ usually registering the lowest pitch within an intonational phrase. Utterance types (such as questions) are marked by particles and affixes, with intonation playing a secondary role. Emphasis may be expressed with high pitch, and this type of intonational effect can also override the expression of lexical tone. Intonational domains that are smaller than the intonational phrase have not been demonstrated for this language, but are likely to exist, given findings for the closely related Dena’ina.¹⁶

The interaction between lexical and intonational tone patterns in Minto Tanana makes lexical tone challenging to hear and to learn. Moreover, because of rich inflectional morphology and grammatical patterns, lexical tone bears almost no functional load; it is simply part of the pronunciation of words. However, we find that lexical tones are recognized in more speech-like forms of music, as will be shown below.

The traditional music of Minto is vocal and often accompanied by an even drumbeat. The vocal expressions vary in style between different genres, ranging from those using words that express a feeling to those for dancing made up entirely of vocables with no lexical meaning. Similarly, the music ranges from slow and mournful melodies to up-tempo strophic forms with a distinct melodic contour: starting high, descending stepwise, and ending with a tone repetition at a low pitch. The Minto tradition shows many similarities with other Athabascan music and with Native American music in general, but it also has its particular character.¹⁷ Categories of vocal expression have been defined on the basis of interviews with native speakers and practitioners of vocal art, and these categories are reflected in terminology within the language (Table 5).

The English translations reflect the structure of the word ch’elik, which is formed from the verb ‘to sing’, as in ch’edelik, ‘she or he is singing’. Since all the people are English-speaking, the English forms are also used in the indigenous culture. This means that terms such as dance song, potlatch song are also seen as indigenous, as well as the names of specific vocal expressions, like Raven song or Caribou song.

The structure of Tanana words for vocal expressions ranges from relatively prose-like to strictly metred, depending on the function and situation. Dratakh ch’elik for mourning displays prose-like qualities, with melody following tone and intonation to a greater or lesser degree, while ch’edzes ch’elik for dancing is highly rhythmic, and the speech tone and intonation are not expressed directly.

A dratakh ch’elik basically consists of two parts, the first of which contains key words expressing relationship (like ‘my father’, ‘my child’) in combination with vocables. The second part contains words honouring the deceased which are repeated three or more times with much parallelism, ending with vocables. The ch’edzes ch’elik have fewer words with lexical meaning, often one or two words that appear at the very beginning of a stanza and are followed by longer passages of vocables, while the melody is usually falling and ends on a tone repetition on a low tone. This stanza will then be repeated a number of times without changes of words or vocables.

Though there are few words with lexical meaning in dratakh ch’elik, the words are extremely important and carefully thought out. The combination of the initial key words and music is also significant: usually a short but poignant melodic and rhythmic motif that is often repeated before the performance continues. There are examples of the same word in different musical settings, so even though there is a fairly close relationship between speech and language in vocal performances, there are big differences as well. That is to say that melodic shape is not driven by word prosody, such as lexical tone or morphological stress, although there is an interaction between music and prosody.¹⁸ The parts of the performance that are totally based on vocables tend to be more formulaic, while still unique for the specific vocal expression.

The musical traditions of the Athabascan people have endured through centuries of contact with European cultures in all the geographical regions where this family of languages is spoken. Vocal expressions are performed using Athabascan languages even where the language is no longer transmitted to the new generation. In the context of extended, intensive contact with Euro-American language and culture, the health of the Minto vocal tradition is particularly notable. While there have clearly been many talented composers and performers in Minto in the twentieth century, we believe that this strong tradition also rests on structural components in composition and lyrics. Describing the important aspects of this system may serve to help support a continuation of the tradition.

Our concern with this music has been strongly encouraged by Minto elders, particularly the late Neal Charlie, who believed that composition and performance could help young people to learn the spoken language. In particular, Mr Charlie believed it was important to work with dratakh ch’elik, because their words contained important advice for living a good life – expressed, of course, in the Minto language. Neal and his wife, Geraldine, along with fellow Minto elders Bergman and Sarah Silas, contributed greatly to the documentation and maintenance of the Minto vocal traditions, and they worked with us in this project. All of them have now passed away.

Our linguistic analysis starts with the translation of words with these elders who knew them and understood the context of the composition. The words are transcribed, and the phrases and words are analysed and compared with the idiomatic translation and meanings as explained by the elders. Pitch, rhythm, and duration in spoken words are compared with the realization of the words in performance.¹⁹

Analysis 11 Raven song

Narratives frequently include vocal expressions. Though they are often referred to in English as story songs, this is not really a clearly defined vocal genre in stylistic terms. Story songs occur in narratives where animals are acting characters and also perform vocal expressions. There are a large number of narratives about the Raven, who plays an important role in Athabascan mythology. In many narratives, the Raven is ‘a supreme being […] and endowed with godlike power. (His wishes come true, he creates Alaska, his paddle strokes create islands)[.] He can make people’s wishes come true and is often partly human or [anthropomorphic].’²⁰ People also make offerings to Raven.

Vocal expressions of Raven that belong to narratives are normally referred to as Raven songs. They are usually very short and often end with an onomatopoeic imitation of Raven’s call or of animals referred to in the tale.²¹ Like other story songs, they may also be performed separately from the story for entertainment or – since the raven is also a bird of omen – in relation to certain activities, like for instance hunting.²²

The Raven song performed by Neal Charlie belongs to a narrative, but it was performed separately and there is no summary of the narrative. It lasts less than 6 seconds and consists of two parts: the first part describes a location and the second part consists of two words that are repeated once (Example 69).

The words łi yettha literally translate as ‘dogs are barking’; but since it is Raven who sings, they are performed as an imitation of Raven. As can be seen in Example 70, the first section does not have a distinct pitch or rhythm pattern, but it is close to speech intonation. The pitch range is, however, wider than normal for the speaker, making it more chant-like. The whole expression forms one phrase, and the final repetition gives it a distinct shape.

Example 70 Melodyne graph of Raven song performed by Neal Charlie, Minto. • 14 Raven song

Characteristics of the Raven song

There are no variations to the segmental structure of the words, and the sentence does not include any lexical low or high tones.

Melody

• Speech intonation with increased pitch range.

Rhythm

• Slightly slowed speech rhythm, with strong syllables emphasized.

Form

• Binary: A–B, where B is a Raven imitation.

Phrasing

• Prosodic.

Analysis 12 Caribou song (Tanana)

The Caribou song is one of a set of senh ch’eliga’, medicine songs, recorded by the highly respected elder Peter John with an unidentified interviewer.²³ This recording is not dated, but it is likely to have been made in the 1970s. It is also an animal song, a term used by George Herzog, who counted them as belonging to an old layer of oral traditions.²⁴ Many pedagogical or etiological tales or myths are about animals, who do things related to such activities as successful hunting or healing. In many of these tales, an animal performs a vocal expression that could, in another context, be used for magical purposes.²⁵ The Caribou song belongs to a well-known teaching story about truthfulness and obedience which explains its meaning. It was recorded by the late Neal Charlie in 2008, and other elders, including Neal’s brother Robert, know and repeat the story. Neal Charlie summarized the tale in English as follows:

The song, given to the shaman by the caribou themselves, restores the proper relationship between the people and the caribou, after the boy’s violation of that relationship with his lie.

The free sharing of this material contrasts with a general caution regarding senh ch’elik on the part of the remaining elders of Minto. Since the introduction of Christianity in the Tanana Valley by missionaries in the late nineteenth and early twentieth centuries, expressions of Native Athabascan spirituality have been handled very discreetly in this area. Neal Charlie expressed his feelings about this music in the interview in which the story was recorded: he stated that medicine songs and other spiritual material could be misused if they were shared without full understanding of their proper application. For this reason, he was not willing to discuss many of those recorded and archived by Peter John. This was an exception, owing to its relationship with the teaching story.

