The most surprising occurrence, however, in recent years concerning the fate of the Beothuk Indians was the accidental discovery of an old Indian woman named Santu, who claimed that her father was one of the last survivors of the Red Indians of Newfoundland. Since considerable discussion was aroused over the innocent claim of the old woman when I had made it public, I shall give the circumstances in some detail, for the benefit of those who may wish to determine to what extent her testimony may be relied on, before making use of the information and the brief vocabulary obtained from her.

Mr. James P. Howley, Director of the Geological Survey of Newfoundland, who for more than forty years has been interested in the history of the Beothuk, during a visit I made him at St. Johns [ sic ] in 1914, expressed his unbelief in Santu’s veracity. ⁴⁸ Notwithstanding the fact that Mr. Howley’s opinions, based on his extensive knowledge of Newfoundland history and physiography, deserve serious consideration, I hardly think, under the circumstances, that the conclusions of one trained in sciences other than ethnology are sufficient to warrant absolutely casting aside information which may be of value, and which on the face of it does bear some semblance of truthfulness.

In July, 1910, I happened to talk over ethnological matters with a family of Micmac who were temporarily camped near Gloucester, Mass. The family consisted of an aged woman, her son, his wife and child (plates xxxi-xxxvi). They all spoke Micmac. The family name was Toney. On inquiring of the young man, Joe Toney, where he was born, he told me in Newfoundland. Then becoming more interested, I inquired if his mother was a native of Newfoundland, he replied that she was. After a few minutes talk with his mother, he said that she was not a true Micmac, but that her father was an Osa’ ( ana Indian from Red Pond, Newfoundland. This naturally startled me, because it referred indirectly to the supposedly extinct Beothuk. Further conversation with the young man, who translated my questions to his mother, disclosed the fact that she was endeavoring to explain to me that, while her mother was a Micmac woman, her father was a member of the tribe which had been exterminated in the island by white men. There was at this time in her statements no idea of boasting, nor of gaining money or favor. She did not claim to know any words of her father’s language, but declared her willingness, if I would give her time, to try to recall some. On one thing she was definite at the very first: that her father claimed that he had been stained red when he was a baby among his own people, and that his people were very crude and were persecuted by the English. He had, it seems, been taken by the Micmac when he was young, reared by them, and converted to Christianity. As for the rest, suffice to say that I spent time when possible during the rest of the summer in following the family about from one summer resort to another, encouraging the old lady, through her son, to endeavor to recall all that she could of what she had heard her father narrate of his early life and people. The old woman was very difficult to work with; because of growing senility she was unable to concentrate her attention on any one thing for a sufficient length of time really to accomplish anything. Petty family troubles and present ills consumed her interest. And so by eking out reminiscences of that period of her life when she lived with her father in Newfoundland, I tediously gathered the information that follows. In September I lost traces of the family, which, I learned later had moved to Attleboro, Mass. ⁴⁹ Knowing the old women’s manner and the circumstances, I am convinced that she was not intentionally fabricating a story. My only distrust of the material she was able to give lies in the accuracy of her memory, especially in regard to her vocabulary.

Santu was born in Newfoundland near "Red Pond" (Red Indian Lake), about seventy-five years ago (dating from 1912). Her father, "Kop" (name of a red root found in the lake, according to her vocabulary), ⁵⁰ was a full-blooded native of a tribe which called itself Osa’ ( ana . With her father she left Newfoundland at about the age of ten, or a little less, and removed to Nova Scotia, where she passed her early womanhood. Her mother was a Micmac woman, one of the band who lived in Newfoundland. She died, it seems, when Santu was quite young. When Santu grew up, she married a Mohawk and spent part of her time in New Brunswick and Nova Scotia, and part in roaming about in the neighborhood of the Great Lakes with her Mohawk husband until the Civil War broke out, when, to escape being drafted, he led her wandering again throughout the northeastern states and eastern Canada. Her husband then died. Santu returned to Nova Scotia and married a Micmac chief near Yarmouth, whose name was Toney. Living there a while, she had four or five children, and finally, with her youngest son, separated from her husband and since then has been drifting about the New England states with him, earning an uncertain living by basket-making, bead-working, and fortune-telling. Her one son, Joe Toney, still lives with her. He has married a Micmac woman of Nova Scotia and they have one child (1912).

