1 What “Really Happened” A Migration Narrative from Southeast Alaska Compared to Archaeological and Geological Data

1. K’áadasteen’s account begins with the death of a chief named Łtaxda’x who owned a valuable feast dish [an at.óow object symbolizing clan identity] and a tináa [copper shield, an at.óow wealth item]. The name of the clan was Ca’dadūx (K).

   a. Łtaxda’x was the leader of a Raven moiety Copper River [Ahtna] clan that lived at Chitina. This clan was called the Gineix Kwáan [rather than or in addition to Ca’dadūx], a name that comes from the Gineix (Big Bremner) River, which joins the Copper River below Chitina (HB).

   b. The clan’s name was Gineix Kwáan but the leader’s name was GuditŁa (SW, KDI).

   c. The dish was made of horn from a giant moose (HB, KDI, EA); the moose horn was decorated with abalone shell (SW, KDI); the dish was made of walrus ivory inset with beautiful stones (MH); it was made of wood with dentalium shells around the edge (KDI). [The bowl combines products of the interior and coast and may be a symbol of trade or migration.]

2. There was a dispute over inheritance of the feast dish, leading to a schism in the chief’s house [hít, “house” and the matrilineal kin group residing there]. A faction of about forty people led by four brothers decided to leave and “go straight for that mountain” [Mount Saint Elias] (K).

   a. One brother kept the dish at Chitina and the other three left with their families, one going down the Copper River and settling at its mouth; one travelling south on the ocean to Yakutat in a skin boat; and the fourth setting out across the glacier. The migration across the ice took a hundred years; the people travelling by boat arrived at Yakutat before them (MH).

   b. “They [women of the clan] moved in a body to the side of the younger brother, which meant they were voting for the younger brother to be their chief. So he leaned over and got that bowl. That signified that he’s chief. And then that migration started.” The brother who was not chosen as leader led the migration to Yakutat (EA).

   c. The migration took place ten generations ago, and ancestors from each generation can be named (MH).

3. The travellers crossing the glacier wore hats, coats made of weasel and marten skins, and nose pins. As they approached the mountain they found a place with many ground squirrels. They clubbed the squirrels, which caused fog to appear. The group became lost in the fog, and some were separated and disappeared (K).

   a. The group that split off in the fog turned back and came out at the mouth of the Copper River, later moving to Katalla up the coast; they were given the name Ganaxtedi (HB).

   b. The people crossing the glacier were starving. In the distance they saw a little hill with trees [a pinnacle projecting above the ice] and mistook it for a wolverine. They later saw what they thought was a rabbit, but it turned out to be the distant peak of Mount Saint Elias. “It was a compass for the people so they wouldn’t get lost” (HB).

   c. The brother and his family who were travelling by sea saw the snowy peak of Mount Saint Elias in the distance and thought it was a seagull on the water (MH).

4. The migrating group climbed toward Mount Saint Elias [across Bagley Icefield], and then found a way around its west side [descending through a pass to Yahtse Glacier]. As they struggled on the ice they sang a song from their Copper River homeland and mourned loved ones who had been left behind or lost in the fog (K).

   a. The people composed songs in the Ahtna language that they sang on the migration. One was a mourning song about a man who accidentally shot his brother with an arrow (MH). Some say he was shot with a gun (KDI), but there were no guns at that time (HB).

5. From the heights they saw the ocean [Gulf of Alaska] for the first time, saying, “What is that so very blue?” They went down to the ocean to save themselves [following Yahtse Glacier to Icy Bay]. At the bottom they crossed a [meltwater] river that was boiling out from under the ice. They claimed Mount Saint Elias as an at.óow crest because they were the first to pass by it (K).

   a. They came upon Mount Saint Elias and adopted it as a crest. “They danced down from that mountain. They were happy when they are coming on this side. Lots of things happen there and there are songs” (SW).

   b. “The glacier was formed so there were steps all the way down to the water, and there was gravel on top of the ice. And every step or platform that was there, they made songs and danced” (KDI).

6. They built a house beside the river to shelter for the winter, naming it Mountain House [Shaa Hít, Swanton 1909, 350] in memory of how close they came to dying on the trek. They resided at Icy Bay for ten years, building a whole town (K).

   a. At Icy Bay “the glacier was all over the bay, way out” [at its maximum extent, larger than today]. The settlement at Icy Bay (Was’ei) was a temporary camp called Teey Aaní (“Yellow Cedar Bark Town,” Thornton 2012, 17), named for bark they brought with them to cover their dwellings. The place was just west of what is now the bay, at Was’ei Dak (“Outside of Was’ei,” Thornton 2012, 17) (HB).

