Skip to main content
sculpin hat header

Smithsonian Uses 3D Tech to Restore a Broken Sacred Object for Tlingit Indians

By Eric Hollinger, National Museum of Natural History on Mon, 09/23/2019 - 12:42

Watch a video below and read on for more information on this project.

On September 25th, 2019, the Tlingit Kiks.ádi clan of Sitka, Alaska, conducted ceremonies to dedicate a new clan crest hat—only this hat was not really new. It was a replica made of Alaskan woods carved by 3D milling machines and ornamented with traditional materials such as deer hide and sinew, ermine skins, copper horns, swan down, and shell inlays. The clan held ceremony in Juneau to put spirit into the newly restored replica so that it could be danced again and put into use for clan ceremonies. The broken hat, in the form of a sculpin or bullhead fish, had rested in the collections of the Smithsonian’s National Museum of Natural History for the past 135 years. The Smithsonian worked closely with the clan to study the broken hat, 3D scan and repair the hat digitally, carve it with a computer numerically controlled milling machine and then finish it by painting it and adding attachments similar to those which had originally adorned the hat. This collaboration is the first cultural restoration of an important religious object for an indigenous community using 3D digitization and replication technology.

The hat was purchased in 1884 by John J. McLean in Sitka, Alaska. McLean was an officer in the US Signal Corps stationed in Sitka from 1881 to 1887 and he made collections of Native Alaskan objects for the U.S. National Museum, the predecessor of the Smithsonian’s National Museum of Natural History (NMNH). The original paper tag attached to the hat described it as a “Wooden Hat, for dancing, Carving Sculping (sic), Hat form”. The hat has a low cone shape, more than 18 inches in diameter, with the face of a sculpin or bullhead fish carved into the front of it and a circular platform on top. The hat rested on the shelves of the Smithsonian for nearly 130 years.

In 2012, when Harold Jacobs, Cultural Specialist for the Central Council Tlingit and Haida Indian Tribes of Alaska spotted it in the museum, he recognized it as a crest hat of the Kiks.ádi clan of Sitka. The hat is badly worn, has a large section missing from the rim, cracks running the length of the piece and sections missing from the ring on the underside of the hat where it rests on the head. Jacobs had previously worked with the museum to repatriate a Killer Whale hat that had belonged to his father’s clan, the Dakl’aweidí. That Killer Whale hat, in collaboration with Dakl’aweidí clan leader Edwell John, Jr., was later digitized and a 3D replica was milled to exhibit for educational purposes in the NMNH. Jacobs asked NMNH Repatriation Office Tribal Liaison Eric Hollinger if the same technology used to replicate the Killer Whale hat could be used to digitize the Sculpin hat, digitally repair it and produce a physical replacement. Jacobs said that the original was too badly damaged and weak to be danced again, but he believed the Kiks.ádi clan would likely wish to ceremonially dedicate a replacement to make it a sacred and ceremonial object known as at.oow. Rather than a repatriation, this would be a form of cultural restoration using digital technologies.

Tlingit crest objects, or at.oow, can be a variety of clan owned objects, but hats in the form of their clan’s crest animals are one of the most common. They are imbued with spirits and are worn and danced by clan leaders on important ceremonial occasions. In Sitka, the sculpin is a secondary crest of the Kiks.ádi clan and their primary crest is the frog. The Kiks.ádi are a Raven moiety clan and clans of the opposite moiety are called Wolf/Eagle. There are numerous clans on either side and each claims ownership or cultural property rights to one or more crests. When a clan “brings out” their crest hats publicly the “opposites” are obligated to bring out their own hats to match them and “hold up” their opposite brethren.