Santu remembers in her childhood having traveled with her father in the skin canoes which seem to have been one of the types of craft in use by the Osa’ ( ana. ⁵¹ While the details of construction given by Santu were very vague, it seems that the canoe was more of a kayak. It was about fifteen feet in length and about two and a half in width, constructed on a wooden framework with a caribou- or seal-skin covering sewed with water-tight seams. The seams were sewed by laying the two edges together, bending them over and sewing the three thicknesses together. Bone awls, she said, were used to perforate the holes for the stitches. The bow of the canoe, she remarked, was straightened and stiffened by a piece of spruce-bark [ sic ], ⁵² and another curved piece held the stern in shape. The bottom was found. At the back sat the man with his paddle. The whole front of the craft was covered with the skin, forming an enclosure large enough to contain the whole family, including women, children, dogs, and property. At his side and in front of him the man had his harpoon and other necessaries fastened on the side of the deck. It is to be understood from this description that a covered kayak-like type of boat is described. The skin-covering of the canoe was so arranged that it could be wrapped around the waist of the man so that no water could come into the hold in rough weather. In this craft the family traveled all over the country by waterways and coast, day and night. When a landing and camp were to be made the cover would be taken off the canoe, poles cut for a wigwam, and a temporary camp made until it was time to move on. Santu herself remembers being bundled in with dogs and members of her family, and traveling by night and day with her father.

The people, she claimed, subsisted largely on sea mammals’ flesh and caribou, using the harpoon for killing the former and the bow and arrow for the latter. When an animal was killed with an arrow, the arrow was never used again, but thrown away as a kind of sacrifice.

Flesh to be eaten was thrown on the fire and only partly roasted. Her father, she remembers, would eat little or no vegetal food nor bread. His diet consisted mostly of half-roasted meat.

A certain species of leaves was smoked in stone pipes. ⁵³

Allowance should be made for the probability that in some of these descriptions the old woman’s memory was so hazy that she could not distinguish between what she intended to claim as applying to the customs of her father’s people and those of the Micmac-Montagnais among whom they lived.

The most interesting information is that describing an annual ceremony participated in by the tribe at "Red Pond." It took place in the spring of the year when the tribe gathered and enjoyed, to use Santu’s phrase, "a big time." Games were played, among them the dice-and-bowl game in two forms. One of these was with seven dice discs and a bowl, ⁵⁴ and seventeen counters — four square ones and a crooked one called the "chief." The other form of the game was played with one large die, about two inches across, and six small ones, which were thrown upon a blanket or a hide and struck sidewise with the hand. ⁵⁵ Men only played the latter. The Micmac and other eastern tribes, she claimed, learned this game originally from her people. It is worthy of note that this game does not occur among the Wabanaki west of the Micmac. Dancing and feasting accompanied the event. At a certain time the men procured quantities of a kind of red root from the lake and squeezed from it the juice which was used for staining their bodies red. The ceremony is said to have lasted about ten days. Every person in the tribe was dyed. Children who were born during the year away in the hunting territories were brought to this ceremony for the first time and received their coat of dye, which was to last them for the year. It is supposed that under certain conditions the dye could be renewed, though the application of the coloring was regarded as a kind of initiation and mark of tribal identity. One good application is said to have lasted six months. Santu’s father, she claimed, was the last child to have been treated in this way. When he grew up he was converted to Catholicism and gave up his belief in the necessity of the red dye. If anyone was observed by the child to have some of the coloring washed from any part of his body, he was ordered to go to water and wash off his dye as a punishment, and not to renew it until the next ceremony.

Santu heard the tradition from her father that in his grandfather’s time [?] a ship was wrecked off the coast of Newfoundland and all hands were drowned except two women who, with the help of the natives, were brought ashore. One of them shortly afterward died; the other remained with the tribe, married one of the men, and spent her life there. Her father thought that he was descended from this woman.

Several opinions expressed by Santu regarding her father’s people may be of value. One was in reply to a direct question as to whether her father’s people were of mixed Eskimo and Indian blood. Friendly relations, she said, were maintained with the Labrador Eskimo and Indians. Some of her father’s people, she said, when dispersed, joined them. She remembers, while living in Nova Scotia, a paternal uncle or great-uncle returning from Greenland where he had emigrated and intermarried with the Eskimo there, and that there were a number of children. He died there within six months after coming to Nova Scotia. Santu stated that she had a relative (I fail to recall whether it was a cousin or a brother) somewhere who knew a great deal of the Osa’ ( ana language.

The Micmac, she said, came to Newfoundland a long time ago and for a while, with the white people, fought her people. Afterward a number inter-married with the Osa’ ( ana some of the descendants of the latter being still scattered here and there among the Micmac of Newfoundland and elsewhere.