The senh ch’elik, or animal songs, are often short. This one resembles Peter John’s other examples in being more speech-like than musical. It consists of two sets of words making up a stanza. The binary poetical form is obvious, since the text is performed twice (Example 71). In the first section, Do sełdini chu is given 8 beats and Bekwlá sełdini chu 10 beats (Example 72). This means that the three extra syllables of Bekwlá are simply added with one beat for each syllable. Apart from this, the principle is isorhythmic and the first motif serves as a building block. The descending melody line is also obvious. There are no real tone repetitions to finish off the lines, but that is explained by the fact that this is a different vocal genre from the following examples. It should be noted that there are no vocables without lexical meaning. Rather, it is one sentence repeated once isorhythmically in a descending fashion.

The melody and rhythm adhere closely to the rhythm and pitch contour of speech. The highest point in the melody falls on the negative stem in the word bekwlá ‘there are none’ (see the arrows in Example 72). In the Minto dialect of Tanana, the negative and a few other specific words are marked with a high-rising, nasalized tone that is very distinctive in the language’s prosody. In this word, the suffix is melded with the stem of the verb ‘to be’ to create a marked negative stem. The melody reflects the tone on this stem, which is the only tonally marked syllable in the text. Rhythmically, the one beat per syllable does not recognize the two light syllables in the text (be- and kw) as meriting different, lesser weight. Nor do the lexical stems in the text get special rhythmic treatment; chu, a conjunction, is elongated, but -ni ‘say’ gets just one beat, as does the high -lá. However, the overall contour, starting higher and finishing lower, with an emphasized negative, closely matches the contour of a Minto intonational unit.

Example 72 Caribou song performed by Neal Charlie, Minto, 10 November 2008. Arrows indicate a higher-pitched final syllable which is also a negative stem in the word bekwlá ‘there are none’. Original pitch: c ≈ 140 Hz. • 15 Caribou song, Neal

Characteristics of the Caribou song

Melody

• Follows the pitch contour of speech closely.
• Descending melody line with a tonal centre.
• Range: 11 semitones.

Rhythm

• Closely follows the rhythm of speech.
• Regular pulse.
• One pulse beat per syllable dominates.
• Isorhythmic organization: the first motif serves as a building block.

Form

• A short binary form made up of one sentence consisting of two sets of words performed twice.

Phrasing

• Verbal metre dominates.

Initial/final formulae

• The final syllable of the phrase (chu, a conjunction) elongated to 3–4 pulse beats.

Word variations

• The two light syllables (be- and kw-) and the lexical stems (-ni, -lá) in the text have a duration of one beat each.

Lexical tones

• The tone of the negative stem in the word bekwlá ‘there are none’, the only marked syllable in the sentence, is reflected by high pitch.

Dratakh ch’elik

The term dratakh ch’elik is a noun compound in which dratakh is a noun derived from a verb meaning ‘to dance a mourning dance, arms moving up and down’. Dratakh ch’elik are made in honour of persons who have passed away and are of particular cultural importance. They would be made by a family member or commissioned from a well-known song-maker. They are performed at ceremonial feasts, particularly the funeral potlatch, and also at memorial potlatches that may occur about a year after the funeral.²⁶ In some cases, a dratakh ch’elik is intended for just one performance, at a memorial potlatch, and is not repeated. These are called khwtitl ch’elik, ‘potlatch songs’. Those that may be remembered and performed at later funerals and potlatches as well are the dratakh ch’elik. Perhaps some khwtitl ch’elik can become re-singable dratakh ch’elik, but we have not observed this process directly. The Minto repertoire includes dratakh ch’elik dating back to the early 1900s, and it is remembered who made them and for whom.

In cases where the dratakh ch’elik is composed by a song-maker, the family member determines the verbal content to be included and the composer assists in organizing the lyrics and then sets them to a melody. The melody is not always new, but it must not too closely resemble important melodies that may be in the repertory of the local community or a neighbouring community. New compositions are evaluated by song-leaders and may be revised if they notice errors. Older compositions, especially really beautiful ones, are often performed at funeral potlatches (held just after a death) when there has not been time to create new ones.

The dratakh ch’elik we are working with have been chosen by elders as important for young people to learn, because of their frequent use for funeral and other potlatch occasions, and because they are considered to contain important advice for life. They are shared by Minto elders when they travel to occasions in other Athabascan communities and are thus familiar around Alaska. It is important for the elders that they are recognized as composed by, and for, Minto-Nenana people; and for this reason, composers are explicitly recognized when the dratakh ch’elik are discussed.

Analysis 13 Dolo k’adi, ‘Missing Dolo’

Dolo k’adi belongs to the type which is called dratakh ch’elik in Tanana and is sometimes referred to in English as sorry songs. Dolo k’adi uses rather complicated ‘high’ language that is not easy for present-day speakers to translate. It is performed widely at women’s funeral potlatches, and is known to members of other language communities in Alaska owing to kinship ties that bring Minto people to potlatches statewide.

Dolo is the name of a woman (Example 73). In most dratakh ch’elik examples, only one kinship term is used in the A section (e.g. en’a ‘mother’). In the case of Dolo k’adi, two terms are used, ‘mother’ and ‘sister’, which emphasize the great importance of the honouree to her family. This memorial song was made by Little Peter in the 1920s for Dolo, who was the eldest daughter of the famous Minto chief Chief Charlie and the mother of Moses Charlie. The performers and translators consulted in this study are her descendants.

The line Oh-oh-oh-oh En’a’a, oh-oh-oh-oh-oh En’a’a, oh-oh-oh-oh-oh En’a’a Soda (A in Example 74) appears three times in the performance, with other stanzas in between as a kind of refrain. The overall poetical and musical form will be A–B–A–B′–A. The A part has a binary form, a building block with the words Oh-oh-oh-oh En’a’a that is repeated twice in a descending motion, marked with double lines at the bottom of the graph in Example 74. The final word Soda is performed as a pulsating tone, that is, with a stress coinciding with each beat, a common trait in Native American music, denoted by a shaded line in the graph. It is phonetically marked by a glottal constriction at the onset of each beat. The descending A section thus ends with a tone repetition on one of the lowest pitches that serves as a tonal centre. This dratakh ch’elik is performed without a drum, and in the sections in which a few syllables are stretched over a number of beats, glottal pulses serve to underline the beat.

Example 74 Stanza A from Dolo k’adi with tone repetitions on vocables at the end. Performed by Neal and Geraldine Charlie, Minto, 5 June 2010. Original pitch: g ≈ 200 Hz. • 16 Dolo k’adi

The stanzas B (Example 75) and B′ (not transcribed) have no vocables but consist of words with lexical meaning praising the departed. Some dratakh ch’elik have four or more stanzas. In this version of Dolo k’adi, two were recorded. Minto elders have suggested that three is a proper number for stanzas in such compositions, and that they may focus on the feelings of the composer, on the important virtues of the departed, and, lastly, on something that the departed person enjoyed doing. Examples of older dratakh ch’elik recorded today may include fewer verses than were originally composed because not all the words were remembered.

Stanza B starts high and descends before it finishes on a pulsating tone repetition. It is longer than A and contains more words. Basically, this section can be seen as an extended variation of A with motifs that are slightly similar to stanza A in the descending section. It will take further research and comparisons between versions and with other examples of dratakh ch’elik in order to establish whether this is a common composition technique.

The B and B′ sections have far more linguistic content than the A sections. In most cases, low lexical tone is not represented in the melody. However, morphologically prominent verb and noun stems have a tendency to be stressed and are more likely to be lengthened or fall on a new pitch (see arrows in Example 75). This reflects the phonetic prominence of stem morphemes in speech, and probably also aids parsing by listeners.

