iDIDJ Australia Didgeridoo Cultural Hub

Sung and spoken: an analysis of two different versions of a Kun-barlang love song.

Article from:Australian Aboriginal Studies
Article date:September 22, 2007
Author:O'Keeffe, Isabel

Abstract: In examining a sung version and a spoken version of a Kun-barlang love song text recorded by Alice Moyle in 1962, I outline the context and overall structure of the song, then provide a detailed comparative analysis of the two versions. I draw some preliminary conclusions about the nature of Kun-barlang song language, particularly in relation to the rhythmic setting of words in song texts and the use of vocables as structural markers.

**********

In 1962 Alice Moyle recorded two versions of a Kun-barlang love song text: a sung version and a spoken version (Moyle 1997). (1) The song belongs to a genre of individually owned public love songs in western Arnhem Land, usually referred to by the term kun-borrk. (2) Like other kun-borrk song sets (Garde 2006), Kun-barlang love songs are sung predominantly in everyday language, although they sometimes contain 'song words' or 'spirit language'. (3) The particular song discussed here is in everyday Kun-barlang, which allows for a detailed comparison of the words in the sung and spoken versions. The fact that differences exist between the spoken and sung versions of Aboriginal song texts has been widely recognised by researchers, (4) and provides the potential to compare song language and everyday language. Although this brief comparison of the two versions of the Kun-barlang love song shows that the texts are very similar phonologically and morphosyntactically, detailed comparison of the texts shows differences in the forms of the texts. An examination of these differences provides insights into the way the text is structured and set rhythmically in song.

Like Kuninjku kun-borrk songs (Garde 2006), this Kun-barlang love song text features typical referential ellipsis, but Aboriginal consultants I have worked with identify a narrative with two characters--one man and one woman. (5) The first line is attributed to the man, speaking to the woman, and the rest of the utterances are attributed to the woman, speaking to the man. (6) A brief synopsis of the narrative contained in the song's text is: The man tells the woman to fetch water and grabs her arm. She tells him not to grab her arm and accuses him of not listening and of having no shame. The reason for the woman's utterances is expressed in the penultimate line: 'You're the one who is the wrong skin for me' (7) and she ends dramatically: 'Now I'm going to die, I'll forget about breathing.'

Several different versions of this Kun-barlang love song exist, as shown in Table 1.

I focus on the first two recordings 1 and 2 above, the sung and spoken versions recorded by Alice Moyle in 1962 (Moyle 1997). These two versions provide interesting data because they are both full versions of the song text and the person singing and speaking is the same. There are some problems in comparing the two versions, however, because the spoken version is hyper-articulated, while the sung text displays considerable lenition, characteristic of song. Because of this, I have not analysed the differences in phonetic realisation between the two versions and my transcription of the sung text does not attempt to show different realisations of words, such as vowel quality. (12) Rather, I focus on the differences in form, particularly the addition of vocables and the deletion and substitution of words in the sung version. The first part outlines the context, composers and performers of the song. The second part examines the musical and textual structure followed by a more detailed examination of the rhythmic setting of the words within this structure. This, in turn, is followed by a systematic line-by-line analysis of the song, in which the differences in form between the two versions are highlighted. These observations are drawn together to present preliminary conclusions that will be important for future research.

Context

The Kun-barlang love song analysed here belongs to a genre of individually owned public love songs in western Arnhem Land, referred to in the related Bininj Kun-wok dialects as kun-borrk. Within the genre of public kun-borrk songs there are numerous song sets, (13) which are made up of song items and are usually owned by an individual songman (the composer, or after his death, one or more of his male relatives or descendants). The song texts are often quite short and are usually repeated in performance. The songs are always accompanied by didjeridu and clapsticks and sometimes a second singer or audience members provide hand clapping. They are sometimes accompanied by dances associated with the particular song set. As previously mentioned, Kun-barlang love songs and other kun-borrk song sets are sung predominantly in everyday language, although they sometimes contain 'song words' or 'spirit language'.

The Kun-barlang love song discussed here has been recorded as one of the songs in the multilingual karrbarda 'long yam (Dioscorea transversa)' song set (14) (Barwick 2005, 2006; Bickerdike 2006a; Marett 2005a) owned and performed today by the two brothers, Bruce and Barndanj Nabegeyo. The status of this Kun-barlang song within the karrbarda song set is unclear. It has differences from the other songs in the karrbarda song set, and it is usually referred to as a 'love song' rather than a karrbarda song when I have discussed it with Kun-barlang, Mawng and Kunwinjku consultants. It is the only song within the song set sung entirely in Kun-barlang, and it is longer and more through composed than the others, which tend to have repeated verses and are more strophic in form. Recent performances of karrbarda songs have contained songs predominantly in 'spirit' or 'song' language, although there have also been some in everyday languages (Kunwinjku, Kunbarlang and one bilingual Mawng/Kun-barlang song). The everyday language songs are said to be 'love songs' and the spirit language songs are said to be 'spirit/dreaming' songs. It seems that in other song-sets in the region, these two categories or genres are mutually exclusive. It may be that the karrbarda song-set now contains more than one original set of songs--a set of 'love songs' as well as a set of 'spirit/dreaming' songs--or it may be that some song sets, such as this one, blur the distinction between these two categories.

The composer, singers and speakers

The composer of the song was Bob Balir-balir. In Moyle's 1967/97 disc notes he is referred to as a Gunbalang (Kun-barlang) man, however, Aboriginal consultants have said he was a Kunwinjku man of the Yalama clan, who was living at Goulburn Island for some time during the 1950s and 1960s. (15) Although he was a Kunwinjku man, he composed songs in Mawng and Kun-barlang, as well as Kunwinjku (Berndt and Berndt 1951:212-40). Another song set, the itpi-itpi 'grasshopper' song set, also has song items in Kunwinjku, Kun-barlang and Mawng. By contrast with the karrbarda song set, however, the itpi-itpi song set consists entirely of songs in everyday language.