There seems little doubt from Santu’s statements that Osa’ ( ana descendants may be found in the maritime provinces and that the tribal name itself is one of the name terms for the tribe known in history as the Beothuk.

Santu, with great difficulty during the summer, remembered the following words in her father’s language:

be r n fi m, woman (Micmac and Malecite épit , Penobscot p { he r n ’ m ).

gu r wa, fat person (Micmac me r gigit ).

gau, rain (Micmac gi 0 kpesa {r ).

h fi ^g , baby cradle, or cradle-board.

tu 0 ^b , baby blanket (Micmac wobi 0 r sun ).

serko, prayer (Micmac alasu 0 r dma ).

si 0 kane r 0 su, whale (Micmac po r d ’ p , Penobscot - es 0 u , "living creature," noun ending in animal names).

NOTE: ^g , ^b , weakly articulated final consonants

Her father’s people, Santu alleged, used their hands a great deal in conversation. The only word in the above list in which any resemblance can be recognized as occurring in any of the published Beothuk lists is the term, be r n ‘ m "woman." Compare emam-(emamoose), "woman" (Peyton vocabulary; Lloyd in Journal of the Anthropological Institute of Great Britain and Ireland, 1875) and enam, "woman," given by Patterson in Transactions of the Royal Society of Canada, vol. x.

Among other reminiscences I add the following song [see 3. above], transcribed by Mr. J.D. Sapir from a phonograph record made by Santu while she was camped at Hampton Beach, N.H., in 1910. It was a rendition of a song that she had learned from her father when she was a girl. She claimed that her father told her that it was an Osa’ ( ana song.

The syllables were too inarticulate to be taken down at the time, I am sorry to say. Santu stated that she was unable to explain them, because they had no sequence of meaning to her.

Again during my trip in Newfoundland I inquired of several elderly Indians about the woman Santu. John Paul, already mentioned, knew of a woman of Santu’s description who had gone to Nova Scotia and was there the wife of a wealthy Micmac chief named Toney. He furthermore, much to my inward surprise, credited the claim that her father had been a man of Red Indian blood. He stated that the thing was not only possible, but that it might well be expected to be true, considering the sedentary habits of many of the Micmac hunters and the secretiveness of the Indians concerning the Red Indians a generation or so ago through fear of retaliation or at least molestation at the hands of the English, since such a stir had been raised over them. From Micmac in Newfoundland I even learned of another man, George McCloud [McLeod?], whom no one could locate at the time. He was said to have knowledge not only of the Red Indian language, but also of where descendants could still be found in Labrador.

If, despite the meagerness of our actual knowledge of the tribe, any conclusions are at all permissible, I believe the indications will increasingly show that the Beothuk formed an archaic member of the culture group which embraced the Micmac and other northeastern Algonkian. This is a strong corroboration of the evidence of linguistic relationship with the Algonkian. As for the likelihood of Eskimo relationship, the links of union, either archeological or otherwise, are not a bit stronger than between the Eskimo and the Montagnais. The next thing to be done in this field, aside from systematic archeological research, is to collect a sufficient quantity of mythological material from the Newfoundland Indians for comparison with that of the Micmac of the mainland in order to determine, if possible, traces of what might be considered Beothuk influence.

In the 1970s, when preparing materials for The Beothuk Vocabularies, which was published at the end of the decade (Hewson 1978), I was searching on both sides of the Atlantic for original materials concerning the Beothuk Indians. From the American Philosophical Society in Philadelphia I received copies of materials deposited by Frank Speck relative to his visit to Newfoundland recorded in his book, Beothuk and Micmac (Speck 1922). There were newspaper clippings and correspondence with a variety of people, including a couple of letters from James P. Howley, author of The Beothuks (1916), and Speck’s handwritten notes taken down in July 1910 from an old lady camped out near Gloucester, Massachusetts, who claimed to be descended from a Beothuk father and a Micmac mother.

Speck also mentions (1922, 67) a song, that he recorded on "a phonograph record," a song that Santu sang and claimed she had learned as a child from her Beothuk father. It never occurred to me that the wax cylinder on which this song was recorded would also have been deposited by Speck along with his other papers at the American Philosophical Society offices. However, Victor DuPree, who at that time was the Director of Language Laboratories at Memorial University, saw the opportunity. He had earlier spent a considerable amount of time trying to track down any trace of the little Beothuk girl Oubee who had been taken back to England in the last years of the eighteenth century by Thomas Stone of Winterbourne Minster in Dorset, but regrettably without any success. He now wrote to the American Philosophical Society inquiring about the wax cylinder, if it still existed, and if so, would it be possible to get a copy. The reply was affirmative and a copy was procured and stored in the language laboratory at Memorial.