7. At Icy Bay a woman adopted a seagull that grew to a giant size. Young men were sent from Icy Bay in a skin boat to explore along the coast to the south. They got to Yakutat Bay and crossed it to a town where Koskedi and Łuẋedi residents [probably Eyak clans living at Lost River] turned them away. They returned to Icy Bay. Some months later, a group of [Galyáx] Kaagwaantaan came to Icy Bay in a skin canoe from the mouth of Copper River and they were welcomed (K).

   a. The leader of the Gineix Kwáan at Icy Bay was concerned that they would become a “lost tribe” because they were Ravens and had no Eagle moiety partners to marry. The arrival of the Galyáx Kaagwaantaan, an Eagle Eyak clan, saved the Gineix Kwáan men from having to marry their “sisters” of the same clan (HB).

   b. The leader of the Galyáx Kaagwaantaan was a Teikweidí [Tlingit Eagle] man named Xatgawet. He married two beautiful Gineix Kwáan sisters, ‘Àndúł and Dúhàn, and became rich because of the copper they owned (SA, EA, LF; see de Laguna 1972, 242–245).

   c. The Gineix Kwáan encountered the Galyáx Kaagwaantaan after finding blood on the ice where hunters had been skinning seals (HB). [Icy Bay is a major harbour seal rookery.]

   d. While they were living at Icy Bay a boy fell into a glacial crevasse and could not be rescued; his mother adopted a seagull in his memory (SW, SA, HB). [A mourning song for the child is still performed at Yakutat in the Ahtna language; see de Laguna 1972, 1157]. According to Harry Bremner, this was the origin of the giant seagull story, which he described as myth (tlaagú) (HB).

   e. The people whom the bothers met living at the town in Yakutat Bay were Łuẋedi, an Eagle clan (KDI) or “Aleuts” [Chugach Sugpiaq people from Prince William Sound] (MH, SW).

8. Six Gineix Kwáan brothers returned to the Copper River to retrieve a tináa [copper shield or plate] whose “real owner” [possibly the clan leader Łtaxda’x from episode 1 above] had died. Bringing the tináa with them, the group [Gineix Kwáan and Galyáx Kaagwaantaan, now intermarried] went to Yakutat Bay by boat. They crossed Yakutat Bay and came ashore on the other side in an area occupied by the Koskedi [an Eyak clan] (K).

   a. [Instead of by sea] the group went on foot “across the ice” all the way from Icy Bay to the east side of Yakutat Bay near Mount Tebenkoff [a route that implies that, at the time, the conjoined Malaspina and Hubbard glaciers extended across Yakutat Bay at a point north of Knight Island; see figure 1.3]. The travellers saw a “beautiful beach” below [possibly Logan Beach] and went down to it, meeting the Hmyedi in the vicinity of Knight Island (HB).

   b. “They came walking overland from Icy Bay and found Yakutat” (GJ). Some boys from Icy Bay ran across the [Malaspina] glacier to Yakutat Bay and discovered that people were there from signs of seal hunting (SW).

   c. A solid glacier covered all of Yakutat Bay and extended north to Icy Bay and beyond; it began receding when the Gineix Kwáan immigrants killed a dog as they approached Icy Bay and threw it into a crevasse (KDI).

   d. “When the people first came to this area, the glacier extended from Point Latouche [see figure 1.3] across to the Manby side [west side of the bay]. The Manby side was apparently then all ice. Knight Island was bare of trees, just as it is now around . . . those areas from which the glaciers have recently retreated.” The sandy places on Knight Island around Old Town were covered with strawberry plants, and there was no forest on Krutoi Island (JB).

   e. The whole of Knight Island was a strawberry patch; there were no trees (OA).

9. The Koskedi were hostile to the new arrivals, and when they discovered a man from the immigrant group fishing at a stream called Kwáashk’ (Eyak, “humpback salmon”) they broke his salmon spear. To settle the dispute the six brothers bought Yakutat Bay with the tináa, which was worth ten slaves. All of the Koskedi and Łuẋedi then left the bay (K).

   a. The original owners of Kwáashk’ creek [variously the Koskedi (K), Hmyedi (HB), Yinyeidi (EA, LF), or Aleuts (SW)] sold the creek, and the new owners [the Gineix Kwáan] thus acquired the name Kwáashk’i Kwáan [people of Kwáashk’] (HB, KDI, EA).

   b. The same group that owned Kwáashk’ creek caught the Kwáashk’i Kwáan daughter of a Galyáx Kaagwaantaan clan leader as she was picking strawberries on Ganawás [Knight Island] and cut the berry basket from her back. Her father then bought Knight Island for her clan (HB). Alternatively, the girl’s brother, Dux, bought the island (JK); or it was Xatgawet who paid with copper for Kwáashk’ creek and Knight Island (EA, LF, SA, KDI).

   c. The purchase of territories in Yakutat Bay was made with the tináa brought from Copper River (worth eight enslaved people) and also sea otter furs (HB), or with a canoe that had fourteen tináa tied to its thwarts, each worth ten enslaved people (SW). “Because they lived up the Copper River the Kwáashk’i Kwáan had tináa then. They used copper for everything—for knives, whenever they had a war” (SW).

