Flutopedia - an Encyclopedia for the Native American Flute

Sign up for our Flute Newsletter


Previous PageNext Page

Anasazi Flutes from the Broken Flute Cave

Map of the area where the Broken Flute Cave flutes were excavated.

Map of the area where the Broken Flute Cave flutes were excavated.

In the summer and fall of 1931, Earl H. Morris led an expedition from The Carnegie Institution of Washington to the Prayer Rock district of NorthEastern Arizona. His team unearthed thousands of artifacts.

Among them were wooden flutes that were constructed between 620–670 CE ([Bakkegard 1961]), which makes them the oldest extant flutes from the Americas made of wood. The cave where the flutes were excavated was dubbed the “Broken Flute Cave”.

The flutes were transferred from the Carnegie Institution to the Arizona State Museum in Tuscon, Arizona in January of 1957.

In June 2001, Ken Light of Amon Olorin Flutes told me about these flutes and a visit that he made to see them with R. Carlos Nakai. He had taken careful measurements of the flutes and made a replica, which he played and let me take careful measurements (provided below).

Then, on August 26, 2002, I had the privilege of studying these flutes, that are still housed at the Arizona State Museum archives. This web page is a collection of information on these flutes, taken from my own observations and various research sources.

My measurements were posted shortly afterward on an earlier version of this web page that was part of the original Flutekey.com, the predecessor web site to this Flutopedia.com.

Since then, many flute makers have begun producing replicas of these beautiful instruments, beginning with Coyote Oldman (Michael Graham Allen). I have one of Michael's replicas made in November of 2005, and the measurements have been incorporated below.

For details of the early development of contemporary rim-blown flutes, see Contemporary Rim-Blown Flutes.

Throughout this page, I use the catalog numbers assigned to the artifacts by the Arizona State Museum.

Here is a more detailed map of various Basketmaker site, from [Coltrain 2007] The Stable- and Radio-Isotope Chemistry of Western Basketmaker Burials: Implications for Early Puebloan Diets and Origins:

Basketmaker Sites in the Four Courners region

Basketmaker Sites in the Four Courners region Larger image

And here is yet another map, from Figure 1.4 of [Hays-Gilpin 1998]:

Location of the Prayer Rock District and other important locations on the Colorado Plateau

Location of the Prayer Rock District and other
important locations on the Colorado Plateau Larger image


Accounts of Morris's expedition identify the area as the Prayer Rock Valley 45 miles SouthWest of the town of Shiprock, New Mexico. The records of the Carnegie Institution name the location as Atahonez Canyon. The valley is on the Navajo Indian reservation, in Apache County, Arizona.

Morris's team excavated fifteen caves. The largest cave contains sixteen dwellings and was later named Broken Flute Cave by the Morris team. Here is a photo of the cave taken by Morris:

Broken Flute Cave. Earl H. Morris photo, Arizona State Museum Negative 6024

Broken Flute Cave. Earl H. Morris photo,
Arizona State Museum Negative 6024 Larger image

Here is a map of the Broken Flute Cave from Figure 5 of [Morris-EA 1980]:

Map of the Broken Flute Cave

Map of the Broken Flute Cave Larger image

From the original field notes of Earl H. Morris:

At the west end of the south wall of Room 4, a slot had been gouged in the floor, its western end extending back under a natural stone in place at the base of the room wall. In this cavity not more than 5 cm. below the floor, were hidden two wooden flutes tied together with strips of yucca. The mouth ends lay toward the east. One of them was stopped at each end with wads of yucca fiber, the other was plugged at one extremity with a corn cob. Enclosing the end farthest distant from the player there was a wad of yucca fiber, placed there to protect the feather ornamentation bound to cover a certain portion of each flute.

These two beautifully crafted and preserved flutes were found almost completely intact. They are cataloged A-13994-A and A-13994-B. Some of the feathers had been removed to identify the species of birds, and are now stored alongside the flutes.

The broken pieces of two other flutes were recovered from another dwelling ("Cistern 7") in the same cave, and were restored by Morris. These flutes are cataloged A-14450 and A-14451.