Example 75 Dolo k’adi. Arrows mark stressed morphologically prominent verb and noun stems. Stanzas B. Original pitch: g ≈ 200 Hz.

Characteristics of Dolo k’adi

Melody

• Descending motion with a tonal centre.
• Range: 12 semitones.

Rhythm

• Regular pulse.
• The pattern short–long, where the long is stressed, dominates.
• Isorhythmic organization: the first motif serves as a building block.

Form

• The overall form is A–B–A–B′–A.
• A consists of vocables and one key word with lexical meaning.
• B and B′ consist of words with lexical meaning

Phrasing

• Both A and B end with tone repetitions.
• Prominent verb and noun stems tend to be stressed, lengthened and/or fall on a new pitch, which reflects speech.

Initial/final formulae

• Ends with a tone repetition on the tonal centre; the very last vocable of the performance is short and stressed.

Word variations

• Vocables prominent.

Lexical tones

• •In most cases, low lexical tone is not represented in the melody.

Analysis 14 Segoya (Bettis, Nenana)

Segoya means ‘my baby’, ‘my child’. This dratakh ch’elik was made by Julius Bettis, Nenana, for a son who died as a child. It was performed by Geraldine Charlie, Minto, and the transcription is based on her performance in an interview situation, recorded on 17 May 2013 (Example 76). In the following texts and translations, the parts dominated by vocables are in the left-hand column and parts dominated by words with lexical meaning are in the right-hand column.

The words

The translation in Example 76 cannot easily be derived directly from the words.²⁷ The language used is elliptical, differing from spoken language by the inclusion of formulaic phrases and varying pronunciations of common words. One word has occasioned some discussion with the elders. Do’ił’an’ is a word found in other Minto dratakh ch’elik, and the elders frequently translate it as ‘why?’ In spoken language, it would be interpreted as ‘What are you doing?’ An elder tells us that the addressee of this question is probably not the person who has passed away but, rather, God – hence perhaps the more general translation ‘why?’ In performance, the word is variously pronounced as do’ił’ani, do’ił’ana, or do’ił’ini.

The vocables, words without lexical meaning, enclose the meaningful words – they precede them and follow them. A sequence containing vocables, perhaps including a kin term as at the beginning, is called bent’aya’, which the elders translate as ‘chorus’. Vocables in Minto vocal expressions can contain vowels that are not found in spoken language, such as the diphthong ei/ey.²⁸ In this case, the vowels in the vocables are low- or mid-level in quality. While our sample is still too small to make a strong claim, we believe that the dratakh ch’elik uses fewer high- and mid-level vowels than ch’edzes ch’elik.

Both vocable syllables and meaningful syllables are elongated on moving sequences of notes. In the case of the meaningful syllables, the moving sequence most frequently corresponds to a morphological stem, but it does not always: udesni, t’anh, do’il’an’, and of course sedena all have a moving sequence on a verb or noun stem. However, the movement on seghw occurs on both the prefix and the postpositional stem.

Another type of elongation occurs on the last repetition of segoya. This type of augmentation consists of the prolongation of a vowel over many beats with the same pitch, divided by glottal pulses. This type of lengthening of a syllable or syllables of lexical words or vocables is used only in the performance of vocal expressions. It may occur both at the beginning and at the end, in this case only at the end.

The music

The performer is sitting down and uses a fairly soft voice relating to the rather intimate situation of the recording session. She accentuates the even beat strongly and rocks slightly forward and back with the rhythm, sometimes in combination with hand movements. Segoya is repeated once later on in the interview session and this repetition is identical with the first one, apart from minor variations in pitches that do not change the main outline of the melody.

Segoya starts with vocables on a rising minor third and with short–short–long tones (A in Example 77). This short motif is immediately repeated with the first word, segoya. The rhythm short–short–long (with minor variations) recurs many times during the performance, as indicated by brackets in the transcription. It functions as an isorhythmic motif and can be regarded as a building block.²⁹

When the highest pitch is reached, it continues with vocables while descending stepwise in a triadic fashion (B, B′ and C, indicated by a dashed arrow). It ends on a tone repetition at the lowest pitch (C, indicated by a dashed line). Here the vocables, while poetically amplifying the vowel sounds, musically serve to bring the melody to an end. The tone repetitions consist of vowels prolonged by a series of glottal attacks. This marks the ends of the stanza: A–B–C.

The second stanza is based on a repetition of this melody, but now with words that carry lexical meaning, and it is prolonged for 8 beats at the highest pitch (A′, marked with an arrow) before it continues with a minor variation on the melody (B′), and finally ends with the same phrase as the first stanza (C). The order of the phrases is A–B–C–A′–B′–C.

Example 77 Transcription of Segoya (Bettis, Nenana). Performed by Geraldine Charlie, Minto, 17 May 2013. Original pitch: final tone f ≈ 170 Hz. • 17 Segoya, Bettis

Characteristics of Segoya (Bettis)

Melody

• Starts low and rises. After the highest pitch it descends stepwise in a triadic fashion with a tonal centre.
• Range: 12 semitones.

Rhythm

• Regular pulse.
• One pulse beat per syllable dominates.
• The rhythm short–short–long (1–1–2 beats) functions as an isorhythmic motif and as a building block.

Form

• Overall form: A–B–C–A′–B′–C.
• A, B, and C consist of vocables and one lexical key word.
• A′ and B′ have words with lexical meaning at a high pitch level.

Phrasing

• Phrases end on a tone repetition.
• In the case of the meaningful syllables, the isorhythmic motif often corresponds to a morphological stem.

Initial/final formulae

• Ends with a tone repetition on the tonal centre. The very last vocable of the performance is short and stressed.

Word variations

• Vocables prominent.
• Both vocable syllables and meaningful syllables are elongated at the end of the isorhythmic motif.
• The last syllable of a tone repetition is elongated over several beats.

Analysis 15 Segoya (Titus John)

Another dratakh ch’elik for a child was made by Titus John, probably in the 1920s. It was performed by Neal and Geraldine Charlie, Minto, and the transcription is based on their performance in the interview situation on 17 May 2013.

Poetics

The words display a structure which is found in a number of dratakh ch’elik and which the elders specifically point to when teaching them. Three specific words ch’eghwtsen ‘love’, chononi ‘hunting luck’, and gholiyo ‘wealth, good fortune’, (gholiyi in singing), are inserted into a line of poetry that is otherwise repeated in three separate verses performed with the same music. The verses occur in this order, and we have no examples of deviations from that order. Elders translate the repeated line as something like ‘[I thought] there would be back and forth downriver’ (Example 78).

These three important words are often found in dratakh ch’elik composed in honour of adult men. In this case, part of the poignancy comes from the presence of the adult words about someone who will not become an adult, and therefore will not have the opportunity to enjoy the love, luck, and wealth mentioned in the verses.

The verb k’ets’eniłno is not exemplified in prose examples in our corpus, but because verbs have so many possible forms in Athabascan languages, this does not define it as a purely poetic word. Do’ił’ini is a different form of do’ił’an’, found in the Bettis’ Segoya (Analysis 14), and is translated as ‘why?’ though the relationship of the meaning to the morphological composition is not clear.

Vocables and words

The vocables are very restricted in vowel quality, comprising only [ei] and [o]. Segoya, ‘my baby’, is lengthened using both changed pitches, in the second measure, and glottal pulses, in the extensions in verses 2 and 6 (see Example 79). Otherwise, 4-syllable sequences are being created, with vocable syllables added where a word does not end in a vowel. So, we have ch’eghwtsen’ ei, with an added syllable, treated in much the same way as k’ets’eniłno, which has the same number of syllables and ends with a vowel. The four 4-syllable sequences ch’eghwtsen’ ei, k’ets’eniłno, ghets’e yodo and do’ił’ini carry the same metrical weight and have a similar rhythmic pattern. Chononi and gholiyi do not have 4 syllables, however, and are not augmented by a vocable – they are just spread over the same number of beats as the other sequences. Overall, however, there is an impression that the meaningful words are delivered one syllable to a beat, an emphatic and highly recognizable pattern found in many performances of this type.