   d. The immigrants built a big town on Ganawás; it had only Kwáashk’i Kwáan houses (SW). Dux, a Kwáashk’i Kwáan clan leader, built the first house there, called Noow Hít (Fort House) and a man married to his sister built Xóots Hít (Bear House) (JK). The village on Knight Island was the oldest one around Yakutat Bay (OA). Its real name was Yéil Áa Daak Wudzigidi Yé (Place Where Raven Fell Down) because “there were so many big houses there, and when it’s calm weather, the smoke goes straight up. So the raven that tries to fly over never gets to the end. It falls down” (OA, SA, EA). The people lived there before the Russians ever came to Yakutat (figure 1.5). [The earliest contact was in 1788, with the Izmailov-Bocharoff expedition.]

   e. It was Xatgawet, a Teikweidí rich man who owned many slaves, who built the village on Knight Island (EA, LF, JE, SA, KDI). He named it Tlákw.aan (Old Town) after Tlákw.aan (Klukwan) on the Chilkat River in Southeast Alaska, in order “to pretend it was a high-class people’s place” [provided by various Yakutat commentators; however, others believe Xatgawet lived during Russian times and had nothing to do with the Gineix Kwáan migration; see de Laguna 1972, 245–247].

   f. Another name for Knight Island was K’ootsinadi.aan (Shaken Land), “because there were so many of them, the land shook when they walked” (EA, LF).

10. A mountain spirit granted the youngest of the six brothers great hunting powers and showed him that the animals of Yakutat Bay, including grizzly bears, black bears, and mountain sheep, lived inside a mountain. Later the brothers all went together in a canoe to hunt for seals in front of the glacier, which was “the seals’ home” and where the animals were abundant [the ice floe harbour seal rookery at Hubbard Glacier]. The position of the glacier at that time was “just at the head of Kwáashk’“ [that is, at the head of the creek’s drainage, also suggesting a mid-bay position similar to that described above in commentaries to episode 8]. Before crossing that glacier [or in front of it], people listened inside a hollow cottonwood tree for sounds of approaching storms that might make the crossing dangerous (K).

    a. “When they came down there it was a foreign country. They didn’t know what to eat; they didn’t know how to live. And the spirits of that place adopted the humans. [. . .] They showed them how to hunt seal, and they became friends of the spirit of the glacier. That is why they have a special connection with the glaciers and the mountains in all that area.” (EA).

The remainder of K’áadasteen’s narrative diverges from the Gineix Kwáan migration story and is not relevant to the present analysis (Swanton 1909, 361–368).

Variations in the Gineix Kwáan migration narrative, including the differing names recalled for places, individuals, and clans, are noted in the synopsis above. Given the continual process of change inherent in oral tradition and the multiplicity of narrators in each generation, it is possible that some of these variants already existed at the time of K’áadasteen’s recounting in 1904, while others may have arisen more recently.

Despite such differences in detail, the same basic sequence of events occurs in all versions, suggesting that long-term fidelity to what “really happened” resides in the main plot elements of the narrative. Among these core elements is the migration itself from the Copper River up the Chitina River basin and over montane icefields to Icy Bay and Yakutat Bay (episodes 2 through 5 above), although a coastal route for part of the group was recalled by one narrator (episode 2a). All versions agree that an interim settlement was built at Icy Bay (episode 6); that the immigrants found Yakutat fjord still filled with glacial ice to a point north of Knight Island (episodes 8, 10); that the immigrants came into conflict with Yakutat residents over access to food resources (episode 9); and that one of their leaders bought territory, including Knight Island and Kwáashk’ stream, using one or more tináa brought from the Copper River homeland (episode 9). The Kwáashk’i Kwáan then built the town known as Yéil Áa Daak Wudzigidi Yé or Tlákw.aan on Knight Island and lived there until abandoning the settlement prior to Western contact (episode 9). They fished for salmon (episode 9), hunted seals along the glacial edge, and harvested land animals, including bears and mountain sheep (episode 10).

The temporal uncertainty of oral tradition is reflected in the difficulty of deriving a secure date for the migration from narrative evidence alone. Maggie Harry (episode 2c) believed that it took place ten generations prior to her own and was able to list ancestors from each cohort (de Laguna 1972, 240). This suggests a date of about AD 1690, although generations are often under-counted in oral tradition (Henige 1974, 27–38; Vansina 1965, 153–154). In connection with the Kwáashk’i Kwáan purchase of Knight Island, the same narrator in the 1950s commented that “three hundred years ago there were no trees at Yakutat—just strawberries” (de Laguna 1972, 236). However, Maggie Harry’s belief that the migration across the montane ice took an entire century is unsupportable given the hostility of that environment to human occupation. Katy Dickson Isaac’s reference to the use of a firearm during or before the migration (episode 4a) is clearly an anachronism, since all other information places the migration in pre-Russian times. Similarly, the Tlingit leader Xatgawet is ambiguously associated with both the migration and post-contact times (episodes 7b, 9e), possibly an example of temporal compression (mixing of eras) in oral narrative (Crowell and Howell 2013) or due to two individuals possessing the same name.