In the Fall of 2012, Hollinger led a Smithsonian digitization team to Sitka for the Sharing our Knowledge Conference where the team demonstrated digitization technology by 3D imaging clan crest objects for clan leaders. Harold Jacobs introduced Hollinger to the leader of the Kiks.ádi clan, Ray Wilson, Sr., (Aanyaanáx), and proposed the project to Wilson. Wilson confirmed that the clan would like to work with the Smithsonian in using the 3D technology to produce a physical replica of the broken original which could be put back into ceremonial use as a clan crest hat. At the conference, Wilson asked that the Smithsonian team digitize several important clan objects as security against damage or loss. Smithsonian Institution Exhibits (SIE) specialists Adam Metallo and Carolyn Thome imaged the Kiks.ádi clan’s Frog Hat, and a helmet and hammer used by Katlian, an important Kiks.ádi ancestor, in battles with the Russians in the early 1800s.

Over the next two years, the Kiks.ádi clan and the Smithsonian engaged in a number of long-distance consultations discussing options and approaches for carrying out the project. Meanwhile, the NMNH researched the record of the accession of the hat. The collector wrote very little about the hat. In the accession file, the museum found a letter from the McLean which described the hat as a “large carved wooden helmet with wooden stove pipe attachment” but no attachment was cataloged with the hat. A search of the rest of the items in the collection sent by McLean was initiated to locate the attachment. More than one hundred catalog entries later in the ledger was listed a “hollow cylinder of wood” which was 22 inches long and 5.5 inches in diameter; an unusually large piece which fit only the platform of the base hat. The cylinder is also badly worn and cracked in several places missing materials that had been attached to it. Drops of green paint on the rim of the hat suggested paint dripped from the cylinder while it was attached to the hat as has wide bands of green paint. Use of portable x-ray fluorescence analysis confirmed the paints matched. Nearly 130 years after they were separated during cataloging the hat base and wood cylinder were reunited once more.

Ray Wilson, Sr., decided that it was critical to visit the Smithsonian to view the hat in person and consult with the museum staff. As a leader of a Raven moiety clan of Sitka, Ray recognized that his words and deeds in something as important as consulting on the hat replication project needed to be witnessed by a clan leader from the Wolf/Eagle moiety of Sitka. He asked Sitka Kaagwaantaan clan leader Andrew Gamble (Anaaxootz) to accompany him to Washington DC to assist and advise in the consultations. Wilson also invited Cyril Zuboff and Garfield George, Deisheetaan clan (Raven moiety) leaders from the village of Angoon to join them and assist with the consultations. The delegation visited the museum and examined the broken hat. They discussed details of its construction, apparent past renovations and how it had been damaged, and repairs attempted. The hat parts had been ct-scanned by Repatriation Office x-ray technician Janine Hinton and she reviewed the scans with the clan leaders. They revealed the grain of the wood, the cracks and a total of 175 holes piercing the hat for former attachments. The scans revealed that the cylinder had been made by splitting a log and then hollowing it out and then lashing the two halves back together.


[image 1 - An iron nail that once served to anchor attachments.]

With a more complete understanding of the existing parts of the hat, Wilson asked that the 3D digitization of the hat proceed. Adam Metallo and Jon Blundell of the Smithsonian’s Digitization Program Office (DPO) explained the digitization process using photogrammetry and structured light scanning. The clan leaders decided that clan protocols for the making of a clan crest hat should be followed as much as possible, which meant that Eagles/Wolves needed to make the hat for the Ravens. Wilson asked that Andrew Gamble, as an Eagle/Wolf clan leader, and Eric Hollinger, as an adopted Dakl’aweidí (Eagle/Wolf), initiate the digitization process by taking the first turns with the scanning equipment. The DPO specialists then finished collecting the data on the hat parts to make sure the files were complete and accurate.

Anaaxootz starting the scanning of the hat.png

[image 2 - Anaaxootz starting the scanning of the hat.]