Some vowels are specific to vocables and do not occur in spoken language in the Minto dialect. In order to transcribe the vocables, it is sometimes necessary to employ spectrograms. For this reason, a solo performance is needed, since sounds are often blurred in communal situations. Many performances in our material were recorded in communal situations. Therefore, two types of transcription are used in this text: one after detailed analysis, the other (more approximate) by ear using the letters u [u], i [i], o [ɔ], a [ɑ], ey [ɛj], dz [ts], b [p], h [h].

In do’ił’ini, it is not obvious whether the final open syllable is an added vocable or a meaningful piece of morphology, since the question marker in this language contains the same vowel. This morpheme is glottal final [-iʔ-], but in Minto, the vocal expression’s final glottal stop is often not realized. Given the narrow range of vowel qualities in the other vocables, we have chosen to translate it with ‘why’.

Example 79 Segoya (Titus John), performed by Neal and Geraldine Charlie, Minto. Original pitch: final tone c ≈ 115 Hz for the male voice.

Music

This Segoya also starts with a rhythmic combination of vocables and a key word that serves as a building block (A, marked with a horizontal line). The key word segoya that follows immediately is performed in a descending motion, so that the final syllable is the lowest. The second time the building block occurs (B), the descending movement, in triads, is prolonged; and in the third repetition, it is further prolonged before ending with a tone repetition at the lowest pitch, followed by a short section at a high pitch which is based on vocables (marked with an X). The B phrase is repeated in a varied form (B′) and prolonged so as to fit the words. Each time it ends with a short high-pitched and descending figure (Y), which is a variation on the ending of section B (X), after which A is repeated. This pattern (B′–A) is repeated three times, one for each line of words. The total poetical and musical form is A–B–A–B′–A–B′–A–B′–A. As is often the case, the very last word of the performance is cut short: sego instead of segoya, with a heavy stress on the syllable go.

Characteristics of Segoya (Titus John)

Melody

• Descending melody with a tonal centre.
• Range: 14 semitones.

Rhythm

• Regular pulse.
• Isorhythmic organization: the first motif serves as a building block.
• Words with lexical meaning are often grouped in 4-syllable units corresponding to 4 pulse beats.

Form

• Starts with a motif with vocables and a lexical key word that serves as a building block.
• The last phrase of the first part is repeated in a varied form and prolonged so as to fit the lexical words.

Phrasing

• The last syllable of a phrase is lengthened.

Initial/final formulae

• Ends with a tone repetition on the tonal centre.

Word variations

• The vocables are restricted in vowel quality, comprising only [ei] and [o].
• When a word is not vowel-final, vowel syllables are added: ch’eghwtsen’ becomes ch’eghwtsen’ ei and is treated in a comparable manner to k’ets’eniłno. Each 4-syllable unit carries the same metrical weight.
• In other cases, words are elongated to fit the 4-beat pattern: chononi and gholiyi are not augmented by a vocable but are simply spread over the same number of beats.

Structural framework of dratakh ch’elik

The different aspects of the two dratakh ch’elik that have been treated separately here are, of course, fully integrated in performance. It is quite obvious that both are built on a framework that is presented in the first stanza (‘A’ in the examples) and then repeated in the following stanza(s) with variation (‘B’).

This framework starts with a key word section consisting of the musical setting of the main word (segoya) that is echoed musically and/or rhythmically with vocables which, in these cases, precede the main word. This is a very poignant motif which serves as a building block for the rest of the framework. Obviously, the originality of the musical setting of the main word and its ‘sonoric double’ in vocables is of great importance for the originality of the composition. The vocables, in combination with pitch and tone-lengths, also appear to have a poetic function in that they help to accentuate the main word. They also introduce syllables with consonants, and particularly vowels that will be used throughout the melody and thus colour it with a unified sonoric form.

Once the key word section is over, the framework continues with a descending unit with vocables, which descends in rather large, mainly triadic steps. This brings the melody to its tonal centre with a final unit consisting of a tone repetition on the last vowel of the main word that ends the stanza (Table 6).

The repetition of the stanza contains words with meaning and could be called a lexical unit. It consists of one sentence that is constructed so that it can be repeated with only one word changed. The principle thus involves parallelism (see Example 78). The words are performed to the melody of the framework, which is flexible enough to be prolonged so as to comprise a full sentence.

The result is that of a very tightly knit vocal expression in which the structural and aesthetic functions of the music and the language are present simultaneously, and all these elements are integrated and interdependent in the organic whole (Table 6). A comparison with some other Athabascan dratakh ch’elik shows that two examples from Minto share similar characteristics.³⁰ Examples from other places have similar structures with regard to the parallelism of the texts and its content, but differ as regards the musical framework.³¹

Vocables have been widely discussed in the literature on Native American vocal practices, and while their origin is still disputed, there is evidence that vocables should be understood as having many different functions, some of which have been touched upon here. The fact that they are so intimately tied to a musical form and to aesthetic content is probably one reason why vocables in vocal expressions tend to be repeated without change in each new performance, and also why the vocables can be remembered. Concerning Navajo ceremonial vocal expressions, Frisbie says:

This may indeed be so in certain vocal genres. In the particular case of dratakh ch’elik, the poetical form is more fixed than a myth would generally be. Frisbie’s reflection is still relevant, though, when it comes to memorability. This study of two dratakh ch’elik has shown that when the vocables are seen as an integrated part of the total linguistic–musical–poetic framework, there are many cues that serve to increase memorability. This may also explain why the performer first searches for the initial building block when she is trying to remember a specific dratakh ch’elik. When she has found it, the dratakh ch’elik unfolds.

The structure of the dratakh ch’elik thus serves as a performance template that facilitates the creation and re-creation of specific vocal expressions. In this performance template matters like parallelism and vocables – in which musical and linguistic factors are integrated – play a significant role.

The dratakh ch’elik performance template

Melody

• Often, but not always, stepwise descending melody with a tonal centre (see Example 74).
• Range: 8–14 semitones.

Rhythm

• Regular pulse.
• One pulse beat per syllable dominates.
• Isorhythmic organization common: the first motif serves as a building block.

Form

• Strophic.
• Section A consists of vocables and one key word with lexical meaning denoting kinship.
• Other parts consist of words with lexical meaning praising the deceased: one line that is repeated with each repetition constructed so that one important word can be exchanged for each repetition (see Example 73).

Phrasing

• Phrases end on a tone repetition.
• A–B with variations.
• Both A and B end with tone repetitions.
• A contains vocables and a lexical key word, B (and repetitions) contains lexical words.
• The words with a lexical meaning are delivered over 4 beats.
• In the case of the meaningful syllables, the isorhythmic motif often corresponds to a morphological stem.
• Prominent verb and noun stems tend to be stressed, lengthened and/or fall on a new pitch, which reflects speech.

Initial/final formulae

• Ends with a tone repetition on the tonal centre; the very last vocable of the performance is often short and stressed.

Word variations

• Vocables prominent and restricted in vowel quality, comprising only [ei] and [o].
• Both vocable syllables and meaningful syllables are elongated at the end of the isorhythmic motif.
• When a word is not vowel-final, vowel syllables may be added: ch’eghwtsen’ becomes ch’eghwtsen’ ei and is treated in the same way as k’ets’eniłno. Each 4-syllable units carries the same metrical weight.