Four Anasazi Flutes from the Broken Flute Cave

Four Anasazi Flutes from the Broken Flute Cave More information

Material and Construction

All four flutes are constructed of Boxelder (Acer negundo) wood. Details on the construction are provided by Elizabeth Ann Morris, the daughter of Earl H. Morris, in [Morris-EA 1959]:

All of the flutes are remarkably similar in nature except for the decorative techniques described above. The wood was identified by the U.S. Department of Agriculture Forest Products Laboratory in Madison, Wisconsin, as boxelder (Acer negundo). Trees of this species were observed growing in the valley bottoms by the Bernheimer Expedition in 1930. The instruments are long, uniformly constructed tubes. The length of the four complete specimens (Fig. 3) varies between 68.5 and 74.1 cm, and the outside diameter between 2.4 and 2.6 cm. The exterior surfaces of the barrels are highly polished. The hollowing out of the cylindrical bore reflects extremely skillful manufacturing techniques. The walls of the flute barrels vary between 3 and 5 mm. in thickness. The interiors are smooth and even, and must represent some closely controlled process of drilling. This would have been facilitated by the presence of a large corky pith in elder wood.

A number of sources (including earlier versions of this web page) have incorrectly stated that “this tree has a large soft pith that can easily be removed”. It was brought to my attention in October 2011 by Randy Goodhew, who provided this information in a personal communication:

Boxelder (Acer negundo) is a common species of maple found throughout North America. Like other maples, it has a solid core – no pith. Because of its name it is frequently confused with Elderberry (Sambucus, various), a small shrub or tree which does have a pithy core. Indeed, the name Boxelder is said to be derived from the observations that this tree has wood that resembles Boxwood and leaves that resemble Elderberry. Other common names for Boxelder are the Ash Leaf Maple and the Manitoba Maple …

Note that she does not discuss a pithy core associated with this species, i.e. Boxelder. Then she muses about how the job could have been made easier if the wood had been elderberry. I think she is clear about the Boxelder not having a pithy core and thus requiring more skills to produce the bore. Perhaps her remarks should awaken a discussion about how pre-industrial peoples would have created the bores in these flutes. Now, there is a real mystery worth pursuing.

One possible answer has to do with the durability characteristics of Acer negundo (thanks to Mike Prairie for suggesting this explanation).

Acer negundo is a soft maple ([Bergman 2010] Wood Handbook — Wood as an Engineering Material, Centennial Edition) which is a hard and heavy, but relatively weak and non-durable wood ([Cassens 1995]) that is susceptibile to insects and trunk decay ([Gilman-EF 1993]). Insects that infest the tree include the Boxelder bug ([Baker-WL 1972]). The percentage of Boxelder trees that experience trunk decay was over 50%, in one present-day study of trees in urban settings ([Luley 2009]).

The native range of present-day Boxelder growth is rather sparse in the Four-Corners area ([Burns 1990] Silvics of North America, Volume 1: Conifers; Volume 2: Hardwoods), but areas of nearby growth do exist (thanks to Barry Higgins of White Crow Flutes for pointing this out).

So, based on this background, Mike Prairie notes that (personal communication): “one possibility is that the Anasazi flute maker had help from nature to start the process.  If the heartwood was eaten by bugs or rotted (to become "pithy") and the sapwood of the branch was intact, a bone or stone tool may have been enough to scrape the bore clean”.


I have found various dates are cited in the literature relating to these artifacts:

B. M. Bakkegard and Elizabeth Ann Morris reported in 1961 that “Specialists were able to date the period of actual house [the single-room pithouse in which the flutes were found] construction between A.D. 620 and 670” ([Bakkegard 1961]).

Bryant Bannister, Jeffrey S. Dean, and Elizabeth A. M. Gell published a comprehensive collection of tree ring dates in 1966 ([Bannister 1966]). This included a general date range of 354–652 CE for 94 tree ring dates in the Broken Flute Cave. Pithouse 4, where the flutes were found, “has two dates, a reused timber cut in A.D. 491 and a log cut in 628.” Additionally, they note that “There are too few dates from Pithouses 4, 14 and 17 to assign definite building dates. However, the range of dates falls within that of the other houses and they were probably built near the time of the latest dates present in each house.”

In 2007, Joan Brenner Coltrain, Joel C. Janetski, and Shawn W. Carlyle published a detailed analysis using radiocarbon dating techniques that included one item from a burial pit in the Broken Flute Cave. The dating placed the artifact in the range 599–769 CE (Broken Flute Cave, Burial 2, sample 31-1-10/N252.0.1, from a 40–49 year old male, from [Coltrain 2007] The Stable- and Radio-Isotope Chemistry of Western Basketmaker Burials: Implications for Early Puebloan Diets and Origins, page 307).