I listened to the tape several times and did not try to decipher any speech sounds because a lot of North American Aboriginal songs have no words, and the recording was so scratchy and indistinct that words, if any, were not obvious. Speck in fact states (1922, 67): "The syllables were too inarticulate to be taken down at the time, I am sorry to say. Santu stated that she was unable to explain them, because they had no sequence of meaning for her."

I also could not make sense of the transcription of the music that J.D. Sapir, son of Edward Sapir (one of the most famous linguists of the twentieth century, who worked early in his career for the National Museum in Ottawa) had transcribed for Speck (1922, 68). Some years later, however, a St. John’s music teacher and pianist, Mary Muckle, expressed interest in the song because she was recording lullabies from a considerable variety of languages, and she did her own transcription of the music on the recording. Looking at this manuscript transcription, I saw at once why I could make little sense of the published transcription of Sapir: the first line of three or four bars was missing from the version published in Speck’s book (1922, 68).

Given that I could now relate Mary Muckle’s transcription to what I was hearing on the tape, I realized that I had given up too easily; a little more perseverance and I would have discovered for myself why I could not make sense of Sapir’s transcription. This gave birth to the thought that I had also taken Speck’s word too seriously, that "The syllables were too inarticulate to be taken down," and that I had consequently given up too easily on making a phonetic transcription of the sounds made by the voice, which might indeed be speech sounds. I returned to the tape, began to distinguish syllables, and to make a draft transcription of what were, in fact, quite obviously speech sounds. It took me two weeks of arduous repetitive listening to produce a transcription that I would consider, given the problems of the quality of the recording, to be a reasonable phonetic interpretation and representation of what I was hearing.

Soon after, while working with Bernard Francis, a native speaker of Micmac, in Sydney, Nova Scotia, I met Franziska von Rosen, a musician interested in Native American music, who sent a copy of the song to Beverley Diamond (then) in the Music Department at Queen’s University in Kingston, Ontario. She did a further transcription and supplied a half page of notes on the style of the song, as Native music. At this time, in the early 1990s, the CBC was also organizing a series of tele-casts on the history of Newfoundland, and it was thought that the segment on the Beothuks might include a recording of the Beothuk song. It was decided that my daughter Jean, a well-known traditional singer, would attempt to record a performance of the song, which she did using the new phonetic transcription and her own transcription of the music.

The recording was made, but the telecast on the Beothuks had too much material and it was not included. It was broadcast, however, and the CBC has rebroadcast it on several occasions. It is now accessible on the internet.[2] The CBC also produced a "cleaned-up" digital recording of Santu’s 1910 performance, which removed a lot of the ambient noise created by the wax disc player.

The deterioration of the audio recording renders a totally accurate transcription virtually impossible. By working from a recording that had a lot of the noise of the wax cylinder recording filtered out, however, and by using digital editing and time-stretching tools, we were able to transcribe much of the text reliably and to determine the range of variability for other syllables or words. In particular, we were able to hear word and line repetitions that had been hitherto elusive. The very filtering that reduced noise, however, also exacerbated the ability to hear certain consonants. The most intransigent problems were as follows:

• bilabials "p" and "m" are hard to distinguish.

• when aspiration occurs, the consonant is obscured; hence, "t" may be "h" or "s". The final word in the second last line, then, might be "halape," "halame," talape," or "talame."

• consonants preceding a long "i" are harder to distinguish than syllables with other consonants. The first word, then, might be "siupe" or tiupe."

• final "l" sounds are virtually inaudible. We think we hear them in m. 17, 18, and 21.

• word divisions are hard to determine except where there is word repetition. Because part three is the most repetitive, the word divisions are the clearest in this section.

• the slurred melisma in m. 15 is particularly unclear. We have transcribed a single vowel sound here but there may be vowel changes that are simply indecipherable.