The clan leaders also recognized that much of the work on the remaking of the hat would involve the programing and operation of the computer numerically controlled milling machine to carve the hat. Therefore, they felt that the CNC machine operator would need to be an Eagle/Wolf. They then interviewed Smithsonian Institution Exhibits (SIE) Exhibits Specialist Chris Hollshwander and Anaaxootz explained to him that the Kaagwaantaan clan would adopt him because it was important that Eagles/Wolves make the new hat for the Kiks.ádi.

During the examination of the hat, the clan leaders noted that the large wooden cylinder atop the hat was unusual. They were used to their hats from the early 1900s being topped with “potlatch rings” made of woven basketry disks stacked in a column. They noted that the wooden “potlatch” cylinder might represent an earlier version of that feature but they had not seen many on Tlingit hats. The delegation visited the rest of the Tlingit collections of the NMNH and the National Museum of the American Indian (NMAI). While in the NMAI’s Cultural Resources Center, Cyril Zuboff noticed a hat among the collections that closely resembled the Sculpin hat in carving style, colors and with a nearly identical wooden cylinder. The hat was from the Wrangell area and the face of the crest figure was in the form of a frog. Because the Kiks.ádi clan was also in Wrangell and was the only clan there that could claim ownership of the frog crest, the delegation was convinced that both hats belonged to the Kiks.ádi and were likely made by the same artist. The NMAI frog hat had also been badly worn and damaged and had been repaired before it was collected. The face of the frog had been cut off of what had once been a larger wood hat and reattached to frame of dowels with the wooden potlatch cylinder attached to the frame. The discovery of this similar hat was important to the delegation and they asked the NMAI staff to allow it to be brought to the NMNH so the two hats could be together for comparison and so they could both be present at a feast held in the museum to end the consultation visit. The feast was attended by NMNH Sant Director Kirk Johnson and other executive staff of the museum, repatriation and collections staff from both the NMNH and the NMAI and staff of the DPO and SIE. In the presence of these ‘witnesses’ Anaaxootz conducted a naming ceremony and publicly adopted Chris Hollshwander into the Kaagwaantaan clan so an Eagle/Wolf could mill the hat. At the feast, clan leaders also asked Director Johnson and the museum staff if, when the new hat was finished, the original broken hat could be brought to Alaska along with it so a ceremony could be held with both the old hat and the new hat together when spirit is put into the new hat. The Director told Wilson and the other clan leaders that the museum would try to make that happen.


[image 3 - The NMAI hat from Stikine closely resembling the NMNH hat from Sitka.]


[image 4 - The delegation and museum staff with hats from NMNH and NMAI.]

The next challenge was finding funding for the project. The digital files still had to be processed to prepare them for milling, large pieces of wood and other materials had to be acquired from Alaska, and Smithsonian Institution Exhibits had to purchase some additional equipment and parts. The Digitization Program Office had completed the capture of the digital data, generously contracted videographers to document the consultation process, and they contracted with SIE to process the digital files. Hollinger submitted a grant request to the Smithsonian Women’s Committee for materials and milling by SIE which received overwhelming enthusiasm from the committee membership and was generously funded.

What had been discussed and agreed to during the consultation process was formalized in a memorandum of understanding signed by NMNH Director Johnson and Aanyaanáx for the Kiks.ádi clan. The clan authorized the museum to make two replicas, one would replace the original and become the property of the clan, while the other would be retained by the museum for exhibition and educational purposes to aid in telling the story of the collaboration. The museum agreed to restrict future digitization of the broken hat and the replicas and access to the raw files while the clan authorized display of the digital models on the web and telling of the story as long as all images and text were first approved by the clan.