Ch’edzes ch’elik

Ch’edzes ch’elik are performed with dancing at gatherings of all types, including formal feasts – potlatches – in honour of deceased or living persons. They may also be performed for audiences. The Tanana expression literally means ‘dance song’, which is a compound of two nouns. It is the term used by elders when designating this type of music. The dance is led by performers and drummers, and many people can join in. Some dances are choreographed – like Minto crow dance or Back and forth – while others may be free, dancers moving individually or in a row. The ch’edzes ch’elik themselves are rather short and usually strophic, consisting of one stanza that is repeated several times. They make abundant use of vocables; some have no words with lexical meaning at all, but consist entirely of vocables.

The Minto Dance and Song Group was formed in the early 1960s. It was one of the first groups in a revitalization movement of Athabascan music through the organization of village dance teams. Sometimes such an initiative was taken by a single dedicated individual.³³ Among prominent early members in the Minto Dancers were the late Peter Jimmie – who was a dance leader and also composed many ch’edzes ch’elik and dances that are still performed – and, over the years, song-leaders such as the late Evelyn Alexander, the late Dorothy Titus who was also a famous song-maker, and Sarah Silas. The Minto Dancers still exist and have performed at pow-wows and on stage at various celebrations. The dance group also serves as a learning opportunity outside the school.

There are some early recordings of the Minto Dancers. One contains 11 ch’edzes ch’elik originally recorded – probably in 1963 – by Gordon Olson, who was then a missionary in Old Minto.³⁴ They were copied by the Swedish folklorist Anna Birgitta Rooth, who recorded a great number of myths and tales, together with some vocal performances, during fieldwork in Alaska in 1966. The Rooth collection also includes some 30 ch’edzes ch’elik recorded at a potlatch in Tanacross.³⁵

In 2009, Tuttle interviewed some of the elders in Minto about the ch’edzes ch’elik in this particular sample, including Neal and Geraldine Charlie and Bergman and Sarah Silas. This provided some new information and, in several cases, also new recordings made by those elders who were present, with better sound quality than the original. The enhanced quality made it possible to transcribe vocables with greater precision. There are also recordings of the Minto Dancers from 1967 in the Alaska Native Language Archive, Fairbanks. Another collection of ch’edzes ch’elik was used by Tony Scott Pearce for his master’s thesis.³⁶ These recordings include about 100 examples recorded in Minto (1972–73), Nenana (1974), and Fairbanks (1983).

The ch’edzes ch’elik are normally accompanied by drumming, using a frame drum. The drumming coincides with the pulse of the vocal performance, but there are small variations in the form of pauses for one beat, double speed, or more complex rhythms. There are no cases of final formulae, but ch’edzes ch’elik normally end with one short and stressed tone that coincides with one strong drumbeat. There are three cases of tertiary division (notated 3/8) with drumbeats on 1 and 3.

The most common overall musical forms of the Minto ch’edzes ch’elik can be described as A–A–A–A, A–B, A–B–B′ or A–B–C, but paired phrasing (A–A–B–B) is rare. Usually the B-part of an A–B ch’edzes ch’elik will be dominated by vocables. The A-part may consist of vocables + lexical words or of vocables only. Phrase endings have tone repetitions, mainly using vocables.

The melodic contour is dominated by downward motion: starting high and ending low. An undulating motion around a central tone is also common. Downward passages make use of sequences, i.e. a series of repetitions of a short motif in a stepwise downward motion so that each repetition starts on a lower pitch than the preceding one. Occasionally, this also occurs in an upward motion. Sequences are especially common in the B section. It is common for sequences to be performed in half as long tones, notated as quavers, while the drum keeps on with the quarter notes. This has the effect of speed doubling and makes the music swing. The variants shown in Example 80 (starting at the arrows) could be characterized as a) consecutive downward, b) pendulum upwards and c) interlocking. This is especially common close to the final tone repetition.

Example 80 Rhythmic sequences in the endings of ch’edzes ch’elik, particularly in Minto (from Lundström 1980: a = E2, b =D5, c = E10).

The Rooth and Pearce samples mentioned above contain ch’edzes ch’elik from Minto, Nenana, Tanacross, and Fairbanks. With many traits in common, they represent a North Athabascan style. It is also possible to distinguish some local traits, for instance a short final tone common in Minto, or paired phrasing which is much more common in the Tanacross material. Comparison between different local traditions is difficult, however. Until the mid-1950s people were seasonal nomads. Even after people began to settle in more permanent villages, there were still many contacts between villages, not least in connection with potlatch feasts when people from several villages gather and take part in the celebrations. Another factor is that, as in most modern societies, young people move away for education or for work. So local practices have spread and mixed in many ways. On the other hand, since people often remember who produced a certain composition, why it was made, and when it was introduced at a potlatch, they also keep track of which local style it represents by comments such as ‘that is a Nenana song’, ‘that song was first introduced in Tanacross’, etc. Sometimes an attribution of origin is made on the basis of stylistic features.

Analysis 16 Joni ło’o, ‘Here it is!’ (Tanana)

In this ch’edzes ch’elik the words are limited to one sentence (Example 81; see also Example 82 for the melodic outline). The ganhok is a tall dance staff. Joni ło’o is said to have been composed on the occasion of the introduction of the ganhok to the Tanana people – the word is of uncertain origin, though there is speculation that the staff was introduced through Tlingit contact.³⁷ While the recording here was made by a small group of elders, normally it would be performed by a hall full of people all shouting together.

In this case, the melody and rhythm overwhelm the words, and all the meaningful syllables are given the same weight: one strong beat. Phrase-final vocable syllables are lengthened over several beats. There are no tonal syllables. There are 8 meaningful syllables, arranged in two groups of 4 (Examples 81 and 82). The following vocable sequences also form patterns in groups of 4. All the syllables are heavy, in accordance with the Tanana stress system, in which full vowels and closed syllables count as heavy. A sequence of 4 heavy syllables would not be impossible in speech, but such a sentence is usually broken up by light syllables (CV, where V is a schwa) in the normal composition of sentences.

The first half of the text (A, Joni ło’o) is performed as three short tones and one long, and the same rhythm is repeated for the second half (Ganhok tolał, Example 82). The final tone of these words finalizes the first section and functions as its tonal centre. The second section (B) is built on the same rhythm, which is repeated twice, but with vocables with no lexical meaning (Iya ho’ey). For the third repetition, these syllables are performed at the same pitch, which stresses the tonal centre of the section.

Example 82 Graphic representation of the melodic movement in the dance song Joni ło’o. The vertical axis shows the pitches. For the most part, there is one syllable per drumbeat; but at the end of each phrase, each syllable is held for two beat-pulses with a glottal stop or a dip for each new beat. Performed by Neal and Geraldine Charlie, Minto, 14 May 2009. Original pitch: g ≈ 200 Hz.

Several characteristics are common in Native American music in general: it is strophic (the diagram in Example 82 shows one stanza that is then repeated with minor changes); it is isorhythmic (that is, built on one rhythmic idea throughout); it has a stepwise downwards or falling motion (the second section of the stanza); and it ends on a tone repetition at one of the lowest pitches that serves as a tonal centre (denoted with a wave-line in the graph).

In his study, Pearce concludes that the overall musical form in this genre is usually binary, i.e. there are two distinct sections within each stanza. He also finds that they are often built around one melodic/rhythmic motif that is repeated at consecutively lower pitches, a pattern which, in a musicological term, is called sequences. Pearce sees this motif as a building block around which the melody is built (denoted by a double bottom line in the graph).³⁸

Joni ło’o may represent a category that starts out with a poignant melodic/rhythmic setting of a verbal phrase. That phrase then becomes the nucleus, whereupon it comes to serve as a building block. It is combined with vocables, sequences in a downward motion, and ends with a tone repetition.

Characteristics of Joni ło’o

Melody

• Stepwise descending melody, with a tonal centre.
• Range: 8 semitones.