In the 1967 versions recorded by Moyle, Bilinyarra, the brother of the composer Balir-balir, is both the singer and the speaker. Like his brother, his main language was Kunwinjku, but it is likely that he also spoke Mawng and Kunbarlang. There is no information about the way the text was sourced from the singer, but it is interesting that Moyle managed to elicit the entire text. (16)

The singers for the Barwick and Marett 2005 recordings and Barwick and Bickerdike 2006 recordings were the brothers Bruce Nabegeyo and Barndanj 'Thomas' Nabegeyo, sons of Bilinyarra. Their first language is Kunwinjku but they have some knowledge of Kun-barlang and Mawng. (17) There are very few changes in the song texts in the versions they sang (Barwick 2005, 2006; Bickerdike 2006a; Marett 2005a) from the Moyle 1967 version; it appears that their limited knowledge of Kun-barlang has not resulted in any significant changes to the song text.

Musical structure of the Moyle 1967 song

The song begins with an instrumental introduction, which is followed by a sequence of vocal sections and instrumental interludes as set out in Table 2. The didjeridu begins the introduction and continues throughout the song. The clapsticks start some beats after the didjeridu and after two initial gapped clapstick beats ([quavev] [??] [quaven] [??]), the clapstick pattern is one clapstick note to every quaver beat ([??] [??] [??] ...) at a rate of approximately 136 crotchet beats per minute. This pattern continues throughout the song except at three points, where the gapped clapstick patterns C1, C2 and C3 are used (Table 2). The interludes in which these gapped clapstick patterns occur (shaded in Table 2) mark breaks between the vocal sections of the song.

Figures 1 to 3 show the gapped clapstick patterns and the handclaps (x) that fall on each crotchet beat. Gaps also occur in the didjeridu pattern and these are shown with arrows. Clapstick Pattern 1 marks the interlude between verses, and clapstick Pattern 2 marks the end of a verse that will not be repeated. Note that the second half of clapstick Pattern 2 (marked B in Figure 2) derives from clapstick Pattern 1 (so clapstick Pattern 2 is made up of two parts: A+B). Clapstick Pattern 3 marks the end of the song item. The first part of this clapstick pattern incorporates clapstick Pattern 2, so clapstick Pattern 3 is made up of three parts, A+B+C.

[FIGURE 1 OMITTED]

[FIGURE 2 OMITTED]

[FIGURE 3 OMITTED]

Textual structure of the Moyle 1967 song

The textual structure of the Moyle 1967 sung text is shown in . It is divided into verses (V), which are subdivided into sections (S), which in turn are subdivided into lines (L). The numbers in the penultimate nodes of the tree diagram show the number of rhythmic cells in each line. Because rhythmic cells are defined here as the textual rhythm occupying a full crotchet beat, the number also indicates the number of beats. The final nodes show the number of syllables contained in each line. The rest of this section provides a detailed explanation of these divisions and how they have been determined.

Verses (V)

In the first branching node from the apical level of song item, (18) shows the division of the song into verses. These were determined on the basis of textual and musical analysis. Verse 1, which may be repeated, is a separate body of text from Verse 2. (19)

[FIGURE 4 OMITTED]

Sections (S)

The level below verses are sections (S1, S2) which are sung in one breath. Between Sections 1 and 2 in Verse 1, there are eight beats of rest in the singing, but between Sections 3 and 4 in Verse 2, there is only half a beat. (20) In both cases, however, the rests provide time for the singer to take a breath. The sections seem to contain separate episodes within the narrative of the song-text.

Lines (L)

Each section is divided into lines. Bilinyarra's pauses between each line in the Moyle 1967 spoken version, were used to determine this division. The repetition of Line 1 provides further evidence of its status as a separate unit. These repetitions are labelled 1.1 and 1.2 since they repeat almost exactly. Each line contains one or more phrases that are grammatically and semantically complete. (21) In addition, lines tend to begin on a metrically strong beat (where a clapstick beat and a hand-clap beat occur together).

Beats and syllables per line and syllables per beat

The penultimate level in Figure 4 indicates the number of beats/rhythmic cells contained in each line, which varies from four to ten. The final level shows the number of syllables per line, which also varies (from eight to 17).

Figure 5, which summarises the text and rhythm of one line (Verse 1, Line 1.1), shows that the number of syllables per beat may also vary (from 1 to 4 syllables per beat). The top line (rh) of the table indicates the rhythm of the sung text, the second line (mr) gives a morphological representation of the text and the final line (ge) provides an English gloss.

Rhythmic setting of the words

In some cases, one word fits into one rhythmic cell, see for example kenda-bonj in Figure 5. In other cases each syllable in a word takes up a whole rhythmic cell (for example kinj-rdam in Figure 5). A word may also be spread across cells (for example kinj-bardi-kadan, Figure 5).

In some environments, the metrical position of a word in a line is a factor in determining its rhythmic setting. As shown in Table 3, each line usually begins with initial syllables set to semi-quavers (22). Moreover, words at the beginnings of the line fall within 1 to 1.5 rhythmic cells.

In contrast, each line ends with a lengthened syllable of at least a crotchet beat or with a vocable syllable that 'steals' time from the previous word, as shown in Table 4 (shaded cells indicate lines ending with a vocable syllable, which is marked in bold). (23) This lengthening of final syllables in a line, or 'rhythmic prolongation' is also common in other Australian songs, including wangga, public songs from the Daly Region (Marett 2005b), and jurtbirrk songs from Croker Island (Barwick, Birch and Evans this volume). As Marett (2005b:198) theorised in regard to wangga. 'To enhance the definition of semantic units, and hence the comprehensibility of texts in performance, word and sentence boundaries tend to be marked by lengthening final syllables'.

Sequences of five syllables in succession (within one word, or a three syllable word plus a two syllable word) seem to require special treatment. Five-syllable sequences always have a lengthened final syllable (a crotchet beat or longer) or are followed by a vocable syllable (no or la), as shown in Table 5.