Anasazi flute detail

Anasazi Detail More information

The two A-13994 flutes have strips of bird skin with red, blue, and black feather fastened near the head (proximal) ends with fiber cords.

This diagram at the right is a detail of the feather decoration of A-13994-B, from [Morris-EA 1959], page 408. The letters indicate the items in order of attachment. The materials are:

  • a. Barrel of the flute.
  • b. Black crown of Stellar Jays (Cyanocitta stelleri).
  • c. Yucca fiber string.
  • d. Light blue breast feathers of Pinyon Jays (Cyanocephalus cyanocephalus).
  • e. Strips of skin with red feathers from the crown of Red-naped Sapsuckers
    (Sphyrapicus varius nuchalis) and the heads of Red-shafted Flickers (Colaptes cafer).
  • f. Yucca fiber string.
  • g. (same as e.)
  • h. Yucca fiber string.

Note that the blowing hole is at the bottom of this diagram. The feathers point toward the player, so the bird feathers are ruffled by the breath of the player.


This section contains my notes from August 26, 2002, from the Arizona State Museum.

Broken Flute Cave flutes in the Arizona State Museum

Broken Flute Cave flutes in the Arizona State Museum More information


Neither of the A-13994 flutes can be seen through, due to a corn cob at the top of A-13994-B, and Yucca fibers at the top of A-13994-A and at the bottom of both flutes.

Due to the fragile nature of the fibers holding them together, these flutes were subject to minimal movement.


This flute has a dull black finish over much of the flute. This finish gives a clear indication of the use pattern, from the wear of the finish through to the underlying wood. From the angle of the wear marks, it is clear that the upper three holes were played with the right hand and the lower 3 holes were played with the left hand. This is opposite of current convention.

This flute has banding, which may have served to prevent or repair cracks.

  • The first 1.9 cm of the flute has no banding and no apparant surface finish.
  • The next 0.92 cm has banding.
  • The next 0.88 cm has the dull black surface finish.
  • The next 1.47 cm has banding.
  • The large central area of the flute has the dull black surface finish.
  • The last 9.26 cm has intermittent banding covering about 70% of this portion of the flute.

While the bore on this flute appears almost perfectly straight, it is not centered. The bore appears “off-center” away from the finger holes, so that the wall at the finger holes is slightly thicker than the wall opposite the finger holes.

The mouthpiece has a flattened area on one side of the flute, approximately 60 degrees counter-clockwise from the finger holes as you look down the flute from the head blowing end.


This flute has no evidence of a finish. It has a polished shiny surface.

This flute has evidence of prior banding.

  • The first 0.87 cm of the flute has no banding.
  • The next 1.7 cm has evidence of prior banding.
  • The next 0.62 cm has no banding.
  • The next 2.94 cm has evidence of prior banding.
  • There is no banding on the remainder of the flute.

Each of the finger holes on this flute has “score marks” emanating from the hole. These score marks appear decorative and are roughly 0.7 cm long. The top holes has about 18 marks, the second and third holes have 14 marks, the 4th hole has 12 marks, the 5th hole has 10 marks on 3/4 of the circumference (the remainder is uncertain due to cracks), and the bottommost hole appears to have similar lines, but they are largely worn away. Only 3 or 4 radial lines are visible on the last hole.

The finger holes on this flute have very wide, smooth bevels in towards the playing hole, most probably worn by playing.

General Notes

All of these flutes were played! There are wear marks where the thumbs rest on the back of each instrument and A-14450 has clear wear marks around each of the six finger holes.

The style of playing must have been similar to the modern Kaval (popular in Bulgaria), and similar in principle to the Japanese Shakuhachi. The mouth hole is largely covered by the lips (maybe 70%) and the player blows across the portion of the rim of the mouth hole that has not been covered by the lips.


These measurements were taken while studying the flutes, and first posted in 2002. Holes are numbered using Flute Player Numbering, all hole measurements are to the center of the hole, and parts of the flute are from Anatomy of the Native American Flute.