Rhythm

• Regular pulse.
• One pulse beat per syllable dominates.
• Isorhythmic: the first motif serves as a building block.
• All meaningful 8 syllables arranged in two groups of 4.

Form

• Strophic.
• Binary form A–B: A has words with lexical meaning, B consists of vocables.

Phrasing

• Vocable sequence patterns in groups of 4.
• Phrases end with tone repetitions.

Initial/final formulae

• Ends with tone repetitions: phrase-final vocable syllables are lengthened over several beats.

Word variations

• Vocables.
• All meaningful syllables are heavy and given the same weight: one strong beat. In speech they would normally be followed by light schwa vowels.

Analysis 17 Christmas tree

Compositions in traditional style with words in both Athabascan languages and English became common in the twentieth century, and are some of the most popular ones; their composers are often remembered. The Minto Airplane song is a perennial favourite and has been popularized in a video – in this case the airplane is a positive force, a ride to the potlatch.³⁹ Another Minto dance song inserts English words such as whiskey in the framework enu’ey! ‘Away with it!’ In this case, the influence of white outsiders is disapproved of.

Christmas tree was composed by the late Dorothy Titus, who was a widely respected song-maker. She was known for her ability to compose on the spot. This is the recollection of Geraldine Charlie, who performed the version recorded here:

There are only two words with lexical meaning: Christmas and tree (Example 83). All the other words are non-lexical syllables or vocables:

As is common in the ch’edzes ch’elik, the lexical contents come in the first phrase as words, not a full sentence. In this case, the first phrase is repeated (Example 84). The vocable part of the first phrase mirrors the lexical part, which makes this quite an original start. The general contour is a stepwise falling: phrase 1: c'–g, phrase 2: g–c and phrase 3: e–c. The two final phrases end with tone repetitions, and the vocable ending is of the hiya-type. It marks the end of the stanza, and before the tone repetitions there is a typical rhythmic sequence. It is a very tight composition, and this is probably a quality that helped to make it remembered. While performing it, Geraldine Charlie made hand movements depicting a Christmas tree.

Example 84 Christmas tree. Performed by Geraldine Charlie, Minto, 17 May 2013. Brackets show tone repetitions at the end of phrases. Original pitch: final tone c ≈ 240 Hz.

Characteristics of Christmas tree

Melody

• Falling melody contour in a series of sequences with a tonal centre.
• Range: 12 semitones.

Rhythm

• Regular pulse.
• Many beats subdivided into two (8th notes); dotted rhythms occur.
• Isorhythmic organization prominent.

Form

• Form: A–A–B–B′.
• A consists of vocables and one lexical key word in a melodic motif.
• B and B′ consist of vocables.

Initial/final formulae

• Ends with a tone repetition on the tonal centre; the very last vocable of the performance is short and stressed.

Word variations

• Vocables dominate.

Athabascan leadership and composing

The maintenance and revitalization of highly endangered languages require a supportive context. While language documentation can create reference materials in the form of resources, and classroom teaching can begin the process of initiating new speakers, these activities do not provide the most important component: a reason to speak in the target language. And without this component, neither maintenance nor revitalization can be achieved;⁴⁰ there will be no viable place for the language outside the classroom.

Reasons to speak come in the form of situations in which the language must be spoken. Political leaders and public cultural representatives for indigenous groups in the United States must often demonstrate fluency in a particular language. Ceremonial contexts may also require proficiency in languages that are otherwise being replaced by more politically powerful codes. A third context is public performance, where the performers serve as cultural representatives of their own communities or of their language community as a whole.

For some members of minority language communities, these three situations become conflated: political leadership may require demonstrations of language proficiency in ceremonial and performance situations as well as in communications with fluent peers or elders. In the interior of Alaska, young people who are moving into leadership positions now face these challenges. This section examines some of the strategies they are using to develop their linguistic proficiency.

Conversation with several young leaders from the lower Tanana River area (Minto and Tanana) suggests that learners develop innovative strategies to carry out their goals, including methods that draw on academic documentation, indigenous learning and teaching, and also digital media sharing. Art and leadership develop together, providing an instructive example for supporters of language revitalization.

The dratakh ch’elik and the ch’edzes ch’elik are well suited to communal performance. The prominent setting of the key words stands out and is ‘catchy’, while the vocable parts contain comparatively common musical material that is fairly easy to learn at the first hearing, particularly if the vocal expression is performed several times in a row. Children can learn them at potlatches by listening, and gradually take a more active part in the performance and dancing.

Communal performance may, at best, keep the vocal expressions in mind; but, if a musical culture is to be alive, there must also be individuals who relate to the music by learning, adapting, and transmitting it. In Northern Athabascan culture, this includes composition, which requires the composition of lyrics in combination with vocables and the handling of the interaction between the lyrics and the music.

The tradition seems to be that composing is something you do alone; but even experienced elders will seek the judgement or advice of other experienced persons, particularly in the case of dratakh ch’elik that have a formal character. There is no evidence of a tradition of more formalized teaching situations.

Learners become involved with the older repertoire and try to create their own vocal expressions. They may seek out elders for counselling concerning the proper use of language, and for advice concerning the quality of the vocal expression and its originality. They will also consult other friends with a good knowledge of the field. In addition, they contact linguists to help them with transcription, and to learn interview techniques to help them access the history of specific vocal expressions. Counsellors, both indigenous and non-indigenous, are seen as resources and colleagues, and learner confidence grows with each achievement.

Community leadership and song-leading go together in Interior Athabascan communities. In Minto, for instance, the late Neal and Geraldine Charlie were both community leaders and song-leaders, and Neal was also a song-maker. The path to leadership in tribal politics or ceremony is not provided for at school. Instead, leadership develops along with verbal and artistic skill when students themselves take on the goals of mastering language or verbal art. These younger leaders are independently carrying out a programme that elders have frequently advocated: using vocal expressions to learn language.

While the idea of learning language through vocal expressions is appealing, it has not been easy to see how to apply it in classroom teaching. Ch’edzes ch’elik contain very few words, though these words are of great cultural importance (e.g. Tanana ch’eghwtsen’ ‘love’, gholiyo ‘prosperity, right living’). In addition, the pronunciation of these words may differ from spoken pronunciation, making them hard to understand in the strong shouting register of group performance. The child-directed classroom songs, with Athabascan words used in language class, also contain little language, and have the additional disadvantage of further adapting the rhythm of Athabascan words to fit into English metrics. Other types of vocal expressions, such as the dratakh ch’elik used at potlatches, contain a considerable amount of text, but the text is often complex and poetic, much too difficult for beginners or intermediate learners to parse. These vocal expressions are often also off-limits outside their ceremonial context.

What we observe is that language learning through vocal expressions involves taking on a role other than that of classroom student. That is, the learners who make progress are those who decide that they want to create vocal expressions, or be song-leaders. These choices require understanding and use of a language that is culturally appropriate and functional, though it may be minimal in lexical coverage.

To understand what it means to dedicate oneself to community learning, it is helpful to consider examples. Norman Carlo, a young man whose family roots are in the village of Tanana, is one such example. In the following section, we document some of the choices he has made in order to gain sufficient language proficiency to compose songs that are acceptable in the linked communities of Tanana, Minto, and Tanacross.

Norman Carlo’s learning project

Norman Carlo holds an associate science degree in process technology from the University of Alaska Fairbanks. Born in the village of Tanana in 1988, he has now joined the International Brotherhood of Electrical Workers and works on the Alaska Pipeline. He also has a major project: to make a comprehensive set of compositions for his native village, where traditions have lapsed since the older generation passed away. He wants to teach them to the people of Tanana and help them to develop the repertory so that they can present them as their own when visiting potlatches.