The rhythmic patterns used to set the words to music mostly conform to the general pattern noted by Marett (2005b) for wangga, which involves syllable lengthening at the end of words as well as at the end of lines. (24)

Line-by-line analysis of the two Moyle versions

The following analysis considers each line in turn, first presenting the spoken version of the text and then the sung version. This is followed by a table (or tables) that summarises the rhythmic setting of the text. Where possible, this data will be used to account for differences between the spoken and sung version.

The main difference between the two versions (Tables 6 & 7) is the addition of the vocable no in the sung version. It has no meaning in spoken Kunbarlang and it seems to act as a filler, particularly after five-syllable words, which, as noted above, require a particular rhythmic treatment. The use of filler syllables required by metrical necessity is also found in other Australian Aboriginal songs such as Marri Ngarr lirrga songs from the Daly region (Ford 2006:28), Dyirbal songs from Cape York (Koch 1987:55-9) and Ngiyambaa songs from New South Wales (Donaldson 1987:31). In this Kun-barlang song, the vocable no is interpolated after the first word in both Lines 1.1 and 1.2 and it is added to the end of Line 1.2. The no syllable does not occur at the end of Line 1.1, but the last syllable rdam is longer (covering a full crotchet beat) whereas in Line 1.2 no steals time from this syllable (making both of them quavers).

The other interesting feature in Line 1 is the substitution of kadan in the sung version for kali in the spoken version. Kadan has no meaning in everyday spoken Kun-barlang, so it is not entirely clear why it is used here. A possible explanation is that there is some kind of syllable harmony in operation, with the consonant [d], which begins the second syllable in the following word kendabonj, being copied and replacing the lateral approximate [l] in -kali.

While the spoken and sung versions of this line have the same number of words and syllables, the sung version lengthens the last two syllables karrme. The final vowel is longer than the lengthening at the end of Line 1.1, which may be related to the fact that it is marking a section break. The principle of iconicity may also provide a reason for this vowel lengthening: the verb karrme in this context has the sense of 'holding' or 'grasping' and this is perhaps expressed by 'holding' the note for longer. It is significant that when this verb occurs again at the end of Section 2 (in Line 5) but with the sense of 'having' rather than 'holding', the vowel is not lengthened.

In Line 3 (Table 9) there is again interpolation of the no vocable after the first five syllables in the line. The addition of the syllable la, at the end of the line, could have several of explanations. It may simply be a meaningless vocable syllable, homophonous with an everyday spoken Kun-barlang word (a conjunction), similar to the vocable ya described by Ford (2006:29) in Marri Ngarr lirrga. Alternatively, the presence of la in this position may be due to the fact that the previous word kadda-walkki-bing has the dispreferred five syllables and therefore needs an additional syllable. However, this la syllable also seems to retain some of its function as a conjunction, linking the lines. It is added at the end of each line in Section 2, except for the last one, suggesting that it is functioning as a structural marker between each of the lines in Section 2.

In the sung version of Line 4 (Table 10) the negative particle ngunda is re-duplicated with an extra syllable no interpolated both times. The second no occurs after the first five syllables. The rhythmic pattern is also reduplicated as the words ngunda no are both set to the same rhythm. The rhythmic pattern of this line and the number of syllables is the same in the Barwick and Marett 2005 and Barwick and Bickerdike 2006 recordings, but instead of singing a leftward repetition of ngunda no Bruce and Barndanj Nabegeyo sing ninda-nuk, the words which begin the following Line 5 (Table 12). This is shown in Table 11 below.

The additional la syllable at the end of the line seems to link it to the next line in this section. Note that the la syllable also follows a sequence of five syllables over two words, ki-djin binungku.

In the sung version of Line 5, the word kundji 'shame' occurs twice, once in the same position as in the spoken version and once as a substitution for the last morpheme of ka-karlmu-barrdja in the spoken version (Tables 12 and 13). Although it also has two syllables, this seems to be a strange substitution, since -barrdja 'stand.NP' is a bound verb and kundji 'shame' a free noun. The sentence in the sung version would thus be ungrammatical. The morpheme -barrdja 'stand.NP' never occurs in any of the sung versions; however, speakers usually include it when a spoken text is elicited from the sung version. This suggests that many speakers have knowledge of an underlying spoken text and are able to produce words that do not occur in the sung texts. It is possible that the singer has used vocables and the repeated word kundji instead of -barrdja in order to cover a lapse in memory, and that these words have then been copied in subsequent performances (replacing barrdja), but this explanation is purely speculative at this stage.

The interpolated la occurs after the two-word, five-syllable sequence ka-karlmu kundji, and also conjoins the two phrases in Line 5 (Tables 12 and 13).

There are no changes in the form of Line 6 (Table 14) except for the loss of the final syllable in wularrud. In Line 3 the word wularrud is set to the rhythm [??] but here in Line 6 the final syllable is lost and it is set to: [??] It is unclear why this change oc-curs; however, it may be to create rhythmic alliteration between the two words karrard and wularrud.

Again, there are no changes in form in Line 7, but note that the final word, which has the lengthened final syllable, is a five-syllable word (Table 15).

Line 8 (Table 16) has some unusual features and this may relate to the fact that it contains the punch line of the song: 'You're the one who is the wrong skin for me'. Here the female character is telling the male character 'you shouldn't be in a romantic relationship with me--it's not appropriate'. (26) This piece of information, not revealed by the singer until here, explains earlier lines of the song, such as the woman saying 'don't hold my arm' and 'you've got no shame'.