Measurement Details A-13994-A A-13994-B A-14450 A-14451
cm inches cm inches cm inches cm inches
Overall Length   73.72 29.02 72.20 28.43 68.30 26.89 73.60 28.98
Outside Diameter   2.50 0.98 2.53 1.00 2.45 0.96 2.46 0.97
Inside Diameter head end 1.74 0.69 1.87 0.74 1.78 0.70 1.62 0.64
  foot end 1.94 0.76 2.06 0.81 1.83 0.72 1.86 0.73
Wall Thickness calculated 0.33 0.13 0.28 0.11 0.32 0.13 0.36 0.14
Hole 1 Location from head end 36.10 14.21 35.26 13.88 31.70 12.48 35.30 13.90
from foot end 37.62 14.81 36.94 14.54 36.60 14.41 38.30 15.08
Hole 1 Size width 0.64 0.25 0.59 0.23 0.68 0.27 0.83 0.33
length 0.64 0.25 0.60 0.24 0.82 0.32 0.80 0.31
Hole 2 Location from head end 40.28 15.86 40.05 15.77 35.80 14.09 39.50 15.55
from foot end 33.44 13.17 32.15 12.66 32.50 12.80 34.10 13.43
Hole 2 Size width 0.60 0.24 0.58 0.23 0.68 0.27 0.79 0.31
length 0.63 0.25 0.61 0.24 0.74 0.29 0.80 0.31
Hole 3 Location from head end 44.38 17.47 43.32 17.06 39.70 15.63 43.30 17.05
from foot end 29.34 11.55 28.88 11.37 28.60 11.26 30.30 11.93
Hole 3 Size width 0.62 0.24 0.58 0.23 0.67 0.26 0.77 0.30
length 0.65 0.26 0.63 0.25 0.75 0.30 0.85 0.33
Hole 4 Location from head end 53.20 20.94 52.59 20.71 49.60 19.53 53.00 20.87
from foot end 20.52 8.08 19.61 7.72 18.70 7.36 20.60 8.11
Hole 4 Size width 0.62 0.24 0.60 0.24 0.66 0.26 0.90 0.35
length 0.64 0.25 0.63 0.25 0.76 0.30 0.78 0.31
Hole 5 Location from head end 57.46 22.62 56.77 22.35 53.50 21.06 57.00 22.44
from foot end 16.26 6.40 15.43 6.07 14.80 5.83 16.60 6.54
Hole 5 Size width 0.54 0.21 0.60 0.24 0.61 0.24 0.73 0.29
length 0.62 0.24 0.63 0.25 0.75 0.30 0.81 0.32
Hole 6 Location from head end 61.64 24.27 60.88 23.97 57.60 22.68 61.20 24.09
from foot end 12.08 4.76 11.32 4.46 10.70 4.21 12.40 4.88
Hole 6 Size width 0.58 0.23 0.65 0.26 0.72 0.28 0.89 0.35
length 0.68 0.27 0.68 0.27 0.83 0.33 0.72 0.28

Contemporary Anasazi Flutes

Since my visit to these flutes in 2002, several makers have begun producing replicas of these beautiful instruments. Here are measurements for two of them. One is from Ken Light and the second is my Coyote Oldman Anasazi replica made in November 2005. All holes on these flutes are round, so only the diameter is shown:

Measurement Details Ken Light Coyote Oldman
cm inches cm inches
Overall Length   75.88 29.88 69.53 27.38
Outside Diameter   2.93 1.15 2.86 1.13
Inside Diameter head end 2.15 0.85 2.25 0.89
  foot end 2.20 0.87 2.25 0.89
Wall Thickness calculated 0.38 0.15 0.31 0.12
Hole 1 Location from head end 35.08 13.81 33.02 13.00
from foot end 40.80 16.06 36.51 14.37
Hole 1 Size diameter 0.91 0.25 0.59 0.23
Hole 2 Location from head end 39.52 15.56 37.15 14.63
from foot end 36.36 14.31 32.38 12.75
Hole 2 Size diameter 0.78 0.31 0.77 0.30
Hole 3 Location from head end 43.05 16.95 41.27 16.25
from foot end 32.83 12.92 28.26 11.12
Hole 3 Size diameter 0.61 0.24 0.79 0.31
Hole 4 Location from head end 58.83 22.38 50.48 19.88
from foot end 19.05 7.50 19.05 7.50
Hole 4 Size diameter 0.77 0.30 0.81 0.32
Hole 5 Location from head end 60.32 23.75 54.77 21.56
from foot end 15.56 6.12 14.76 5.81
Hole 5 Size diameter 0.78 0.31 0.80 0.31
Hole 6 Location from head end 63.70 25.08 58.74 23.13
from foot end 12.18 4.79 10.80 4.25
Hole 6 Size diameter 0.79 0.31 0.80 0.31

For details of the early development of contemporary rim-blown flutes, see Contemporary Rim-Blown Flutes.