Observations about environmental change are embedded in the narrative and in associated place names, a regional cultural-linguistic pattern (Connor et al. 2009; Cruikshank 2001; Monteith et al. 2007; Thornton 1997, 2008; 2012, xi–xxiii). The extent of Yahtse Glacier in Icy Bay (episode 6a) and of the Hubbard/Malaspina ice mass in Yakutat Bay (episodes 8, 10) are noted in addition to glacial retreat (episode 8c), the exposure of new land (episodes 8d, 8e), and post-glacial plant succession (episodes 8d, 8e). Katy Isaac Dickson’s explanation for the glaciers’ retreat—that a dead dog was thrown into a crevasse (episode 8c)—is based in the belief that glaciers are sentient beings who respond to an invitation “to eat” by advancing and to pollution or human disrespect by pulling back (Connor et al. 2009; Cruikshank 2001, 2005; de Laguna 1972, 286–287).

A guiding cultural theme of the story is reciprocity between Raven and Eagle moieties (Raven and Seagull among the Athna), the complementary “halves” of matrilineal Tlingit, Eyak, and Athabascan societies that intermarry, exchange resources, and support each other during times of loss and grief (de Laguna 1972, 1990a, 1990b; Worl 2010). In the Gineix Kwáan migration story, the shared moiety structure supports intermarriage and the exchange of wealth and knowledge across cultural-linguistic boundaries. Thus, an Eyak Eagle clan (the Galyáx Kaagwaantaan) shares its skin boats, weapons, and sea mammal hunting skills with its Ahtna Raven marriage partners, the Gineix Kwáan (episode 7a), enabling them to learn “how to live” in the coastal environment of Yakutat Bay where they become seal hunters (episode 10), aided by protective mountain and glacier spirits (episode 10a). In turn, the Gineix Kwáan use tináa made of native copper from the riverine interior to buy land at Yakutat and to fund their social partnership with the Galyáx Kaagwaantaan. It is also said that Xatgawet, a Tlingit Eagle of the Teikweidí clan and leader (in some versions) of the Galyáx Kaagwaantaan, “became rich” in copper through his marriage to two Gineix Kwáan sisters (episode 7b).

In oral tradition, Tlákw.aan is described as the first and most important Yakutat Bay settlement of the Gineix Kwáan and is given the name Yéil Áa Daak Wudzigidi Yé because of its “many big houses” with smoking hearths. The Tlákw.aan site should therefore be extensive, containing the remains of numerous multi-family houses.

The YAK-00007 site (figure 1.6) occupies a forest clearing of about 200 × 100 metres (2.0 hectares, or 4.94 acres) bordered by a mature spruce and hemlock forest with trees up to 1.5 metres in girth. Ages of the largest trees have not been determined but are in the range of two to three centuries. Younger trees have encroached on the clearing over the last century or more, and vegetation in open areas includes rye grass, wild celery, and salmonberry bushes.

De Laguna mapped seven square or rectangular house pits at the site (Houses 1–7) as well as smaller surface depressions left by underground storage caches. Four areas of midden were found containing animal bones, shell, charcoal, and fire-cracked rock (Mounds A–D). Average midden depth was less than 75 centimetres (30 inches), but some pits extended up to 230 centimetres (90 inches) below the surface. Over forty subsurface pits and buried structures were uncovered, representing earlier phases of occupation. The two largest dwellings at the site were House 1 and House 7, both about 15.6 metres (50 feet) long. The other houses were 5.9 metres to 9.1 metres (18 to 30 feet) long.

A second group of cultural depressions was discovered in 2014 to the northeast of the main occupation area (see figure 1.6, Features P–KK). These are circular to rectangular and range in size from 2.0 metres to 4.4 metres in maximum dimension. While some of the larger pits could represent small semi-subterranean houses, most or all were probably winter storage caches for meat, dried fish, berries, and other foods utilized by the inhabitants of Tlákw.aan.

The areal extent and number of houses at the YAK-00007 site confirm that Tlákw.aan was at least a moderately sized village, although not as vast as oral tradition suggests. It is more extensive than most other Yakutat sites, although Diyaaguna.éit on the Lost River is far larger, with twenty-six house pits (Davis 1996, 192–200).

An approximation of the population of Tlákw.aan may be derived from comparative ethnohistoric information. The largest traditional houses in the Tlingit region were 15 metres to 18 metres square (49 to 59 feet square) and sheltered forty to fifty people (de Laguna 1972, 294–299, 1990a, 207–208; Emmons 1991, 59–68). This suggests an average of about 6 square metres of interior space per person. If one extrapolates from the total floor area of the houses at Tlákw.aan (707 square metres, without including any of the northeastern pits), the resident population would have been around 118 persons. This assumes that all the surface houses were occupied at the same time, although some evidence (discussed below) indicates that this was probably not the case. Also, because it is based on surface features, the estimate would apply only to the final period of site occupancy rather than to deeper, older layers. K’áadasteen (episode 2) stated that the original group of Ahtna migrants included 40 people, who were joined by an unknown number of Eyaks at Icy Bay (episode 7). The descendants of this relatively small founding group might have increased in number over time, expanding the village.