One of the unusual features in this line is the additional syllable la at the start of the line on an upbeat. This may be functioning either as a conjunction, with the sense of 'but,' or simply as a vocable syllable, perhaps due to the fact that the previous line finished with a five-syllable word. This previous word is, however, in a separate section. Another unusual feature is that the final syllable of Line 8 is tied over to the next line. While this could suggest that these two lines are actually part of the same line, in the spoken version the speaker makes a clear pause between ki-ngan-kakkin-wandarrkbum and balkkime, so I have labelled them as separate lines. The last word of Line 8 is also the only seven-syllable word in the song, making it a more difficult word to set rhythmically. The tie over into the following line allows Line 8 to still have a sense of final lengthening. Lengthening of this type occurs at the end of all lines that do not have a final vocable syllable.

As previously mentioned, Line 9 begins with the unusual feature of a tied-over word from Line 8, which is followed by the syllable la, which may again be either functioning as a conjunction linking the two lines, or as a vocable syllable, providing an additional syllable to the previous seven-syllable word (Table 17). There is also an interpolated nuk which may be operating both as a dubitative particle ('maybe') and also as an additional syllable for the following two words which make up a dispreferred five syllable sequence.

Conclusion

The changes in form between the Moyle 1967 sung and spoken versions include the repetition and substitution of words, and, most noticeably, the interpolation of the vocable syllables no and la. There are not enough examples of the repetition and substitution of words to clearly understand their role in this song. The vocable syllables, however, contribute to the structural organisation of the song and the setting of the text to music.

While the number of beats and syllables per line and syllables per beat are clearly variable in this song, there are some identifiable patterns in the rhythmic setting of words, particularly at the beginnings and ends of lines. Short syllables at the beginnings and longer syllables at the ends of words and lines were shown to conform generally to the patterns noted for other Australian songs, including wangga (Marett 2005b) and jurtbirrk (Barwick et al. this volume). Within Sections 1 and 2 of this song, however, the vocable syllables no and la occur at the ends of lines instead, stealing time from the final lengthened syllable. These vocables seem to mark the ends of lines within a section, while lengthened syllables seem to mark the ends of sections, as shown in Table 18. While this holds in Sections 1 and 2 (Verse 1), it does not hold true for all the lines and sections of Verse 2.

This use of vocables as structural markers is similar to what Koch (1987:55-9) found for Dyirbal gama songs, where there was a nga vocable inserted between each line in a musical phrase. Donaldson (1987:31) also noted that there was a nga vocable inserted between lines in the sung versions of certain Ngiyambaa songs. Vocable syllables used as structural markers have also been noted in many indigenous songs in North America, including Navajo, Kwakiutl and Coast Tsimshian (Mulder 1994).

The vocable syllables no and la also seem to contribute to the rhythmic setting of the text by providing an additional syllable to the end of five-syllable sequences, which seem to be dispreferred. This use of vocables as fillers to fit in with metrical constraints conforms to some general patterns found in other Australian Aboriginal music (Donaldson 1987; Koch 1987; Ford 2006; Barwick et al. this volume). Future investigation, taking into account the broader repertoire of Kun-barlang songs and other kun-borrk songs, will no doubt provide further insight into the use of vocables and other variations between sung and spoken song texts across the repertoires. This will provide points of comparison with other song traditions in Australia and elsewhere.

Appendix 1: Abbreviations

APPL Applicative BEN Benefactive COM Comitative applicative conj Conjunction DUB Dubitative EMP Emphatic FUT Future IMP Imperative KB Karrbarda song LL Land gender MA Masculine gender NEG Negative NP Non-Past O Object P Past PROX Proximal demonstrative R Realis sg Singular SW Song word (vocable)


Appendix 2: Spoken Version (Moyle 1997)

Verse 1, Section 1

Line 1

kinj-bardi-kalikenda-bonj kinj-rdam

2sg.FUT-liquid-fetch.NPLL.PROX-EMP2sg.FUT-put.NP

Fetch water, put it right here.

Line 2

kanj-yuwa njan-kundu-karrme
3sg.FUT-lie.NP 2sg.IMP-arm-hold.NP

Leave it, don't hold my arm.

Verse 1, Section 2

Line 3

kinj-ngayi barndanwularrud kadda-walkki-bing
2sg.FUT-listen.NP fight already3p1.R-APPL-arrive.P

Listen, they have come already with trouble.

Line 4

ngunda karlmuki-djin bi-nungku
NEG ears 2sg.R-stand.P BEN-2sg

You've got no ears; you don't listen for yourself.

Line 5

ninda-nukka-karlmu-barrdjangunda kundji ki-bun-karrme
MA.PROX-EMP 3sg.R-ears-stand.NP NEG shame 2sg.R-3sg.0-have.NP

He's the one who doesn't listen. You've got no shame.

Verse 2, Section 3

Line 6

ngayi-nuk karrard wularrud ka-warrenj ka-ngan-ngunda
1sg-EMPmother already 3sg.R-go.along.P 3sg.R-lsg.0-threaten.P

But my mother used to go along growling at me.

Line 7

ka-warrenj ka-ngan-karlmu-kalng
3sg.R-go.along.P 3sg.R-1sg.0-ear-take.P

She used to discipline me [lit. take my ear].

Verse 2, Section 4

Line 8

ngudda-bonj ki-ngan-kakkin-wandarrkbum
2sg.-EMP 2sg.R-1sg.0-meat-make.mistake.P

You're the one who is the wrong skin for me.

Line 9

balkkime nganj-warrmiberrngek nganj-wakwanj
now1sg.FUT-forget.NP breath 1sg.FUT-die.NP

Now I'm going to die, I'll forget about breathing.

Appendix 3: Sung Version (Moyle 1997)

Verse 1, Section 1

Line 1: V1-S1-L1.1

kinj-barli-kadan no kenda-bonjkinj-rdam
2sgFUT-liquid-fetch.NP SW LL.PROX-EMP 2sg.FUT-put.NP

Fetch water, put it right here.

V1-S1-L1.2

kinj-barli-kadan no kanda-bonj kinj-rdamno
2sgFUT-liquid-fetch.NP SW LL.PROX-EMP 2sg.FUT-put.NP SW

Fetch water, put it right here.