Finger Hole Position

It is interesting to look at these six flute (the 4 prehistoric flutes and the two contemporary flutes) in terms of the position of each finger hole as a percentage of the physical bore length (also called the station). In this case, the physical bore length is a straightforward measurement from the head end to the foot end:

Hole Number A-13994-A A-13994-B A-14450 A-14451 Ken Light Coyote Oldman
Hole 1 49.0% 48.8% 46.4% 48.0% 47.5% 46.2%
Hole 2 54.6% 55.5% 52.4% 53.7% 53.4% 52.1%
Hole 3 60.2% 60.0% 58.1% 58.8% 59.4% 56.7%
Hole 4 72.2% 72.8% 72.6% 72.0% 72.6% 74.9%
Hole 5 77.9% 78.6% 78.3% 77.4% 78.8% 79.5%
Hole 6 83.6% 84.3% 84.3% 83.2% 84.5% 83.9%



After the Arizona State Museum acquired the flutes, B. M. Bakkegard played some notes on A-13994-B and measured the pitches ([Morris-EA 1959], page 408 and [Bakkegard 1961], page 185). I also have pitches measured on my Coyote Oldman replica made in November 2005.

Here are Bakkegard's measurements of A-13994-B and my measurements of my Coyote Oldman replica, along with the interval of each finger relative to the root / fundamental note of the instrument. The number after each note is the octave number on the IPN scale, with C4 being middle C:

Fingering A-13994-B Coyote Oldman
Note Interval Note Interval
Finger diagram open open open open open open A4 major 7th G4 major 7th
Finger diagram closed open open open open open G4 major 6th F4 major 6th
Finger diagram closed closed open open open open F4 5th Eb4 5th
Finger diagram closed closed closed open open open D4 major 3rd C4 major 3rd
Finger diagram closed closed closed closed open open C#4 minor 3rd B4 minor 3rd
Finger diagram closed closed closed closed closed open C4 major 2nd Bb3 major 2nd
Finger diagram closed closed closed closed closed closed Bb3 root Ab3 root

As with all instruments of the rim-blown style such as a Kaval or a Shakuhachi, changes in embouchure can produce a wide range in the pitch. Also, cross-fingerings could have been used for additional notes within an octave.

Discovering a Flute

What is it like for an archeologist to discover a pre-historic flute? Earl Morris led a team to Canyon del Muerto in 1923, an area that is now part of Canyon de Chelly [kan-yahn duh-shey] National Monument in Northeastern Arizona. He gave this account from [Morris 1925], page 293-294:

The body was that of an old man, surely once a priest or chief. Beside the usual offerings of beads, baskets, and sandals, there lay above his buckskin wrapping a flute, one end beneath the chin, the other between the thighs. …
Along the left side was a mass of wooden objects, all readily perishable, hence extremely rare in perfect condition. Conspicuous among them were bone-tipped flint flakers with whch knives and projectile points were made, several spears, four handsomely wrought spear throwers, and three more flutes.
I picked up one of the flutes, shook the dust and mouse dung out of it, and placed it to my lips. The rich, quavering tones which rewarded even my unskilled touch seemed to electrify the atmosphere. In the distance Navajo workmen paused with shovels poised, seeking the source of the sound. A horse raised its head and neighed from an adjacent hillside and two crows flapped out from a crevice overhead.
Our little group was motionless for a dozen heartbeats, which seemed as many minutes. In the weird silence it was as if time had been halted in its flight — nay turned back — for in swift array there crowded through my consciousness the scenes of grief and mourning, of savage pomp and ceremonial, amid which the tones of that instrument had last echoed from the selfsame cliff that now glistened under the rays of the setting sun, which for a brief moment had broken through the dark clouds maksing the November sky.


My thanks goes to:

  • Ken Light of Amon Olorin Flutes, for informing me of this amazing resource and allowing me to measure his replica.
  • The staff of the Arizona State Museum, for providing access to their collections and library resources:
    • Suzanne Griset, Head of the Collections Division, 520-621-2676
    • Mike Jacobs, Archeological Collections Curator.

Previous PageNext Page