The primary line of oral tradition holds that a Copper River Ahtna Raven clan (the Gineix Kwáan) joined an Eyak Eagle clan (the Galyáx Kaagwaantaan) during the migration, and that members of both groups built the village of Tlákw.aan (episode 7). In an alternative version, Xatgawet, a Tlingit Teikweidí clan leader, constructed the settlement (episode 9e). House remains at YAK-00007 should permit identification of the cultural origins of the site’s founders, to the extent that Ahtna, Eyak, and Tlingit houses differed from each other in style and construction.

The traditional residential structures of the Ahtna, Eyak, and Tlingit—variously known in English as lineage, chiefs’, or winter houses—shared basic features of design as well as variations that can be archaeologically distinguished. All were plank buildings with wooden frames and housed multiple families related by matrilineal descent and marriage. The Ahtna chief’s house of the nineteenth century was reported to be rectangular (5–10 metres long) with a floor excavated up to 1 metre below ground level (Allen 1887, 130; de Laguna and McClellan 1981, 645; Ketz 1983, 145–149; Shinkwin 1979, 40–50). Wooden sleeping platforms lined the walls and were partitioned into family compartments. The house had a central cooking hearth, overhead smoke vent, vertical and/or horizontal wall planks, and a bark roof supported by single or double ridge poles. A rectangular annex for steam bathing was often connected to the main house.

Eyak multi-family dwellings were similar in size and construction to their Ahtna counterparts but had only a single ridge pole (Birket-Smith and de Laguna 1938, 32–43; de Laguna 1990b, 181). Birket-Smith and de Laguna described historic Eyak house floors as level with the ground, but Davis (1996, 210–309) interpreted pit houses at Diyaaguna.éit and Wulilaayi Aan on the Yakutat foreland as Eyak because they antedated Tlingit migration into the area. These were associated with calibrated radiocarbon dates as early as the tenth century AD and were built inside pits up to 2 metres deep with vertical plank walls that extended to the bottom; several had earthen benches and sleeping platforms around the central hearth. The Eyak practice of using grooved base frames to secure the lower ends of vertical wall planks (de Laguna et al. 1964, 73) was not observed at these sites.

Tlingit lineage houses of the eighteenth and nineteenth centuries varied in size (6–18 metres long) but were typically larger than Ahtna or Eyak dwellings of the same period (de Laguna 1972, 294–299, 1990a, 207–208; Emmons 1991, 59–68; Russell 1891, 79–80; Seton Karr 1887, 156–157). The house had vertical plank walls based at ground level without a bottom frame. Entry was through a circular opening in the front wall. The interior pit was up to 4 metres deep and surrounded by stepped residential platforms that were divided into family apartments. A wood-burning hearth occupied the centre of the floor with a smoke hole above. The roof was covered with spruce planks and supported by two heavy beams. Posts that held up the ridge beams were often carved and painted with clan crests, as was the house front. In some houses a screen with crest emblems divided the house leader’s apartment from the rest of the interior space.

De Laguna interpreted all the houses at YAK-00007 as Tlingit in design, apparently because they had excavated floors and lacked the basal wall frames she believed were indicative of Eyak construction (de Laguna et al. 1964, 43–76). Nonetheless, all had Eyak- or Ahtna-like aspects and did not conform entirely to the Tlingit ethnohistoric model. House 8, found buried under Mound B deposits and built early in the history of the site, had vertical wall planks inside a pit as at Diyaaguna.éit, a bark-covered roof, and a single ridge pole. House 9, which dated to a later period based on its superposition over House 1, was similarly constructed with a single ridge pole and no side benches. House 1 had a side bench along its north wall and a double-beam gabled roof “as on the large Tlingit houses of historic times” but lacked a stepped pit and central hearth. House 7, also interpreted as a Tlingit-style lineage house, had a deeply excavated floor but no side benches.

Given these variations, and taking into account the architecture of Eyak houses built at Diyaaguna.éit and Wulilaayi Aan, de Laguna’s identification of the Tlákw.aan structures as Tlingit is questionable. The structures instead seem to reflect the mixed Ahtna-Eyak heritage that the oral tradition would project, although most of the houses were larger than either of those groups built during the post-contact period after they had suffered severe population decline. It is notable that House 7, one of the largest dwellings at Tlákw.aan, was interpreted by de Laguna as one of the oldest because of the mature spruce trees that have overgrown it, making it unlikely to have been a product of late Tlingit influence.