Line 2: V1-S1-L2

kanj-yuwa njan-kundu-karrme --
3sg.FUT-lie.NP2sg.IMP-arm-hold.NP

Leave it, don't hold my arm.

Verse 1, Section 2

Line 3: V1-S2-L3

kinj-ngayi barndan no wularrud kadaa-walkki-bing la
2sg.FUT-listen.NP fightSW already 3pl.Real-APPL-arrive.P conj

Listen, they have come already with trouble.

Line 4: V1-S2-L4

ngunda no ngunda no karlmu ki-djinbi-nungku la
NEG SW NEG SW ears 2sg.R-stand.P BEN-2sg conj

You've got no ears; you don't listen for yourself.

Line 5: V1-S2-L5

ninda-nuk ka-karlmu kundji la ngunda kundji ki-bun-karrme
MA.PROX-EMP 3sg.R-ears shame conj NEGshame 2sg.R-3sgO-have.NP

He's the one who doesn't listen. You've got no shame.

Verse 2, Section 3

Line 6: V2-S3-L6

ngayi-nuk karrard wularr(ud) ka-warrenj ka-ngan-ngunga
1sg-EMP motheralready 3sg.R-go.P 3sg.R-1sg.0-threaten.P

But my mother used to go along growling at me.

Line 7

V2-S3-L8

ka-warrenj ka-ngan-karlmu-kalng
3sg.R-go.along.P 3sg.R-1sg.0-ear-take.P

She used to discipline me [lit.take my ear].

Verse 2, Section 4

Line 8: V2-S4-L8

la ngudda-bonj ki-ngan-kakkin-wandarrkbum
conj 2sg.EMP 2sg.R-1sg.O-meat-make.mistake.P

(And) You're the one who is the wrong skin or me.

Line 9

V2-S4-L9

la balkkime nganj-warrmi nuk berrngek nganj-wakwanj
conj now1sg.FUT-die.NP DUB breath 1sg.FUT-forget.NP

(And) Now I'm going to die, maybe I'll forget about breathing.
REFERENCES

Barwick, Linda 2005 'Karrbarda (yam) kunborrk performed at Kunbarllanjnja', Bar05-1, deposited with AIATSIS, Canberra.

-- 2006 'Karrbarda (yam) kunborrk performed at Kunbarllanjnja', deposited at Endangered Languages Archive, London.

Barwick, Linda, Bruce Birch and Nicholas Evans (this volume) 'Iwaidja Jurtbirrk Songs: Bringing language and music together'.

Berndt, Ronald M and Catherine H 1951 Sexual Behavior in Western Arnhem Land, Viking Fund, New York.

-- 1970 Man, Land and Myth in Northern Australia: The Gunwinggu people, Ure Smith, Sydney.

Bickerdike, Isabel 2006a 'Karrbarda (yam) kunborrk performed at Kunbarllanjnja', to be deposited with AIATSIS, Canberra.

-- 2006b Fieldnotes and field recordings, to be deposited with Endangered Languages Archive and AIATSIS.

Donaldson, Tamsin 1979 'Translating oral literature: Aboriginal song texts', Aboriginal History 3(1):62-84.

-- 1987 'Making a song (and dance) in south-eastern Australia', in M Clunies Ross, T Donaldson and SA Wild (eds) Songs of Aboriginal Australia, University of Sydney, pp. 14-42.

Ford, Lysbeth 2006 'Marri Ngarr Lirrga songs: A linguistic analysis', Musicology Australia 28:26-8.

Garde, Murray 2006 'The language of Kun-borrk in western Arnhem Land', Musicology Australia 28:59-89.

Koch, Grace 1987 'Dyirbal Gama songs of Cape York', in M Clunies Ross, T Donaldson and SA Wild (eds) Songs of Aboriginal Australia, University of Sydney, pp.43-62.

Marett, Allan 2005a 'Karrbarda (yam) kunborrk performed at Kunbarllanjnja', Mar05-07, deposited with AIATSIS, Canberra.

-- 2005b Songs, Dreamings and Ghosts, Wesleyan University Press, Middletown.

Merlan, Francesca 1987 'Catfish and alligator: Totemic songs of the western Roper River, Northern Territory', in M Clunies Ross, T Donaldson and SA Wild (eds) Songs of Aboriginal Australia, University of Sydney, pp. 142-67.

Moyle, Alice M 1997 [1967] Songs from the Northern Territory 1, Australian Institute of Aboriginal and Torres Strait Islander Studies, Canberra (LP disc recording re-released on audio compact disc).

Mulder, Jean 1994 'Structural organization in coast Tsimshian music', Ethnomusicology 38(1):81-125.

Strehlow, TGH 1971 Songs of Central Australia, Angus & Robertson, Sydney.

Turpin, Myfany 2005 'Form and Meaning of Akwelye: A Kaytetye women's song series from Central Australia', unpublished doctoral thesis, University of Sydney.

Isabel O'Keeffe (nee Bickerdike)

university of Melbourne

NOTES

(1.) The full texts of both versions of the Moyle 1967 recording can be found in Appendices 2 and 3.

(2.) The word kun-borrk is used in the dialects of Bininj Kun-wok. Kun-barlang speakers often use this word too, but some Kun-barlang consultants have said that it is really 'just a Kunwinjku word' and that the Kun-barlang word is manyardi (cognate with the Mawng word manyarti that is used to describe any song).

(3.) 'Song words' or 'spirit language' refers to the language used in songs that is unknown to the singers. The language is usually attributed to the spirits who have given the composer the song in a dream.

(4.) For example Strehlow 1971; Koch 1987; Merlan 1987; Donaldson 1979; Turpin 2005; Ford 2006; Barwick, Birch and Evans in this volume.