Before considering the content of her performance, it is important to describe the technology of the day. Speck recorded her on a wax cylinder, itself a fragile format that deteriorated severely with every replaying. Hewson has described the process by which it was copied with a cassette copy eventually reaching MUNFLA. Cassettes also deteriorate markedly over time and so, by the time we were able to digitize this audio artifact, much audio data was irrevocably lost and the noise of the recording process itself gives us a heavily masked image of her voice. Using digital technology, we can cut some of the noise but we may simultaneously be changing her voice.[9] I worked with three different restorations to attempt transcription. The consonants, in particular, which are hard to distinguish, were the main points of difference between my initial transcription and that of Hewson. For example, he wrote the second word as "tanitulati," where I heard "kamiskunapit" or perhaps "ka mistanapeo" (Innu for big man, spirit helper, or grandfather). The transcription notes (above) describe other problems.

It is also important to note that the recording technology of the day sometimes led scholars to ask performers to sing without accompaniment, so that the voice could be heard more clearly. A well-known instance of this was Frances Densmore who advised that "an Indian drum does not record well, and a rattle does not record at all. A short stick on a pasteboard box gives the percussion without resonance, which is all that is wanted unless the records are for exhibition use ..." (cited in Gray 1988, ix). A stick tapping either birchbark or a small wooden, hand-held saucer, on the other hand, is a traditional Mi’kmaq/Malecite accompaniment.[10] It is impossible to ascertain whether Santu might have used such accompaniment had she been allowed to.

Ethnography is central to the work of contemporary ethnomusicologists. Hence, the challenge of trying to make some sense of a single text takes me outside my usual methods and frames of reference. I have not worked as a textualist in recent years and when I did,[11] I relied on methods that I no longer regard as adequate. Unsettled then, I had to ask hard questions about what the internal evidence of a song actually could mean, and what frames of reference were relevant for determining the history of oral traditions. My approach is necessarily that of an outsider, but with some knowledge of both archival and contemporary recordings of First Nations singers in Atlantic Canada.

The question I wanted to answer of this song was: What might we assume about Santu’s agency in selecting what she would record for Speck? Is this unanswerable or were there clues?

Initially, I queried what the song itself could tell us if we listened/looked carefully at its pitches and rhythms, its repetition patterns, and text. I questioned why we should assume that this was a "piece" rather than a bit of play or a process of communication. Why should we assume that it was singular — a unified whole — rather than a series of fragments? Santu claimed that she learned "it" from her father, but what was that "it"? A capacity for singing, a corpus of songs, or a single song? I look for instance at whether there are clear points of closure within the performance, or whether there are differentiated styles for different segments of the performance.

Speck (1922, 66-67) tells us that the words were likely vocables since she was "unable to explain because they had no sequence of meaning to her" and were "too inarticulate to be taken down." As he describes earlier, Hewson tried with little success to relate the words to those in several fragmentary Beothuk word lists.

The pitches and rhythms of the tune, on the other hand, are discernible.[12] Speck published a transcription (see Figure 1) by Sapir that does not start at the beginning, as if Sapir thought the first phrase was extraneous. He seems only to have "heard" the song at the point where the metric structure clarifies. Since, at that point, there are two measures of four, he writes the song in 4/4 metre, even though the accent structure and motivic repetition of the rest of the song defies this consistency. In the first two lines of my transcription, I have indicated a metric grouping that reflects shifting accents and repetition patterns. The middle part of the song (lines 3-5 of my transcription), however, sounds "freer" to me, as if the delivery of the text rather than a steady pulse governed the presentation. The durations are still transcribable into quarter and half notes, but the underlying pulsation is not easily grouped as it was in part one. The third part of the song changes rhythmic structure again, now having a clear and consistent triple pattern. These very different rhythmic qualities lead me to suspect that Santu performed, not one, but perhaps three different songs, or possibly only song fragments. Further analysis supported this hypothesis.

The scales and formal structures[13] of the three parts of the song differ. Part one (ending at the first double barline) uses six notes, transcribed here as EG¹A¹B¹C¹D¹.[14] While I considered that this very short section might be a song fragment, or an uncertain beginning, it does seem to have some elements of repetition that suggest a discrete structure. Phrases one and three are parallel (the first two words and the melodic motives to which they are sung are mirrored at the end). That is "siapes kamiskunapit" sung initially to motives marked "a" and "b" is repeated "kamiskunapit siapes" sung to motives "b" and "a" at the end of "part one." The mirror relationship is coincidentally like the double curve design that features so prominently in Mi’kmaq design. Analogies between visual patterns and aural ones are arbitrary but such relationships are sometimes noted by First Nations singers. Hence, I offer the observation without putting too much weight on it. There are other instances of an initial word mirrored at the end of Mi’kmaq and Wabanaki songs, in archival audio collections.[15] These few examples, however, are slight support for an identification of part one as Mi’kmaq "in feel," and certainly no specific genre can be identified.