Maggie Harry believed that Tlákw.aan was founded ten generations, or about 200 years, before her own birth in 1892, and that people lived there until shortly before Russian or Spanish contact (episodes 2c and 9d). The implied span of occupation is thus about a century, approximately AD 1690–1780. Archaeologically, multiple layers of cultural debris would be expected for a site that was inhabited for this period of time or longer, and the exact chronology of occupation should be determinable from calibrated radiocarbon dates on samples of plant material (charcoal or wood, for example) from different levels of the midden.

De Laguna reported two uncalibrated dates on charred wood from the lower levels of Mound B: 136 ± 62 radiocarbon years before “present” (or RCYBP; this would mean 1950) and 328 ± 78 RCYBP (de Laguna et al. 1964, 206). The older of these dates falls in the early seventeenth century AD, about as expected from Maggie Harry’s oral information. However, the two dates are quite different and do not adequately define the occupation period.

One purpose of the 2014 test trench was to record midden stratigraphy and collect a vertically controlled series of radiocarbon samples for more precise dating. The trench deposits (figure 1.7) were 60 to 65 centimetres deep and corresponded to the general site stratigraphy reported by de Laguna (de Laguna et al. 1964, 36–41). Stratum 1 consisted of brown humus accumulated from vegetal growth over the centuries since the site was last inhabited. Stratum 2 was the uppermost cultural layer, composed of sand mixed with large amounts of charcoal and fire-cracked rock generated by cooking fires and steam bathing. Stratum 3 was sand mixed with lesser amounts of fire-cracked rock, charcoal, and fragments of burnt animal bone. Stratum 4 at the base of the midden (the earliest period of occupation) was the thickest and most complex deposit, composed of beach sand interlayered with charcoal, fire-cracked rock, and well-preserved fish and animal bones, as well as small piles of marine shell (littleneck clams, cockles, mussels, sea urchins, and snails). De Laguna reported that bone, stone, and copper artifacts occurred throughout the midden but at a somewhat higher frequency in the upper cultural strata (de Laguna et al. 1964, 85–86).

Seventeen samples of wood charcoal, wood, bark, and conifer needles were collected from Strata 2, 3, and 4 of the trench, as indicated in figure 1.7, and submitted for species identification and AMS (accelerator mass spectrometer) dating to the PaleoResearch Institute (in Boulder, Colorado). All samples, including those from the base of the deposit, were identified as Picea (spruce) except for one fragment of Populus (aspen or cottonwood) (Kováčik and Cummings 2015).

AMS dates were run on nine of these samples and fell into two groups, with the oldest seven ranging between 371 ± 23 RCYBP and 310 ± 24 RCYBP (table 1.1; figure 1.8). Six of the samples in this group were from Stratum 4 and one was anomalously from Stratum 2. These estimates intersect a plateau in the dendrochronological calibration curve and so have bimodal calendrical distributions. The earlier ranges start in the mid- to late 1400s and the later ranges in the early 1500s to early 1600s (figure 1.8). The early peaks are close to the deglaciation of Knight Island about six hundred years ago and older than the local availability of spruce, so are implausible on that basis. The younger peaks are quite similar to each other without the variability that random growth ring sampling of centuries-old trees (the “old wood problem”) would produce, so it is likely that the source trees were young, first-growth spruce that germinated around AD 1500 or later. Overall, this set of seven dates identifies the mid-1500s to mid-1600s as the most likely span of deposition represented by Stratum 4.

The second group includes two dates from Stratum 3: 234 ± 26 RCYBP on a conifer needle and 145 ± 24 RCYBP on spruce charcoal. The former is the most reliable because needles stop growing and absorbing atmospheric carbon after only five to seven years, avoiding the old wood problem. The earliest two calibrated date ranges for the needle (AD 1630–1690 and AD 1730–1750) are the most reliable, since there is no evidence at Tlákw.aan of Spanish or Russian contact. Therefore, the most likely dates for accumulation of Stratum 3 are the late 1600s to early 1700s, with no indication of a temporal gap between Stratum 3 and Stratum 4 below.

SOURCE KOVÁČIK AND CUMMINGS 2015.

In sum, radiocarbon dates from the 2014 test trench in Mound B near House 1 suggest that Tlákw.aan was inhabited by the mid-1500s and that people lived there through the 1600s and into the early 1700s. These results would push the beginning of occupation back more than a century before the AD 1690 estimate inferred from Maggie Harry’s genealogical account.

Another time marker embedded in the migration story suggests an even earlier date for the founding of the settlement. In episode 8d, Knight Island is described as having been so recently deglaciated that it lacked trees and was covered with strawberries, a stage of plant succession that would probably have obtained no later than the early to mid-1400s (Barclay, Calkin, and Wiles 2001). This oral evidence does not appear to be consistent with paleobotanical data from the 2014 test trench, where spruce—a species that does not take hold for at least a century after deglaciation—occurs from top to bottom and where no early succession tree species such as willow and alder were found. The likely explanation is that Mound B was not formed during the earliest phase of occupation at the site, as also supposed by de Laguna, who thought that House 7 and associated Mounds C and D might be older (de Laguna et al. 1964, 85). Thus, while the exact date for the founding of Tlákw.aan remains unresolved, a date somewhere in the mid-fifteenth century seems possible.