(5.) The characters are indexed only by the use of pronouns (mostly second person pronominal prefixes) so that it is impossible to identify the characters, their genders, or which character is associated with which utterance simply from the text. Character identity and direct speech attribution was determined via the shared background knowledge and pragmatic inferences of Aboriginal consultants with whom I have worked. Much entertainment value is derived from speculating about the actors and the situation to which the singer was referring.

(6.) In Line 5 of the song text (Appendices 2 and 3) there is a third person pronominal prefix that is co-referential with the male character. This seems to indicate that, in this line of the song, the female character is speaking directly to the audience, or to other unknown participants in the song narrative.

(7.) 'Wrong skin' refers to the fact that the man is in the wrong marriage class for the woman. Traditionally, fairly strict rules governed the choice of marriage partners (Berndt and Berndt 1951, 1970).

(8.) Track 6b is a recording of the didjeridu accompaniment alone, played by Lazarus Lamilami. However, it is beyond the scope of this paper to analyse this recording.

(9.) The Barwick and Marett recordings are of the same song session. I am grateful to them both for providing me copies of their field-recordings, photos from the sessions and information about the recordings.

(10.) I am very grateful to Ngal-Kangila Margaret Marrangu, Ngal-Kangila Mary Mundanmari, Ngal-Kangila Mary Gurden-gurden (d.2007), Ngal-Kamarrang Margaret Marlingarr, Ngal-Wakadj Roseanne Gameraidj, Na-Kodjok Frank Ambidjambidj, Ngal-Ngarridj Sandra Makurlngu, NgalNgarridj Shirley Mawuli, Na-Bangardi Solomon Imeraganyan, Dora Diaguma, Linda Najinja, Daisy Nadjandanga, Na-Kangila Bruce Nabegeyo and Gary Djorlam, for their help in transcribing and translating this song text and discussing it with me at various times. All errors or omissions are, of course, my own.

(11.) It appears to be a transcription of the spoken version since it contains the full word ka-karlmu-barrdja which does not occur in any of the recordings of the sung version.

(12.) As part of my doctoral research, I am planning analysis of the differences in the phonetic realisation of sung and spoken song text versions.

(13.) In using the term 'song set,' I refer to what has often been called in the literature 'song series' or 'song cycle', that is, a particular group of songs (song items), which belong together. However, following Garde and others I use the term 'song set' since 'song series' or 'song cycle' 'tends to suggest obligatory sequential performance of songs in a set which is not the case with kun-borrk' (Garde 2006:68).

(14.) Karrbarda is the Kunwinjku word for 'long yam' and it is the name by which the song set is usually referred to, probably because most of the owners and singers of the song set are Kunwinjku people. The Kun-barlang word for 'long yams' mabudj is only occasionally used. As Garde (2006:69) pointed out, many kun-borrk song sets are named after plants with edible tubers, such as this one.

(15.) Balir-balir may have had some Kun-barlang relatives, for example his mother may have been Kunbarlang, but children are classified by their father's language and clan grouping.

(16.) In my experience, it is often hard to get a singer to speak all of the words of a song--often they will only say parts of the text or will give a summary of the text.

(17.) While I was working with Bruce Nabegeyo in 2006, he was reluctant to help in transcribing the song as he said did not speak Kun-barlang. However, he was able to confirm some of the transcriptions I had already made with other consultants and he was able to give some translations of the Kun-barlang text into Kunwinjku and English.

(18.) A 'song item' refers to the entire sung text.

(19.) In the Barwick and Marett 2005 recordings there is no repeat of Verse 1. In the Moyle 1967 recording the repetition of Verse 1 has a slightly different melody but the text is exactly the same.

(20.) Similarly, Marett (2005b:93) noted that for wangga the singer may vary the length of the instrumental sections and the coda.

(21.) There is one exception in the Moyle 1967 sung version ( and related discussion).

(22.) Line 8 is an exception, although even here the first word, after the initial vocable, starts with a semiquaver.

(23.) The one exception is Line 8, which still has a sense of final lengthening since the last syllable is tied over into the next line (and related discussion).

(24.) The only words that do not conform to this convention are two words that appear in the last line (Line 9).

(25.) This vocable occurs only in L1.2--the second repetition of Line 1.

(26.) As previously mentioned, character identity and direct speech attribution can only be determined by shared background knowledge and pragmatic inference. Aboriginal consultants attribute this line (Line 9) to the female character.

(27.) Elsewhere I have glossed it as EMP 'Emphatic' however, nuk seems to have a dual function as an emphatic marker and a dubitative marker.

Isabel O'Keeffe is a doctoral student at the University of Melbourne working on a thesis called 'The Language and Music of Kun-barlang Songs'. She recently spent nine months living at Warruwi, South Goulburn Island, doing fieldwork research and making recordings for the Hans Rausing Project Classical Song Traditions of contemporary Western Arnhem Land in Their Multilingual Context, for which she is a student researcher.


Table 1: Different versions of the Kun-barlang love song

RecordingRecorderVersion Location
date/
Publication
date

1962/1967Moyle (1997)sungOenpelli
(track 6a)

1962/1967Moyle (1997)spoken Oenpelli
(track 6c) (8)

2005 Barwick (2005) sungAdjumarlarl
and Marett (near Oenpelli)
(2005a) (9)

2006 Barwick (2006) sungAdjumarlarl
and Bickerdike (near Oenpelli)
(2006a)

2006 Bickerdike 2006 spoken/ Warruwi (South
(2006b) written Goulburn Island)
(partial) and Oenpelli

?/1951 Berndt and Berndt written Warruwi (South
(1951:23) (partial) Goulburn Island)?

RecordingSinger/ Context/Details
date/Speaker
Publication
date

1962/1967Bilinyara Didjeridu
(track 6a) accompaniment by
Lazarus Lamilami.
Performance context,
possibly with dancing.
Audience members
can be heard in the
background and some
participants provide a
handclap rhythm.

1962/1967Bilinyara Text is presumably
(track 6c) (8) elicited by Moyle.

