The Breckenridge Flute

The Ozark Plateau, located in the present-day Southeastern United States, was home to Native Americans from as early as 10,000 BCE until the early 1800s. They inhabited rock shelters on bluffs along rivers of the Plateau, leaving behind rock art and artifacts. Because of the arid, protected conditions of the bluff shelters, many artifacts from these sites are unusually well-preserved ([Blouin 2011] ).

Nine artifacts from excavations done in 1932 were recently identified as musical instruments by James A. Rees, Jr. ([Rees 2011]). They come from five bluff shelters in Northwestern Arkansas and Southwestern Missouri and include one partially-intact duct flute of river cane from Breckenridge Bluff in Northwestern Arkansas. This artifact is known colloquially as “The Breckenridge Flute” and has been radiocarbon dated to 1020–1160 CE.

Acknowledgements

All the artifacts described on this page are in the University of Arkansas, Fayetteville, Museum Collections, which supported all the archeological expeditions that retrieved them, and has given kind permission to publish information and images of the Breckenridge flute.

Thanks to James A. Rees, Jr. for initially bringing his research to my attention, providing the present-day photographs of the flute, and assisting with the development of this web page. Thanks to Deborah Sabo of the Arkansas Archeological Society and editor of their Field Notes Newsletter for making [Rees 2011] available to me.

Elevation map of the Ozark Plateau, showing the site of the Breckenridge Bluff Shelter

Please note and respect the copyrights for images on this page, which are held by the Arkansas Archeological Survey, James A. Rees, Jr., Leslie Walker, or otherwise, as noted.

General background references for this page are [Anderson-DG 2002] and [Mainfort 2008].

The Site

Site 3CR2, known as the Breckenridge Bluff Shelter, is located on the shore of Beaver Lake in Carroll County, Northwestern Arkansas.

The map at the right shows the four sections of the Ozark Plateau and the location of the Breckenridge Bluff Shelter excavation site (formerly known as the “Pine Hollow Shelter”). This image is based on an elevation relief map of the Ozark Plateau provided by Wikimedia user Tosborn on January 2, 2007.

The site was excavated in 1922 by the Museum of the American Indian and in 1932-1933 and 1960-1962 by the University of Arkansas Museum ([Crane 1968], pages 88–89).

The 1930s Excavations

The excavation that yielded the Breckenridge flute was led by Samuel C. Dellinger (1892–1973). Dellinger was professor and chairman of the Zoology department at the University of Arkansas and curator of the University of Arkansas Museum from 1925 to 1960. From [Blouin 2011] :

Dellinger was an archeological pioneer in a day when almost no professional standards existed in the fledgling field. Most excavations were result-oriented and put little to no emphasis on the placement or preservation of the location of artifacts. But, although Dellinger had no formal archeological training, he kept careful notes on his own excavations, which occurred in the Ozarks, northeast Arkansas, the Arkansas River Valley and along the Ouachita River. … During Dellinger’s tenure at the University of Arkansas from the 1920s to the 1960s, he built up an incredible collection of pre-Columbian artifacts from the Mississippi River Valley and the Ozark Plateau, which remains today one of the premier collections of its kind in the country.

The Breckenridge flute was excavated by one of Sam Dellinger's crews in 1933. Here is a picture of the Breckenridge flute, Accession no. 32-2-344, in the field in 1933:

Field photograph of the Breckenridge flute in situ, 1933 ©1933, The University of Arkansas, Fayetteville, Museum Collections

Description of the Flute

The Breckenridge flute is a two-chamber duct flute made of river cane. (This section is based on information in [Rees 2011], and the direct quotes are from the same paper).

Although much of its blow chamber has broken off and the slide is missing, the Breckenridge flute has all the features [of a Native American flute]. It is 31.5 cm (12.4″) long and includes one complete section of cane and two partial ones. The complete section of cane forms the sound chamber of the flute. At the proximal end [head end] of the instrument are the remnants of the blow chamber [slow air chamber]. The blow chamber is separated from the sound chamber by an intact node of the cane culm. A groove or channel has been carved over the node on the exterior of the cane. At the proximal end of the channel is what appears to be the end of the blow chamber duct hole [SAC exit hole]. At the distal end of the channel [towards the foot end of the flute] is the sound chamber duct hole [sound hole], which is intact, but whose shape and size may have been distorted when the flute was damaged sometime after its manufacture. The distance between the edges of the two duct holes [i.e. the length of the flue] is 1.5 cm (0.59″).

Dorsal side of flute showing flue channel at proximal end, decoration zones, stop holes and direction hole at distal end:

Dorsal side of the Breckenridge flute photo by Leslie Walker, Arkansas Archeological Survey, courtesy of The University of Arkansas Museum Collections.

Closeup of proximal end showing flue channel, edge of proximal duct hole, distal duct hole, gouge marks on proximal node and decorations:

Proximal end of the Breckenridge flute photo by Leslie Walker, Arkansas Archeological Survey, courtesy of The University of Arkansas Museum Collections.