The second part, on the other hand, bears a close resemblance to a subset of the rhythmically free Innu hymns (often based either on French cantiques that are named in Innu language hymnbooks, or on Latin liturgical music) that have been recorded in Labrador. Santu may have encountered such repertoire in the dealings with Labrador Indians that she described to Speck. Indeed, among recordings of Innu hymns that I made in the 1980s, there are many with similar opening formulas — an initial leap up to an elongated note (see Figure 4 below), followed by stepwise motion as the performer declaims a phrase of text, and a cadential formula. Furthermore, the cadences ending each phrase in Santu’s part 2 have both rhythmic and pitch similarities vis-à-vis these Innu hymns. Some parallels are noted in Figure 4. Unlike part 1, the scale of this section, with a range of a minor seventh, is distinctly modal, becoming clearly Aeolian, especially after the first phrase.[16] While these are details that might also be found in many other repertoires,[17] their configuration and consistent use in the aforementioned group of Innu hymns is noteworthy.

The third part is in a faster tempo with a steadier triple metre, as noted above, giving this section a dance-like quality that is distinctly different from the earlier parts of the song. The steady rocking rhythm may have been what reminded John Hewson of a lullaby. When I first heard this, I noted that the first phrase sounds a bit like "London Bridge is Falling Down." These associations indicate how we automatically map repertoire from our own experience on to new and unfamiliar songs that we encounter. The fact that this part of the song uses a familiar major scale (albeit spanning an octave from the sixth degree of the scale), may make it still easier for Euroamerican associations to come to mind. The final text phrase of this third part repeats, although the two melodic phrases differ. This gives a distinct sense of closure to this section of the performance. Once again, however, I have chosen to explore whether this part of Santu’s song might be related to a variety of repertoires that she might possibly have encountered, acknowledging that the intuitive responses of both Hewson and myself are not to be discounted since European children’s songs or lullabies might well have been part of Santu’s cosmopolitan experience. This is the only section of her song with repeated syllables ("lu lu lu ...") that are almost surely vocables rather than lexically meaningful text, as noted above. Such syllables are common in both English-language and Innu lullabies, the latter called bébé ataushu, for instance. (On the other hand, vo-cables might indicate that she simply forgot the words in this one phrase.) My search among other Wabanaki or Innu repertoires revealed that Malecite "wedding dances" recorded by Mechling in 1911 have phrases in triple metre, rare among other genres of Algonquian music, but there is nothing else to connect this part of Santu’s song to the Malecite genre. I am aware of a single Innu hymn that switches from a free rhythm section to a triple metre section, but one example is hardly evidence. So the question of possible links to specific genres remains open.

An additional question is whether other Wabanaki songs ever had multiple sections with different melodic and rhythmic characteristics. By the 1950s, when Nicholas Smith recorded Wabanaki (mostly Penobscot) songs, there were, indeed, a number of songs with two distinct sections. But other features such as consistent triple or more often quadruple metres and parallel phrases with open and closed cadences contrast quite markedly with the style of Santu’s song.

The internal evidence that is recoverable by analyzing this performance closely, then, suggests that Santu may have sung three songs for Speck rather than one. Double bar lines indicate the end of each of the three sections in my transcription. The notation tries to reflect some of the differences between the three sections of the song, by using no bar lines in the middle section, for instance. Was it usual for singers to run songs together in this way? While it is hard to know, we might again note that Frances Densmore’s instructions for collectors includes the following: "Singers should not be allowed to ‘run their songs together’" (Gray 1988, ix).

The nature of my analysis is clearly Eurocentred, focused on pitch- and rhythm-based stylistic features, and these may not be the most relevant song elements to explore at all. Indeed, most First Nations musicians with whom I have worked speak about genre in very different terms: relating to dance steps, to places of origin for specific repertoire, to individual stories that generated specific songs, or — and this is the marker that I thought might be useful for Santu’s song — they point to distinctive phrases of vocables that mark specific types of music. Hence, the phrase "Gainawiyo heja" identifies the eskanye (women’s shuffle dance) of the Haudenosaunee, or the final phrase of vocables "he ye no we" ends a "peyote" song of the Native American church. But as yet, I have not found any vocable phrases in Mi’kmaq archival sources that help identify any parts of this performance.