2005 Bruce and Didjeridu
Barndanjaccompaniment by
NabegeyoStephen Kanerradj.
Two separate
recordings.

2006 Bruce and Didjeridu
Barndanjaccompaniment by
NabegeyoStephen Kanerradj.
Two separate
recordings.

2006 various Partial versions of
Aboriginal the texts elicited by
consultants Bickerdike--audio
(10)recordings and
fieldnotes.

?/1951 unknown Partial version. Appear
to be a transcription of
the spoken version. (11)

Table 2: Musical structure of Moyle 1967 sung version

Instrumental introduction

Verse 1.1Vocal S1.1

interlude

Vocal S2.1

C1 interlude

Verse 1.2Vocal S1.2

interlude

Vocal S2.2

C2 interlude

Verse 2 Vocal S3

C3 instrumental final

Table 3: Rhythms at the start of each line

Verse/ Line # Word # syllables
Section #

V1.S1 1 kinj-bardi-kadan no 5 + 1
2 kanj-yuwa 3
3 ki-ngayi 3
V1.S2 4 gunda no 2 + 1
5 ninda-nuk ka 3 + 1
6 ngayi-nuk 3
V2.S1 7 ka-warrenj 3
8 la ngudda-bonj 4
V2.S2 9 la balkkime4

Verse/ Line # rhythm # of cells
Section #

V1.S1 1 [??] [??]1.5 (+ no)
2 [??]1
3 [??]1
V1.S2 4 [??]1
5 [??]1
6 [??]1
V2.S1 7 [??]1
8 [??] [quaven] [??]1.5
V2.S2 9 [??] [??]1.5

Table 4: Rhythms at the end of each line

Verse/ Line # Word# syllables
Section #

VLS1 1.1 kinj-rdam2
1.2 kinj-rdam no 3
2 njan-kundu-karrme5
V1.S2 3 kadda-walkki-bing la 5
4 bi-nungku la 3
5 ki-bun-karrme4
V2.S1 6 ka-ngan-ngunga 4
7 ka-ngan-karlmu-kalng la 5
V2.S2 8 ki-ngan-kakkin- 7
wandarrkbum la
9 nganj-wak-wanj 3

Verse/ Line #rhythm # of cells
Section #

VLS1 1.1[crochet][crochet]2
1.2[crochet][quaven] 2
[quaven]
2 [??] [crochet][crochet] 8
[crochet] [??]
V1.S2 3 [??][??] [crochet]3.5 (+ 1a)
[quaven][quaven]
4 [??] [quaven][quaven] 1.5 (+ la)
5 [semi-quaven][crochet] 3.25
[crochet][crochet]
V2.S1 6 [??] [crochet][crochet] 4
[crochet]
7 [??][??][crochet] 3.5 (+ la in
[quaven][??][quaven] the next line)
V2.S2 8 [??][??][??][quaven] 3.5 (+ la in
[semi-quaven] the next line)
9 [crochet][crochet]3
[crochet]

Table 5: Five syllable sequences

Word Rhythm

kinj-bardi-kadan no[??][??]
kenda-bonj kinj-rdam [??][crochet][crochet]
kenda-bonj kinj-rdam no[??][crochet][??]
njan-kundu-karrme--[??][crochet][crochet] [crochet] [??]
ki-ngayi barndan no[??][crochet][??]
ka-ngan-karlmu-kalng [??][??] [crochet]

Table 6: Verse 1, Section 1, Line 1.1

Line 1

Spoken version

kinj-bardi-kali kenda-bonj kinj-rdam
2sg.FUT liquid-fetch.NP LLTROX-EMP 2sg.FUT put.NP

Fetch water, put it right here

Sunq version

kinj-bardi-kadan no kenda-bonj kinj-rdam no (25)
2sg.FUT liquid-fetch.NP SW LL.PROX-EMP 2sg.FUTput.NP SW

Fetch water, put it right here

rh [??] [??][??]
mr kinj-bardi-kadan no kenda-bonj
ge 2sg.FUT-liquid-fetch.NP SW LL.PROX-EMP

rh [crochet] [crochet]
mr kinj-rdam
ge 2sg.FUTput.NP

Table 7: Verse 1, Section 1, Line 1.2

rh[??][??] [??]
mrkinj-bardi -kadan nokenda-bonj
ge~2sg.FUT liquid-fetch.NP SWLL.PROX-EMP

rh[crochet] [quaven] [quaven]
mrkinj-rdam no
ge2sg.FUT put.NP SW

Table 8: Verse 1, Section 1, Line 2

Line 2

Spoken version

kanj-yuwa njan- kundu-karrme
3sg.FUT lie.NP 2sg.IMP-arm- hold.NP

Leave it, don't hold my arm

Sung version

kanj-yuwa njan-kundu-karrme ee
3sg.FUT-1ie.NP 2sg.IMP-arm-hold.NP

Leave it, don't hold my arm

rh[??][??] [crochet]
mrkanj-yuwa njan-kundu-karr
Xe3sg.FUTIie.NP 2sg.IMP-arm-hold.NP

rh[crochet] [crochet] [crochet] [??]
mr-me --
Xelengthening

Table 9: Verse 1, Section 2, Line 3

Line 3

Spoken version

kinj-ngayi barndan wularrud kadda-walkki-bing
2sg.FUT listen.NP fight already 3p1.R-COM-arrive.P

Listen, they have come already with trouble.