The remaking of the hat was a long and deliberative process. The clan decided that it would be important to make the new hat for the clan from alder while the replica to be retained by the museum could be yellow cedar. Both woods were traditionally used for crest hats, but the clan felt that the cracks in the original may have occurred because the cedar was too weak for such a wide hat. They felt they were learning a lesson from their ancestors and that alder would make for a stronger hat since they intended to dance with it. The museum’s replica would not be danced in ceremony, so it was fine for it to be made of the weaker cedar. The clan found an alder tree with a large diameter trunk being felled near Juneau, Alaska, and they cut, split and sealed two large rounds for the museum to ship to Washington DC where it was moved into freezers to reduce the risk of splitting while it dried. Icy Straits Lumber in Hoonah, Alaska was commissioned to provide yellow cedar rounds for both the base of the museum’s replica and the potlatch cylinders for both new hats. Enough wood was shipped to provide for two backups in case pieces split during the milling or flaws were discovered in the wood.

Meanwhile, at the SIE shops, Exhibits Specialist Carolyn Thome processed the digital files to prepare them for milling. File preparation usually takes considerably more time than the capture of the digital data. She digitally repaired the hat using a variety of software to erase iron nails, sinew knots and straps, and leather lashings. She joined the cracks and bridged the missing sections by digitally sculpting the gaps. The two halves of the potlatch cylinder had become warped and offset so she had to reshape and move the halves to align them to the position they had been in when it was first made. Because the photogrammetry and scanning could only see the edges of the interior of the cylinder, Thome used the ct-scans of the interior to complete the digital model. Thome was careful not to fill in holes that marked where the hat had been pierced for attachments because the clan had asked that those be left as traces marking the work of the original artist and prior renovations.

 SIE Exhibits Specialist Chris Hollshwander roughed out the wood to fit in the CNC machine and programmed in the digital files prepared by Thome. To test the files and machine settings he milled a ½ scale prototype of both the base and cylinder. The prototype was shown to Aanyaanáx via video conferencing and then later in person. The computer milling process uses drill bits of increasingly smaller sizes to remove finer and finer material in a very controlled fashion. As is the practice of traditional carvers, the wood needed to rest in freezers between carving episodes to prevent uncontrolled warping and cracking. The milling process was observed live by clan leaders via video conferencing. Like the original, the logs for the cylinders were split and then the interior was milled for a total of three pieces per hat.


[image 5 - Discussing the ½ scale prototype.]

The museum and clan also used video conferencing to consult remotely on the painting and materials used on the original hat and the sequence of modifications that had been made to the hat over the course of its use-life. Ct-scans of the hat allowed the mapping of all 175 holes in the hat and revealed the sizes and angles of each piercing. Some holes had been painted over or filled with a mud slip. Rows of small holes on the mudded over back of the hat suggested sea lion whiskers, representing spines of the sculpin, had once been present but were removed and covered over. Because the back had not been painted and nails on the upper back had bits of leather under them, the clan felt it was likely that it had last had a cape affixed to the back which probably draped down over the shoulders of the wearer. No teeth are present on the original hat but when clan leaders asked the museum to look closer at the empty sockets in the mouth a fragment of operculum shell was spotted indicating it had once had shell teeth. Inch by inch examination of the original hat under magnification revealed isolated hair and bits of feather embedded in paint, cracks and holes. There were even several hairs stuck under splinters on the underside of the hat where previous wearers had probably lost them.


[image 6 - The broken original sculpin hat.]


[image 7 - The broken potlatch cylinder.]


[image 8 - The angles of the sea lion whisker holes revealed.]

Clan leaders felt that it was important to know as much as possible about what materials had been attached to the hat. Aanyaanáx wrote a letter to the Anthropology Department’s Collections Sampling Committee asking to have minimally destructive analyses applied to identify the paints, hairs, feathers, sinew and leather remnants on the hat. Small samples of the materials were extracted and NMNH Zoologist Suzanne Puerach microscopically identified hairs as belonging to goat, deer, ermine, and human. NMNH Zoologist Carla Dove microscopically identified feather bits from both the base and potlatch cylinder as swan down. Materials scientists Timothy Cleland and Asher Newsome with the Smithsonian’s Museum Conservation Institute (MCI) applied mass spectrometry to the samples. Proteomic analyses revealed ermine, walrus, elk or deer and goat. One bit of material extracted from a small hole was even identified as eagle. Direct Analysis in Real Time (DART) mass spectrometry was applied to paints and various surfaces of the hat and revealed a wide range of chemicals. Perhaps the most interesting was the discovery of traces of nicotine on both the base and the cylinder in locations that suggest it was exposed to tobacco smoke while the pieces were articulated.