Raising even more questions than the criteria I used to analyze the song itself are the historical connections that I offered. Here the enormous question of how and why we might try to reconstruct histories of oral traditions and oral performances is evoked. While all musical traditions change and develop over time, the sorts of style changes that we associate with written traditions (classical European music, for instance) and with societies where innovation has been highly valorized are not usually appropriate expectations for oral traditions. Hence, while composers from Brahms to Reich might have changed the whole language of classical music in the course of a century, oral traditions maintained during the same time period may have added some repertoire from new sources, changed performance practice, and slowly altered versions of well-known songs, but the change happens differently, and arguably more slowly. Hence, we have dozens of variants of "traditional" English ballads or Irish dance tunes, variants that migrated across national borders and instrumental techniques, variants that were localized or exoticized, variants that responded to the availability of musical instruments or new media. In the currently maintained English- and French-language oral traditions of Newfoundland and Labrador, for instance, some repertoire can be traced to the sixteenth or seventeeth century, much more to broadsides of the eighteenth and nineteenth centuries to American sheet music of the early twentieth century, or to local composers who draw on all of the above styles when they made up new songs.

There has, thus far, been surprisingly little work done on specific histories of Native American repertoires, but a few things about transmission processes are evident. Like other oral traditions such as the ballads or dance tunes alluded to, variants migrated and transformed in relation to a variety of factors but specific tunes, or shorter motifs, contours, or rhythmic patterns are sometimes recognizable across these variants. For this reason, I do not think it is advisable to regard Santu’s statement to Speck that she learned her songs from her father as an indication that her repertoire originated at a specific time or in a specific place. He maintained his repertoire in the last half of the nineteenth century but it is likely that some of it predated his lifetime while some might have been quite "modern." There is an argument to be made, then, that sources from divergent time periods can justifiably be compared when trying to determine histories of orally transmitted songs. On the other hand, I would be remiss if I were to suggest that the historical evidence is either clear or adequate. The connections I make are well considered but ultimately speculative.

First Nations repertoires that were associated with ceremony rather than social occasions have often been resistant to change since "errors" in performance had severe consequences. Even Christian hymns in Aboriginal languages, especially those that were taught using music notation, changed slowly (as Tom Gordon’s paper in this volume demonstrates). Some hymns changed more than others. The Innu language version of "Veni Creator Spiritus" is melodically and rhythmically like the Latin hymn. The contemporary Innu version of the "Huron Carol," on the other hand, differs substantially from the tune that has been frequently reprinted.[18] The hymn form that I reference in relation to part two of Santu’s song is not quite like either of these but is rather a formula-like mode of delivering a text. The initial leap up to rest on a long note, the subsequent utterance of the text in a conjunct melodic pattern that curves in and around the long note’s pitch, and the use of conventional phrase ending patterns can be adapted to many different texts.

There is also evidence that certain phrases from older repertoire are now incorporated in contemporary social dance music, sometimes moving to a new "genre" in the process.[19] This transmission process is also found in many other oral traditions. In such cases, then, the history of a specific genre is cross-cut by the collective memory of musical fragments that are reused in new contexts.

A significant dimension of song histories for Native Americans is often in the stories that accompany songs. The gifts of songs from other nations are carefully noted. Often the individuals who brought a certain form of traditional knowledge are remembered. In some traditions — the Inuit drum song tradition, for instance — composers are carefully acknowledged and repertoires can be dated relatively accurately. In others, where songs are "received" in dreams, even the concept of a "composer" is not relevant. Santu acknowledges her father as the source of her songs but further detail about the acquisition of her repertoire is lost.

In sum, then, many questions remain about the historical roots or generic associations of the song. Its three contrasting styles, however, seem quite clear. I do think it likely that Santu was trying to demonstrate to Speck her eclectic repertoire, partly derived from her own people and partly from those she encountered on her travels. The concept of recording one’s voice, making it available to others, must have been such a powerful experience. It seems likely to me that the choice of what you would sing in such a momentous circumstance would not be arbitrary. Santu is perhaps one of the first indigenous singers to use audio recording not to represent a static culture but to bridge cultures, as she did with her baskets, beadwork, and fortune-telling. She saw cultural interaction as dynamic and was clearly proud of her hybrid heritage.