Projectiles. Copper arrow points from Tlákw.aan (n = 5) have leaf-shaped blades, sloping shoulders, and narrow, pointed tangs (figure 1.9A–C). They are closely comparable to leaf-shaped points from Ahtna sites, including GUL-00077, which dates from approximately AD 925 to 1485 (Hanson 2008, Figure 9; Workman 1977) and the early nineteenth-century Dakah De’Nin’s Village (Shinkwin 1979, Figure 10). Trace element analyses of two of the Tlákw.aan points (Cooper et al. 2008; Veakis 1979) indicated that the metal probably came from a Chitina River source.

Large ground slate endblades for lances (n = 3; figure 1.9D) and smaller slate endblades for arrows (n = 6; figure 1.9E) are relatively uncommon at Tlákw.aan in comparison to late prehistoric Eyak (Davis 1996, 466–471, and Figures 95 and 96) and Sugpiaq sites (e.g., de Laguna 1956, 1975; Clark 1974; Crowell and Mann 1998; Knecht 1995). Workman (1977) noted that among the Ahtna, copper replaced stone and bone as the material used for many types of tools.

Unilaterally barbed arrow points with conical tangs made of bone or antler (n = 19; figure 1.9H–J; de Laguna et al. 1964, Figure 17) are an archaeologically documented Ahtna type (Hanson 2008, Figure 16; Shinkwin 1979, Figures 13D and 14D; VanStone 1955), although also widespread during the past millennium among most Alaskan Athabascan and Inuit groups as well as the Tlingit. Unilaterally barbed bone harpoon heads with tapered tangs and line holes, used for taking seals, dolphins, and sea lions (n = 12, figure 1.9F), reflect the maritime focus of the Tlákw.aan subsistence economy and were used by all southern Alaskan coastal groups including the Tlingit (de Laguna 1960), Eyak (Birket-Smith and de Laguna 1938), and Chugach (de Laguna 1956), but not by the Ahtna in their original inland territory. Small, barbed harpoon arrowheads for sea otters (n = 9; figure 1.9G) had a similar cultural distribution. Pointed bone pieces with flattened sides, identified by de Laguna as gaff hook points but which actually served as barbs for halibut hooks (n = 5, figure 1.9K), represent another facet of coastal adaptation.

Woodworking tools. Splitting adzes (n = 14, figure 1.10A and B) made of pecked and ground greenstone or schist with hafting knobs or grooves were broadly distributed after 1000 BP across southeastern and southern Alaska as far west as Kodiak Island, but have not been reported for the Copper River Ahtna. Other Tlákw.aan woodworking tools include stone planing adzes (n = 13, figure 1.10C), stone chisels (n = 76, figure 1.10D and E), and beaver or porcupine teeth (n = 13, figure 1.10G and H) used as carving knives. These types have been found at Ahtna sites (Rainey 1939; Shinkwin 1979; Workman 1977), Eyak sites on the Yakutat foreland (Davis 1996), and in Prince William Sound (de Laguna 1956). Nails and small fragments of iron, almost certainly derived from driftwood or wreckage (n = 19; not illustrated), were probably used as bits for wood and bone-working tools (de Laguna et al. 1964, 88–90).

Cutting and scraping implements for skin and meat. The Tlákw.aan assemblage includes paddle-shaped scrapers made of flaked slate or schist (n = 5, figure 1.10F) as well as semi-lunar slate scrapers (n = 7, figure 1.10I), both comparable to Chugach types (de Laguna 1956, 131–135) and used for hide preparation. Boulder spall scrapers (n = 6, figure 1.10J) used for preparing skins are a predominant artifact in the Ahtna region (Hanson 2008, 122–123; Ketz 1983, 174–175, 187–188; Shinkwin 1979, 61–62; Workman 1977) and occur in other Athabascan, Eyak, and Sugpiaq areas.

The most distinctively Ahtna cutting tools from Tlákw.aan are semi-lunar knives with wooden handles and crescentic copper blades (n = 9, figure 1.11A and B) used for slicing salmon and other fish, a type that is duplicated at the pre-contact GUL-00077 site (Hanson 2008, Figure 11). There is also a unique Tlákw.aan copper semi-lunar knife with a grass-wrapped tang (figure 1.11C).

Lamps. Undecorated oil lamps (n = 51, figure 1.12A–C) hollowed from limestone, basalt, and other rocks are abundant in the Tlákw.aan collection and were used to light house interiors. Stone lamps for burning sea mammal oil are a coastal trait unknown in Ahtna territory. Stone lamps are rare in Tlingit collections, although a few were found at the Daax Haat Kanadaa site near Angoon (de Laguna 1960). They were universally used by other Alaskan coastal peoples, including the Eyak (Davis 1996, 490–496), Chugach (de Laguna 1956, 143–146), and Kodiak Island Sugpiaq (Clark 1984).