Closeup of distal end showing stops and direction hole, notice gouge marks on node:

Distal end of the Breckenridge flute photo by Leslie Walker, Arkansas Archeological Survey, courtesy of The University of Arkansas Museum Collections.

Ventral side of flute showing cracks:

Ventral side of the Breckenridge flute photo by Leslie Walker, Arkansas Archeological Survey, courtesy of The University of Arkansas Museum Collections.

Distal end of flute showing perforated septum:

Distal end of the Breckenridge Flute photo by Leslie Walker, Arkansas Archeological Survey, courtesy of The U. of Arkansas Museum Collections.

Rees notes that “an attempt was made to carve away and smooth the exterior ridge of the cane node separating the two chambers. This was probably done to facilitate the attachment of a slide to cover the duct”.

Barry Higgins of White Crow Flutes offered a perspective on using the apparent stains on an artifact to glean information on its construction and use. In a personal communication on January 9, 2012, with my editorial comments in [brackets]:

One of the thing we look closely at on old flutes are stains. These can be grease and dirt smudges which give us an indication of how the instrument was held and by which hand where, variations in wood color change due to differences in amount of UV exposure (plates/birds/wraps), and residual stains from glue/pitch/ leather tanning. In the flute pictured there is an obvious change in the coloration of the wood at the plug/TSH area which would indicated to me it was covered or wrapped and may even have been done with true sinew which stained the wood as well.

The sound chamber is intact and has a second river cane node near the foot end of the flute. A 1.8 cm (0.71″) diameter hole has been pierced through this lower river cane node. The distance between the two river cane nodes is 27.3 cm (10.75″). From [Rees 2011]:

There are four finger holes in the sound chamber. The first of these is located 12.4 cm (4.88″) from the distal edge of the sound chamber duct hole. The distance between each of the stop holes [finger holes] is approximately 4 cm (1.6″). All of these holes are surprisingly small with an average diameter of only 2 mm (0.08″). … The last section of the instrument beyond the distal node has one very small hole in line with the finger holes.

A counterpoint from Barry Higgins of White Crow Flutes (personal communication on January 9, 2012):

[Regarding] the postulation that the bottom hole is a playing hole: Most traditional makers regardless of tribe use a “formula” for hole spacing be it thumb width or any other method. One thing that seems to be quite obvious, regardless of tribal specificity and absence of precision measuring devices, is that they were generally uniform in spacing as the formula used was consistent. This flute seems outside of that observation. What it does suggest is a design possibility with elements similar to those of the Iroquois flute that had a plug in the hole closest to the foot which was more often disproportional in spacing (some longer some shorter). The consistent spatial relationship of the other holes makes me feel this is not a playing hole and more likely a tuning hole or Iroquois like plugged hole with plug missing.

Measurements

These measurement were provided by Jim on a sketch drawing that I've re-rendered, substituting some of his terminology for the standard terms used on this web site. Additional notes on the sketch drawing are provided below. Please note that Jim is using flute player numbering for the finger holes.

Measurements of the Breckenridge Flute

• The distance between the two septums is 27.3 cm The depth of the flue varies from 5mm to 2 mm at the distal (foot) end of the flute.
• The proximal (head) end of the sound hole begins 1 cm from the septum and is 1 cm long and 3 mm wide. It is located in the bottom of a groove that is 3.5 cm long at present.
• The distal end of what may be the SAC exit hole may be present at the broken end of the groove. The distance between the broken edge of the groove and the proximal end of the proximal hole is 1.5 cm.
• The diameter of the distal septum is 1.8 cm, with a 4mm hole in the center.

Dating of the Breckenridge Bluff Shelter

Artifacts and cultural materials have been excavated from the surface of the shelter to a depth of 3.5 m (11.5′). No datable organic remains have been recovered from the lower levels, but samples collected in a burial pit in 1932 by S. C. Dellinger 35 cm (13.8″) below the surface were dated to 1130±110 CE.

Three samples collected from charcoal in a hearth and from other shelter deposits in 1961 and 1962 yielded these dates ([Crane 1968] pages 88–89):

• 1810±100 CE at 20-40 cm (7.9″–15.7″) below the surface
• 1430±100 CE at 40-60 cm (15.7″–23.6″) below the surface
• 360±100 CE at 68-83 cm (26.8″–32.7″) below the surface

Rees originally conjectured that, based on the context of the artifacts shown in the 1933 in situ photograph shown above and on the detail page for the field photograph, the flute should date in the range 750–1350 CE ([Rees 2011], page 8).

This conjecture proved to be accurate when, in 2013, a sample from the artifact yielded a radiocarbon date range of 1020–1160 CE (95% probability calibrated date range, [Rees 2013]).

A Reproduction of the Artifact

Jim, assisted by Devin Pettigrew, an undergraduate assistant at the Arkansas Archeological Survey, crafted two reproductions of the Breckenridge flute. This section is based on and quoted from personal communication with Jim on December 13, 2011.