 [image 9 - Select pictures of DART analysis of hat and cylinder.]

Once the milling of all parts of the two hats was completed, it was time to paint them and attach materials to them. This required more traditional techniques and interpretations of how it should appear. Hollshwander and Hollinger were to complete the work on the replicas, but lacked experience with making Northwest Coast art, so Aanyaanáx asked Cyril Zuboff to travel to the museum to teach and advise the Wolf/Eagles in finishing the hats. Adopted Kiks.ádi clan member George Reifenstein sent the hide and sinew of a deer he had killed in Alaska as well as resin from a spruce tree in his yard. Zuboff donated shell for the eyes and some of the ermine skins. The milled hat served as an exact model of the original hat as a foundation, but the clan decided that some modifications to the colors and design were appropriate to have it serve the clan as a contemporary hat. Aanyaanáx decided that he preferred the dark green used on more recent clan hats over the bright blue surrounding the eyes of the original. The original red paint was vermillion-based and contained toxic mercury, so non-toxic contemporary acrylic paints were used on the new hats. Although the eyes of the original hat did not have inlays, they did exhibit incised circles that suggested the original artist intended to add inlays at some point but was unable to complete it. Therefore, the eyes were carved out and inlayed with abalone shell. The deer hide was painted in the form of a sculpin and attached using sinew as a cape on the back of the hat. Ermines were attached to drape from the cape. Many sculpin form hats and images are depicted with horns and the original had indicators that elements had been attached behind the eyes, so copper covered cedar horns were attached behind the eyes. Sea lion whiskers, representing water spray or spines, had probably once adorned the back of the original hat to the clan decided to add seven whiskers on each side of the head. The whiskers for the Kiks.ádi hat were obtained by a Kiks.ádi clan subsistence hunter in Ketchikan, Alaska.

Digital repairs to the broken hat.png

[image 10 - Digital repairs to the broken hat.]

Hollshwander milling the new hat.png

[image 11 - Hollshwander milling the new hat.]

Videoconferencing with clan during milling.png

[image 12 - Videoconferencing with clan during milling.]

Hollshwander transferring the pattern.png

[image 13 - Hollshwander transferring the pattern to the new hat as Zuboff advises on the work.]

Hollshwander painting the replica hats..png

[image 14 - Hollshwander painting the replica hats.]

Zuboff advising Hollinger on the sinew for chin straps.png

[image 15 - Zuboff advising Hollinger on the sinew for chin straps.]

Hollshwander assembling the new potlatch cylinder.png

[image 16 - Hollshwander assembling the new potlatch cylinder.]

Hollshwander trimming the deerskin cape.png

[image 17 - Hollshwander trimming the deerskin cape.]

Zuboff assisting Hollshwander with attaching the cape.png

[image 18 - Zuboff assisting Hollshwander with attaching the cape.]

Attaching the potlatch cylinder.png

[image 19 - Hollshwander attaching the potlatch cylinder.]

Attaching the whiskers.png

[image 20 - Hollinger and Hollshwander attaching the whiskers.]

Preparing the whiskers.png

[image 21 - Hollinger preparing the whiskers.]

Fitting the ermine skins.png

[image 22 - Hollinger fitting the ermine skins.]

The nearly finished replicas with the old hat.png

[image 23 - Nearly complete replicas with original hat.]

Nearly complete replicas with original hat.png

[image 24 - Hollshwander and Hollinger standing with the nearly complete replicas with original hat.]


[image 25 - The Kiks.ádi clan’s restored hat.]