Geographical and language names can be confusing here. Tanana, originally a Native village situated at the confluence of the Tanana and Yukon rivers, has been heavily impacted by contact with non-natives due to its location on the rivers and the fact that rail lines run through it. The language native to the area is Upper Koyukon, a dialect of Koyukon that shares many phonological and lexical features with the Lower Tanana language spoken in Minto. Language shift has moved more quickly in Tanana than in Minto. Minto is now on the road system, but the village was moved there from a place on the Tanana River in 1969, owing to repeated flooding.

Partly because of this historical difference, more Athabascan elders who have stayed active in song-making and song-leading remain in Minto. Until very recently, elders have been able to lead performances and to assist with the composition of dratakh ch’elik. This has not been the case in Tanana. The result is that far fewer vocal expressions from the area have been represented in potlatches in recent years, and fewer older ones are remembered and performed. Norman says:

Norman’s process in developing these compositions has drawn on the talents and mentorship of both Athabascan elders and University contacts. He attends all possible occasions for dancing and vocalizing in the presence of knowledgeable elders; and he has studied old videos to observe conservative styles of performance.⁴¹ These videos feature performances and other activities carried out by Lower Tanana elders from Minto. In addition, he has worked on Koyukon literacy with Tuttle in an independent study where his goal was to create lyrics in this language. As he began to build his collection, Norman recorded each composition in a laboratory environment with Tuttle at the Alaska Native Language Center, providing copies of the material and allowing us to keep track of his progress and talk about it.

The potlatch functions as a context for language and music. Before starting to compose, Norman had learnt vocal performance, drumming, and dancing in situations like potlatches and other celebrations, both as an observer and as a participant. While he gradually acquired these performance skills, he noticed that people watched him and listened to him. This gave him confidence to continue. He learnt about structures from listening to older recordings and present performances, but he had not acquired the local language as he was growing up; so, in order to compose he needed to develop his language proficiency. In this case, after acquiring performance skill at potlatches, Norman found a reason to learn the language of his home village – and what is more, something of neighbouring languages as well. Hence, musical activity and the potlatch as a functioning context led to language learning. An important element in Norman’s story is the continued vitality of the Minto Dance and Song Group. Without this continuing strand of language-infused culture, Norman’s quest to become a song-maker and leader might not have been realized. (Figure 4)

The crucial context for language use, as demonstrated by Norman’s journey, is the potlatch, a domain where outsiders have no significant role. As Hinton demonstrates, no strategy for language revitalization that requires motivation or organization from outside the speech community can persuade people to learn and use a language.⁴² Those who take on the challenge of learning and using a fragile minority language need internal motivation. However, the roles that Hinton identifies for external helpers (help with literacy and documentation) are exactly those that Norman chooses for his academic supporters.

Which language is he learning? Norman’s lyrics often use forms that could belong to several different languages of the Athabascan family, since the languages are distinguished by relatively few consonant contrasts and the vowel systems are very similar. However, his lexical choices reflect the content of the lyrics, extending across languages. For example, in a composition he prepared for a Tanacross elder, he included both Koyukon and Tanacross language in the lyrics in order to honour this mentor from upriver (Tanacross is the Athabascan language spoken east of Fairbanks; the name refers to ‘Tanana crossing’). This Athabascan language is phonologically and lexically more different from Upper Koyukon than Tanana; so, to translate to it, Norman consulted speakers from Tanacross. Morphology, syntax, and spoken phonology are hard for him in all the languages. Even so, distinguishing between the three, and working with important concepts and words from each of them, is something he expects and indeed needs to do.

It may seem strange to approach three languages as opposed to just one when all three are highly endangered. However, the people of the Tanana River have always lived in a multilingual environment. Being comfortable with language differences (which Athabascans often refer to as dialect differences) is part of their cultural heritage. Norman’s language learning is still developing, but he does not seem to be at all confused by learning from people who speak different languages; instead, he seems free from a common problem among language learners: the need to learn ‘my grandmother’s language’ and no other. This hang-up about locality may cause severe problems for language learners when mentors are scarce. Therefore, Norman’s approach may have a great deal to recommend it.

Figure 4 Norman Carlo dancing at the Alaska Federation of Natives Convention in Anchorage, in the autumn of 2015.

Norman, along with other Alaskan Athabascans, is developing the skills he needs to move into a changed world: one where the village is no longer limited to physical locality, but is defined by blood and marriage relationships and shared history. In order to get his lyrics and melodies out to the stakeholders he wishes to contact, Norman created a closed Facebook group into which the lab videos of his performances are linked. This strategy allows him to obtain feedback from a fairly widespread, discontinuous village. His revised Caribou song, a composition for his mother – one part related to sobriety, and one inviting his friends to dance – is linked to this page. This web development is relatively new, and it is not clear exactly how the feedback will develop. Participation on Facebook is high among Alaska natives in all except the oldest generations, and its interactive format may provide a way for other learners to comment more freely than they might do in a formal, local context.

Analysis 18 Caribou people version 2 by Norman Carlo

Norman Carlo has created a set of compositions that he continues to revise and refine as he receives feedback from people he considers stakeholders. He recently changed the melody of one of them (originally Caribou song in honour of his clan, not to be confused with the Caribou song (Analysis 12)), to make it easier to distinguish it from the work of another, older song-maker, Robert Charlie, whose musical experience includes the heritage of his father, Moses Charlie, and his grandfather Chief Charlie. The revised version is transcribed in Examples 85–86. This version was questioned again for lyrical content when an elder remarked that it is boastful to speak about one’s clan. Norman revised the lyrics accordingly, the result being The people’s song, which is both more inclusive and more culturally appropriate.

The genre is that of ch’edzes ch’elik. There are three words with lexical meaning. They are:

Example 86 Musical transcription of Caribou people version 2. Performed by Norman Carlo, 17 February 2015.

Most of the discussion and revision have centred on the introductory words and their musical realization. Example 87 shows how the original phrase a) was quite similar to that of an existing dance song b), while that of the revised version is quite different c). The second part, consisting mainly of vocables, was not altered at all; see Example 88 (one exception is the first tones of c) which are higher in order to link the first high-pitched phrase with the second phrase). Apparently, considerations of meaning, originality, and aesthetics focus on the initial part of the vocal expression, whereas the longer falling phrases tend to be more common or formulaic in nature. This pattern explains how ch’edzes ch’elik have the capacity to combine the unique with the everyday, a capacity which enables the composition of new songs as well as communal performance and learning. The same pattern accounts for the memorability of these songs and their inclusion in the general repertory.

Example 87 The initial phrase of a) Caribou people version 1, and b) a dance song by Robert Charlie performed by Norman Carlo • 18 Caribou people, v1 1st phrase a and b and c) the final phrase of the first stanza of Caribou people version 2. Performed by Norman Carlo. Original pitch: final tone f ≈ 188 Hz. • 19 Caribou people, v2 1st phrase

Example 88 The second (and final) phrase of a) Caribou people version 1, b) a dance song by Robert Charlie performed by Norman Carlo, and c) Caribou people version 2 performed by Norman Carlo. • 20 Caribou people, v2 2nd phrase

Characteristics of Caribou people 2

Melody

• The overall contour is falling with a tonal centre.
• Range: 6 semitones.

Rhythm

• Regular pulse.
• Syllabic except in tone repetitions.
• One pulse beat per syllable dominates.
• The first part (A) is isorhythmic.

Form

• Form: A–A–B–B′.
• The initial part, A, contains vocables and key words with lexical meaning.
• The final part, B and B′, consists of vocables and a song-word.

Phrasing

• The final syllable of a phrase is lengthened.

Initial/final formulae

• Ends with a tone repetition on the tonal centre; the very last vocable of the performance is short and stressed.

Word variations

• Vocables dominate.
• Lexical words are pronounced differently from ordinary speech.
• Final tone repetitions consist of vocables and one lexical word, haaliyaa, that functions as a song-word in this context.