Sung version

kinj-ngayi barndan no wularrud kadda-walkki-bing la
2s.g.FUT-listen.NP fight SW already 3pl.R-COM-arrive.P conj

Listen, they have come already with trouble.

rh [??][crochet] [??] [??] [??]
mr kinj-ngayi barn- dan no wularrud kadda-
ge 2sg.FUT-listen fight SW already3p1.R-

rh [??] [crochet] [quaven][quaven]
mr walkki- bingla
ge COM- arrive.Pconj

Table 10: Verse 1, Section 2, Line 4

Line 4

Spoken version

ngunda karlmu ki-djin bi-nungku
NEG ears 2sg.R-stand.P BEN-2sg

You've got no ears; you don't listen

Sung version

ngunda no ngunda no karlmu ki-djin bi-nungku la
NEG SW NEG SW ears 2sg.R-stand.P NEG BEN-2sg Conj

You've got no ears; you don't listen to it

rh [??][??][??] [??] [??]
mr ngunda no ngunda no karlmu ki-djinbi-nung
ge NEGSW NEGSW ear 2sg.R-eat.NP BEN-2sg

rh [quaven][quaven]
mr -ku la
ge conj

Table 11: Line 4 comparison of three versions

Spoken ngunda karlmu ki-djin bi-nungku
Sung 1 ngunda no ngunda no karlmu ki-djin bi-nungku la
SSung 2 (2005) ninda nuk ngunda no karlmu ki-djin bi-nungku la

Table 12: Verse 1, Section 2, Line 5

Line 5

Spoken version

ninda-nuk ka-karlmu-barrdja ngunda kundji ki-bun-karrme
MA.PROX-EMP 3sg.R-ears-stand.NP NEGshame 2sg.R-3sg.O-have.NP

This one here, he doesn't listen. You've got no shame.

Sung version

ninda-nuk ka-karlmu kundji la ngunda kundji ki-bun-karrme
MA.PROX-EMP 13sg.R-ears shame conj NEGshame 2sg.R-3sg.O-have.NP

This one here, he doesn't listen to it. You've got no shame.

rh [??] [??][quaven] [quaven]
mr ninda-nuk ka- karlmu kun-dji la
ge MA.PROX-EMP 3sg.R ear shame conj

rh [??] [??] [crochet][crochet] [crochet]
mr ngunda kun- dji ki- bun- karr- me
ge NEG shame 2sg.R- 3sg.O- have.NP

Table 13: Line 5 comparison of sung and spoken

Spoken ninda-nuk ka-karlmu -barrdjangunda kundji ki-bun-karrme
Sungninda-nuk ka-karlmu kundji la ngunda kundji ki-bun-karrme

Table 14: Verse 2, Section 3, Line 6
Line 6
Spoken version

ngayi-nuk karrard wularrud ka-warrenjka-ngan-ngunga
1st-EMPmother already3sg.R-go.along.P 3sg.R-1sg.O-threaten.P

But my mother used to go along growling at me.

Sung version

ngayi-nuk karrard wularr(ud) ka-warrenjka-ngan-ngunga
1sg-EMPmother already 3sg.R-go.along.P 3sg.R-1sg.O-
threaten.P

But my mother used to go along growling at me.

rh [??] [??] [??] [??]
mr ngayi-nuk karrardwularr ka-warrenj
ge 1st-EMPmother already3sg.R-go.along.P

rh [??] [crochet] [crochet] [crochet]
mr ka-ngan-ngu -nga --
ge 3sg.R-1sg.O-threaten.P

Table 15: Verse 2, Section 3, Line 7
Line 7
Spoken version

ka-warrenj ka-ngan-karlmu-kalng
3sg.R-go.al3sg.R-1sg.O-ear-take.P

She used to discipline me [lit. take my ear]

Sung version

ka-warrenj ka-ngan-karlmu-kalng
3sg.R-go.al3sg.R-1sg.O-ear-take.P

She used to discipline me [lit. take my ear]

rh [??] [??] [??] [crochet]
mr ka-warrenjka-ngan -karlmu -kalng
ge 3sg.R-go.along.P 3sg.R-1sg.O-ear-take.P

Table 16: Verse 2, Section 4, Line 8
Line 8

Spoken version
ngudda-bonjki-ngan-kakkin-wandarrkbum
2sg-EMP2sg.R-1sg.O-meat-make.mistake.P

You're the one who is the wrong skin for me.

Sung version

langudda-bonj ki-ngan-kakkin-wandarrkbum
conj 2sg.EMP 2sg.R-1sgl.O-meat-make.mistake.P

(But) You're the one who is the wrong skin for me.

rh [quaven] [??] [??] [??] [??] [quaven]
mr Langudda-bonj ki-ngan -kakkin -wandarrkbum
ge conj 2sg-EMP 2sg.R-1sg.O-meat-make.mistake.P

Table 17: Verse 2, Section 4, Line 9

Line 9
Spoken version

balkkime nganj-warrmi berrngek nganj-wakwanj
now 1sg.FUT-die.NP breath 1sg.FUT-forget.NP

Now I'm going to die, I'll forget about breathing.

Sung version

(-um) la balkkime nganj-warrmi nuk berrngek nganj-wakwanj
conj now 1sg.FUT-die.NP DUB27 breath 1sg.FUT-forget.NP

(And so) now I'm going to die, maybe I'll forget about breathing

rh [quaven] [??] [??][crochet]
mr -umlabal --kkime nganj -warr

rh [??] [crochet]
mr -mi nuk berr -ngek

ge conj now 1sg.FUT-die.NP

rh [crochet] [crochet] [crochet]
mr nganj -wak-wanj
ge 1sg.FUT-forget.NP

Table 18: Summary table showing structural markers

Verse Section Line

1 and optional repeat 1 Line 1.1
of 1 (slapstick pattern Line 1.2
1 in between these two) no
Line 2
lengthened syllable
2 Line 3
la
Line 4
la
lengthened syllable

Clapstick pattern 2

2 3 Line 6
lengthened syllable
Line 7
lengthened syllable + la
4 Line 8
tied-over lengthened syllable + la
Line-9
lengthened syllable
Clapstick pattern 3

Figure 5: Line 1.1 rhythmic cells

Verse 1 Line 1

rh[??] [??][??]

mrkinj-bardi- kadan nokenda-bonj

ge2sg.FUT-liquid-fetch.NP SWLL.PROX-EMP

rh[crochet] [crochet]

mrkinj- rdam

ge2sg.FUT-put.NP