We tried to match all of the measurements as precisely as we could. However, most of the SAC is missing including all but the distal edge of the duct hole at that end of the windway or flue. Following tradition we made the SAC or blow chamber approximately as long as the width of a man's fist. We did this because the sound chamber also seemed to follow tradition being almost exactly the length of my arm from the elbow to the wrist. There was no evidence of what was used to cover the flue, but the outside ridge of the node of the cane that formed the barrier between the two chambers had been gouged off in an apparent effort to facilitate the attachment of some sort of cover over the flue.

I tried using my finger as a cover [block] with no success. In addition it would have been nearly impossible to cover all the stop holes [finger holes] with one hand while covering the flue with the other unless you had very large hands. After some experimentation I found that using a small piece of leather to cover the flue worked best when properly adjusted and tied.

Jim notes that the reproduction has a “pleasant, but slightly breathy voice”. Here are the approximate pitches, with the leather flue-cover positioned to produce the best tone quality:

Approximate pitches for the Breckenridge Flute Reproduction

Fingering	Approximate Note	Interval
	G5	major 6th
	F5	perfect 5th
	D5	major 3rd
	C5	major 2nd
	Bb4	root

This scale matches the pitches that a contemporary G minor Native American flute would produce with the fingerings:

Recordings of the Reproduction

Jim sent me one of the reproductions crafted by Devin Pettigrew and himself. I was able to photograph it and record this improvisation on January 13, 2012:

Improvisation

Clint Goss. Reproduction of the Breckenridge flute by Jim Rees.

Reproduction of the Breckenridge flute crafted by Devin Pettigrew and Jim Rees

Detail of the Pettigrew/Rees reproduction of the Breckenridge flute

Here are the pitches I measured on the reproduction, with the note and its bias in ± cents, including the cross-fingerings:

Measured pitches on the Breckenridge Flute Reproduction

Fingering	Measured pitch (A=440)
	Eb6 -20
	C#6 +0
	C6 +0
	G#5 +5
	G5 +5
	F#5 -20
	F5 +30
	F5 -40
	E5 +30
	D5 +50
	D5 +10
	C5 +30
	B4 -20

Comparison and Conjecture

The most likely candidates for comparison are the cane flutes of the Papago and Pima cultures. They used a windway and nest area crafted across a septum between segments of the river cane, used a finger or cloth wrap for a block, and many even had a second node near the foot end of the flute:

Yuma Transverse and Vertical Flutes from [Densmore 1932], plate 25

Pima Flutes from [Russell 1908], figure 80

The Breckenridge flute appears to have been designed with these design element, but based on the available artifcacts, significantly pre-dates the Pima and Papago design.

Jim Rees opines on a number of issues relating to the significance of the Breckenridge flute in establishing that the Native American flute has prehistoric roots. Quoting from [Rees 2011], page 8:

The other controversy surrounding the Native American flute concerns its geographic origin. As can now be seen with the Breckenridge flute, the structure of the instrument contains clues to its place of origin. The Beltrami flute, for example, is more-or-less an attempt to turn a piece of wood into a piece of river cane by boring it out from both ends and leaving what amounts to an artificial node between the two chambers. This was necessary in places where river cane was not available. My argument is that the odd design structure of the Native American flute is based on and was originally necessitated by the natural structure of river cane. This is demonstrated by the structure of the Breckenridge flute, which utilizes the natural hollow sections of river cane and the solid nodes that divide them to dictate its two-chambered form. In other words, the maker of the flute exploited the natural structure of river cane to make a two-chambered duct flute. Oddly, in this case, function seems to follow form. The place of origin for the Native American flute therefore must be somewhere within the natural range of river cane. This would place it somewhere within the southeastern United States, which includes the Ozark Plateau, or, more intriguingly, in northern South America.

The latter possibility was raised by Robert Hall ([Hall-RL 1997], pages 114–118), who sought to relate bird stone atlatl weights to bird-shaped blocks/slide often found on nineteenth and twentieth century wooden Native American flutes. He suggested that two-chambered cane flutes ethnographically reported by Izikowitz ([Izikowitz 1970]) in northern South America are the possible prototypes for the Native American flute. However, these South American flutes do not have bird-shaped blocks/slides. Hall suggests that this trait evolved from bird sone atlatl weights only in eastern North America. The rediscovery of the Breckenridge flute neither confirms nor refutes Hall's hypothesis, since its slide system can only be guessed at, but the question of a link between the cane flutes of northern South America and those of eastern North America remains open. What can be said at present is that the Breckenridge flute appears to be the oldest flute of its kind found so far in the Americas.

Other Instruments

In addition to the Breckenridge flute, [Rees 2011] describes a number of other instruments found in Ozark Plateau bluff shelters:

• two notched-cane rasps;
• a gourd rattle, from Brown Bluff in Northwestern Arkansas. This is the earliest recognizable gourd rattle known in the Eastern United States, radio-carbon dated to 1160±110 CE ([Crane 1968] pages 88-89);
• two rim-blown flutes and one possible three-hole rim-blown flute;
• a possible “internal duct, single-chamber whistle” from Edens Bluff.

I am sure that further research will shed more light on these instruments as well as identifing more instruments from the collections of the Museum Collections of the University of Arkansas.