Ch’edzes ch’elik endings or ‘refrains’

Norman Carlo presented some twenty ch’edzes ch’elik in two recording sessions, most of which he had not presented before (30 August 2016 and 8 September 2016). He used recordings he had made on his cell-phone in order to recall them. Norman does not verbalize much on the process of composition, except in connection with revision. Usually he says that it ‘just came to me’; but he has said that ‘usually when I start off, I start with the tune and then I think of words to put in there’. Most of the ch’edzes ch’elik in these recording sessions had not yet been tried out in actual dancing situations. Some of them may never be used; but it is just as likely that some will be used and become part of Norman’s regular repertory, and also be learnt by others.

The musical form is similar to ch’edzes ch’elik in general, and A–B, A–A–B, A–B–B or A–B–C dominate. Some of the newer ones were longer and had more sections repeated: A–A–B–B′–C–C′ or A–A′–A″–B–A‴–B′. Perhaps this shows a progression in his composing. Some have key words with lexical meaning, while some consist of vocables only. The majority have the same haaliyaa-type ending as Caribou people (Example 89). Haaliyaa (gholiyo in speech) means prosperity and right living. Usually further vocables are added, for instance e-e-ey haaliya h-o-o-hey ho or e-e-ey haaliya hi-hey-ho-o-hey-ho (Example 90). Sometimes the last vocables in the chain are performed one octave higher. Such octave doubling is common in the very last refrain of Athabascan ch’edzes ch’elik, and Evelyn Alexander of the Minto Dancers was particularly well known for octave doubling.

Example 89 Final vocable refrain Eey haaliyaa hiheyho heyho ooho. Prolonged haaliya-type. Vertical: pitch, horizontal: time (1 column = 1 second). Performed by Norman Carlo, 2016.

Example 90 Final vocable refrain Hiya hooheyho using the vocable hiya. Performed by Norman Carlo, 2016.

Example 91 Long final vocable refrain with the vocables ho and ha dominating. Performed by Norman Carlo, 2016.

A couple of his ch’edzes ch’elik have the ending Hiya ho-o-hey-ho or Hey hiya ho-o-hey ho. There is also a long ending, dominated by the vocables ho and ha (Example 91).

The first part of the ch’edzes ch’elik, consisting entirely of vowels, is less formulaic. This is – as seen above – the section that characterizes a composition by the combination of the key word(s) and a pregnant or original melodic or rhythmic motif (Example 92).

Example 92 Initial melodic motif. Performed by Norman Carlo, 2016.

There are two main types of motifs for the beginning:

• a melodic motif with much (often mainly descending) movement between pitches (Example 93; compare for instance Example 84, Christmas tree), and
• a flat motif at a high pitch with little melodic movement, usually at one interval and/or micro-intervals (Example 94; compare for instance Example 86, Caribou people 2).

Musical form

Though there is much variation in detail, there is also a good deal of repetition of material, particularly in the final vocable parts. This can be illustrated as follows. The basic form A–B may start with either a melodic motif or a flat motif. Both A and B may continue as low final vocables or as high (Example 95). The A section may contain lexical words (usually one or two), vocables only, or a combination of both.

Example 93 An initial melodic motif is often repeated, resulting in ‘paired phrases’: a–a, marked with a vertical line. Performed by Norman Carlo, 2016.

A ch’edzes ch’elik may be made longer by means of repetition of the motif in A, either identical or with minor variation, resulting in an A–A–B form (Example 96).

Prolongation may be achieved by descending more slowly and inserting an intermediate vocable section (Example 97). This is usually called an A–B–C form; but since the B and C parts are usually quite similar, A–B–B′ would also be a possible description.

There are many ways to vary a composition of this type, for example by combining and repeating the form components in various ways.

Performance templates and composition

Among the vocal expressions, dratakh ch’elik are composed for specific memorial situations, and they are sometimes commissioned from people known to be skilled song-makers. The dratakh ch’elik studied here are old and rather well known, but others are still being composed in the Minto tradition. The framework that has been described here combines comparatively set patterns with flexibility and shares many characteristics with improvised musical styles.⁴³ This is not an improvisatory genre per se, but the process of composition and that of improvisation have much in common.

Example 94 Initial flat motif repeated with variation: a–a′, marked with a vertical line. Performed by Norman Carlo, 2016.

Example 95 Main outline of the A–B form of composition.

Example 96 Main outline of an A–A–B form of composition.

It seems to be the linguistic–musical–poetic framework that makes it possible to compose a stylistically satisfactory ch’edzes ch’elik in rather a short time. This framework thus has the basic function of a performance template. Coupled with the fact that ch’edzes ch’elik are collectively performed and heard in potlatches, the existence and efficiency of this method of composition are likely to play an important part in the survival of this Minto tradition, as well as for the transmission of Minto culture and language to younger generations.

Example 97 Main outline of the A–B–C form of composition.

Ch’edzes ch’elik for the dancing situation are composed either ex tempore on sudden inspiration, or by developing a basic idea by trying it out on others, privately or in the dancing situation. Most ch’edzes ch’elik share a similar basic structure and are made up of a few words with lexical meaning at the beginning, while vocables make up the remainder; there are also some that consist entirely of vocables. The most original part is the very beginning, where the words with meaning are combined with a melodic/rhythmic motif. The latter part tends to be more formulaic, and the same or slightly varied endings may be used for many different ch’edzes ch’elik. However, these parts also tend to become stable once a composition is fixed.

One may look on the process of composition as a form of improvisation that differs from ex tempore improvisation in that it will be further perfected until – as in Athabascan practice – it achieves a final form, which is then retained basically without changes each time it is performed. If this position is taken, it is relevant to use the concept of a performance template in this case as well. What has been outlined in the analysis is the approximate basic character of such templates. It should be noted that there are other vocal genres in the Athabascan tradition, such as the senh ch’eliga’, ‘medicine song’ (Analysis 12), which have a completely different form and are more closely related to intonation, with neither vocables nor tone repetition. There is no evidence to the effect that vocal expressions of this genre are ever newly composed.

By way of conclusion, the Athabascan vocal genres dratakh ch’elik and ch’edzes ch’elik may be approached via performance templates, and this approach leads to an understanding of them as vocal expressions in which language and music are one. One important function of the vocables proves to be that of making composition of these genres possible, even composition of ch’edzes ch’elik completely made up of non-lexical words, which are then repeated identically each time the composition is performed. Yet another function on the part of the vocables in Athabascan tradition has been demonstrated: they form the basis of the performance template that facilitates composition. Perhaps this was what Geraldine Charlie meant in an interview when she said that ‘the vocables are there to keep the words in place’.⁴⁴

The ch’edzes ch’elik performance template

Melody

• Melody generally descending stepwise in a sequential manner and with a tonal centre (see Examples 84, 86).
• Range: 6–16 semitones.

Rhythm

• Regular pulse.
• One pulse beat per syllable dominates.
• Pulse beats may be subdivided into two; dotted rhythms occur.
• Isorhythmic organization prominent.
• Vocable sequence patterns often arranged in groups of 4 beats.

Form

• Strophic.
• Binary form A–B or A–B–C with variations: A has vocables and key word(s) with lexical meaning, other sections consist of vocables (see Examples 95–97).
• In A, the key word and ‘corresponding’ vocables form a poignant musical motif or ‘building block’.

Phrasing

• The final syllable of a phrase is lengthened, and phrases end with tone repetitions.

Initial/final formulae

• Ends with a tone repetition (often on the tonal centre); the very last vocable of the performance is short and stressed.

Word variations

• Vocables dominate.
• Lexical words may be pronounced differently from ordinary speech.
• Final tone repetitions consist of vocables only, sometimes with one lexical word (haaliyaa) that functions as a ‘song-word’.