After Speck recorded Santu in the company of her son and granddaughter in Connecticut, probably at their camp along the Penobscot River, he cast and recast information about her for different purposes. First, consider his field notes: initially, fragmentary words and phrases, mostly Beothuk nouns for people (Osaganna, described here as "own name"), Santu (old ladies name — not "Old Lady’s name"), Kop ("her grandfather’s name") or "alasudma" (Micmac), and things he probably saw around their camp (blanket, cradle, gun). We see hints about the weather (rain), activities ("leaves smoked," "pray, made cross"). The fragments of vocabulary suggest that he solicited Beothuk words perhaps by pointing to objects or people and recording names. There is little sense of coherent conversation conveyed here. Later in the notebook (right page) we begin to get much fuller explanations, as if the communication improved. She describes the spring red dye ceremony and dice games (perhaps related to Mi’kmaq waltes games) that were part of it.

The rest of his notebook entry is more literary. He writes an entire paragraph about canoe construction, a sentence about sacrificing arrows used to kill an animal, another returning to the red ochre ceremony. These passages found their way into the section of his published book that is replicated at the beginning of this article. What does the fluent, more literary style tell us? Were these passages written later? Or did they perhaps indicate that Santu and probably her adult son started, at this point, to speak English to Speck?

His published account of these same topics suggests that he relied a great deal on his memory for additional detail. I wonder, for instance, how he concludes that "Osaganna" — "own name" in his notebook (Figures 5 and 6) — was a localized name for the Beothuk people rather than a family name, clan name, or name for a regional group. A second context for Speck was the news media. Undoubtedly with newspapers or the popular press in mind, he took photographs of Santu and her family in their camp. The cropped head and shoulders reproductions in a newspaper story about the family show little of the context of their lives.[20] The headline (Figure 7) exaggerates her age (saying she was almost 100, when she was probably about 60, judging by the ages of her son and granddaughter, or in her seventies if Speck’s own estimate of her birthday is accurate). He casts their relationship with the Mi’kmaq as adversarial ("living among those they abhorred"), ignoring that they spoke Mi’kmaq and had equally strong Mi’kmaq as Beothuk roots.

A third context for Speck is the academic one, and here his correspondence with colleagues indicates that he expected scholarship would be argumentative and adversarial. At this stage of his career, he was confronted by James P. Howley, Director of the Geological Survey of Newfoundland, who from his office in St. John’s had compiled all the data on the Beothuk that would constitute his book of 1915. Speck and Howley corresponded in 1911 and 1912. Howley challenged the basis on which Speck determined Santu’s Beothuk lineage, particularly because those names (including Santu) and vocabulary items noted in Speck’s fieldnotes "didn’t sound Beothuk." Howley neither references Santu nor includes Speck in the bibliography of his book, published four years later (1915).

But while Howley was ready to discount her credibility because she was apparently hybrid, Speck too was clearly devoted to thinking of the Beothuk as a separate and distinct people, noting "the Beothuk was an independent stock and tribe. They were not related to any of the other Indians about them, either in language or in race." In the face of the mixed parentage of Santu herself as well as her son and granddaughter, the statement indicates a determined effort not to look at hybridity, encounter, and the complexity of lived experience. In 1917, Mechling (with whom Speck had a rather cordial relationship, judging by their correspondence) comments that "your table of similarities between the Eastern Algonquians and Beothuk seemed to me to be a very false sampling as there were many points of similarity as well as difference not included in the table."[21]

In conclusion, it is essential to emphasize that all representations of texts, including musical texts, are deeply connected to what we expect to find. Speck, like Howley, sought evidence of "authentic Beothuk" culture and had no interest in looking at the internal consistency of this song, or even ensuring that his transcription was complete. My approach was to search for traces of encounter, but my tools are constrained by Eurocentred notational practices and style analysis. Our different approaches illustrate what studies recently published in an anthology by Brown and Vibert (1996) demonstrate: how both the uses of texts and the ever-changing contexts that surround their creation, re-creation, and representation are essential components of historical study. They quote Johannes Fabian’s argument that "in acts that produce ethnographic knowledge, creations of text and creations of context are of the same kind" (xx). And they play with the image of the window, suggesting that texts are not simply open windows on the past but "textured panes of glass, framing outlines of distant shapes, subtle hints of movement. If we approach the texts alert to their refractions as well as to our own angles of vision, there is much to learn" (xx). In the case of Santu’s song, those refractions have included the mediations of both audio recording and print. The angles of vision have been shaped by different academic expectations and different sociocultural experiences. New means of contextualizing her life story, in relation to ongoing debates about the hybrid and complex histories of Aboriginal people in Newfoundland, in relation to the biographical recovery of the lives of Eastern Woodlands women as cultural mediators, in relation to different disciplinary constructs and post-colonial ways of reading texts change the image of Santu Toney and, hopefully, help give her back her voice.