Ornaments. Tlákw.aan ornaments made of native copper included bracelets (n = 6, figure 1.13A), rings (n = 4, figure 1.13B), coiled wire beads (n = 2, figure 1.13C), tinkler cones (n = 4, figure 1.13D), and pins (n = 4, figure 1.13E). The rings and cones have close parallels among the pre-contact Ahtna copper jewelry from GUL-00077 (Hanson 2008, Figures 12 and 13). The Eyak and Sugpiaq also wore copper ornaments and examples have been discovered at archaeological sites, including a bracelet from Kachemak Bay (de Laguna 1975, Plate 49-10). These pieces may have represented the wealth and high status of Tlákw.aan’s inhabitants, as recalled in oral tradition. They also wore coal beads (n = 36, figure 1.13F–H), possibly made from coal collected at seams along Esker Creek on the west side of Yakutat Bay. Unfinished beads and coal fragments suggest on-site manufacture. Holes through the beads are straight (figure 1.13G), suggesting possession of metal drill bits. Coal beads were common in Prince William Sound, Cook Inlet, and elsewhere on Alaska’s southern coast.

Overall, artifacts from the site appear to be a combination of Ahtna implements and Eyak, Sugpiaq, or Tlingit types, the latter primarily related to maritime hunting and fishing. This result, while not definitive, is consistent with the combined Ahtna-Eyak origin of the migrants as described in oral tradition and with indications (episode 10) that the Ahtna adjusted to coastal life by adopting maritime technologies (such as sea mammal harpoons, ground slate knives, stone lance points, halibut hooks, and seal oil lamps) from their Eyak affines.

Distinctive Ahtna identity at Tlákw.aan is most strikingly represented by diverse native copper items (arrow points, knives, earrings, bracelets, rings), which have only rare Eyak, Tlingit, or Sugpiaq counterparts. Copper artifacts at Tlákw.aan link the site’s inhabitants to their Copper River homeland, and the relatively high frequency of these items in the upper levels of the site (de Laguna et al. 1964, 87–88) indicates that a trade connection must have been maintained long after the original migration (Pratt 1998). No direct references to post-migration contact with the Copper River region appear in the oral narratives except the incident in which six Gineix Kwáan brothers return home to retrieve a copper tináa for the purchase of Knight Island (episode 9). Perhaps this element of the story has a larger meaning, signifying an enduring connection between Yakutat Bay and Lower Ahtna territory on the Copper River. Sarah William’s statement that the Gineix Kwáan “used copper for everything—for knives, whenever they had a war” (episode 9c) might refer not just to copper items the migrants were able to bring with them on their original journey but to continuing importation and use. Finally, it is notable that YAK-00007 is the only site in Yakutat Bay where pre-contact copper artifacts have been found, supporting its identification as Tlákw.aan village.

According to the migration narrative, the residents of Tlákw.aan fished for salmon (episode 9), hunted harbour seals at their glacial rookery (episodes 10 and 10a), and took terrestrial game including bears and Dall sheep (episode 10). At YAK-00007, where animal bones are well-preserved due to the buffering of soil acidity by calcium carbonate from marine shells, these and other species should be represented.

Species identifications of faunal remains from the 2014 trench (Etnier 2017) are generally consistent with earlier findings from the site (de Laguna et al. 1964, 77–84). The total number of identified specimens (NISP) from all strata was 10,632 (table 1.2). Fish remains (NISP = 6,669, or 63% of the total assemblage) were dominated by salmon, with the addition of a few cod and dogfish. Mammals (NISP = 3,255, or 30%) were predominantly marine species, including harbour seals, fur seals, dolphins, sea lions, and sea otters. Harbour seals (NISP = 1,044) were by far the most numerous. There was a minor representation of terrestrial animals (deer, artiodactyl [possible mountain goat], black bear, beaver, or porcupine, NISP = 40, or 0.4%) and of birds (NISP = 18, or 0.2%).

The age distribution of the harbour seal remains gives a strong indication of rookery harvesting, with a concentration on the hunting of pups or a practice of bringing these smaller animals back to the village site as whole carcasses. Harbour seal specimens of determinable age (n = 173) included 28 newborn pups (16%), 27 transitional pups up to 2 months old (16%), 27 weaned pups up to 7 months old (16%), 52 subadults (30%), and 39 adults (22.5%) (Etnier 2017).

The faunal assemblage verifies oral accounts of subsistence practices at Tlákw.aan and indicates that hunting and fishing extended from spring (harbour seal, shellfish, sea otter, and fur seal) through summer (salmon, sea mammals) and into fall (mountain sheep or goat, black bear).

SOURCE: ETNIER 2017.