LITNIK ARCHAEOLOGY AT THE TSUNAMI SITE
CONTINUED EXCAVATION IN 2003

Donald W. Clark
August, 2003

Prepared at Kodiak, Alaska, for the Native Village of Afognak

LITNIK ARCHAEOLOGY AT THE TSUNAMI SITE
CONTINUED EXCAVATION IN 2003

Description of the site and its surroundings is to be found in the report entitled “Litnik Archaeology at the Tsunami Site—2001,” placed on file in 2001.

Objectives
Modest objectives were set for 2003, inasmuch as the field season was to be only three weeks long and the crew to consist of just two persons, Jimmy Charliaga and Donald W. Clark, the writer. The Native Village of Afognak’s Dig Afognak field camp provided significant logistical support, including meals, lodging and transportation. That extended the capabilities of the crew. Additionally, on three days, staff and participants from the camp assisted with the excavation.

A large, shallow oval house depression was known at Litnik, near the crossing of the covered bridge, which was destroyed in 1964. This was to be examined briefly to determine which prehistoric people had occupied it.

The Tsunami site is located farther down along the tidal reach of the Afognak River. It had been excavated in 2001, but the season ended before the house entrance could be located and, apparently, part of one end wall remained to be uncovered. A second house depression lay near its southwest end. The next year, we found a pair of connected houses across the river at the Salmon Bend site. Thereupon, we suspected that the two Tsunami houses might also be connected. Double houses had not been investigated previously for the Kachemak culture. The resumed excavation addressed this question.

Work Accomplished

Methods
The excavation trench at the Tsunami site was only 1 meter wide and it was possible to cast out the dirt directly by shovel, thus eliminating double handling, i.e., the necessity to “bucket” the dirt out of the hole and away to the spoil pile. Depth of artifacts below the base of the Katmai ash was recorded in the catalog of noteworthy artifacts. At the conclusion of excavations, the pits were filled in. In about three years, the ground will become well revegetated.

Oval Depression
A 1.5m-long test pit in this feature revealed that ancient construction had cut into the underlying fine soil. After one day’s work only, a large chert flake, a cobble spall (boulder flake), and a scrap of slate had been recovered through screening. These probably belong to the Kachemak culture tradition of 900 to 3,800 years ago. It was felt that the results did not warrant further work here.

Site examination
Two early- and late-winter storm-driven high tides during 2002-2003 had caused erosion along many Kodiak beaches. We were concerned that sites in the Afognak Bay area might have been damaged. Foot traverses of the beaches and near-shore observations by boat showed that damage at eight inner Afognak Bay sites was negligible.

Tsunami Site
Work at the Tsunami site continued from July 21 to August 5. Exclusive of back-filling time, a day at the oval depression, and one day when the weather prevented work, the crew excavated 10.5 days. No shelter was erected over the excavation. There was only one day during which any shelter would have been of use.

Ten and a half one-square meter sections were excavated: 1.5m as an isolated unit and 8m in two intersecting narrow trenches. They were placed to abut the west edge of the 2001 excavation. Two persons from the Science Camp, Justin Kewan and Arthur May, assisted one day each. We did not make greater use of camp youth because they could not be accommodated very well within the small excavation and because previous experience had shown that youth shifts longer than 3 hours’ duration would not be practical.

Collections
Artifact recovery was modest. Implements are listed in the table below. Comparative data for the 2001 excavation also are given.

COLLECTIONS FROM THE TSUNAMI SITE.

2003 2001 Total
Adze bit 0 0 4 2% 4
Ulu blades 33 40% 50 20% 83
Whole (9) (19) 28 Fragment (24) 27% (31) 55
Cobble spall 11 13% 26 10% 37
Some are unused or fragments
Notched pebble sinker 8 10% 50 20% 58
Whetstone, shaft smoother 6 7% 30 12% 36
Pumice abrader, whetstone 2 2% 11 4% 13
Ground slate points 5 6% 12 5% 17
Flaked stone points 0 0 11 5% 11
Grooved cobble 2 2% 5 2% 7
Red stone for jewelry, beads 3 4% 5 2% 8
Labrets 0 0 5 2% 5
Flaked ulu or blank 6 7% 17 7% 13
Various slate pins 3 4% 3 1% 6
Double-edged slate knife 2 2% 7 3% 5
Lamp 1 1% 4 1.5% 5
Chert implements 0 0 10 4% 10
Chert flakes (37 500+) 537
Boulder maul 1 0 NC
Total w/o chert flakes, slate
scrap and rounded pebbles 82 250 332

In addition, there is one sawn-off scrap of slate – a piece of trim, and an ulu blade fragment scraped in the Ocean Bay II mode of fabrication. These items may be contamination from the adjacent 3,800-year-old terminal Ocean Bay-Kachemak transition site. They are out of place in the late Kachemak period.

There is hardly anything on the 2003 list that was not also found in the 2001 Tsunami excavation. The artifact types and styles indicate an age during the first millennium A.D., which is late in the Kachemak tradition. One notable difference is the lack of flaked stone or chert points and other flaked artifacts in the 2003 collection. The ratio of chert flakes to other artifacts, as listed above, is less than 1:2 for 2003, whereas the ratio is the reverse, more than 2:1 for 2001. This figure, using the relatively abundant flakes, suggests that more is involved here than the vagaries of sampling. Abundance of ulu blades support the interpretation, based on the site’s location, that the primary activity there was catching and processing salmon.

Soils and Structural Details and Interpretation
In the 2001 excavation, there was evidence for a pre-tsunami house and later minor post-tsunami occupation. This model of occupation is compared with the data from the 2003 excavation.

The 2003 excavation encountered an old structure floor that had been excavated down to hard glacial till. It is thought to be part of a house, here called House 2. The old structure edge is clearly defined where it cut through fine sediment (probably volcanic ash) orange and yellow brown soil. Some of the orange soil was left at the river end of the excavation, showing that the floor did not reach glacial till everywhere. A similar substrate was found in 2001.

House-pit edges were reached at each end of the line of five sections excavated in 2003, showing that two sides of the structure had been reached. In the till floor, there were two post holes and three pits or basins. The smaller basin was covered by a slate slab. Inasmuch as the excavation was only one meter wide, insufficient floor was exposed for any pattern of features to be apparent. At one end, there also was a boulder floor that nearly reached the pre-Katmai surface. It has counterparts in boulder piles uncovered in 2001 at the corners of House 1, and probably was a support for a beam that would have run the length of the wall below roof level. There would have been additional higher level support for the roof.

Above the floor there was a thick layer of fine dark brown artifact-bearing soil about 20 cm thick. It probably is derived from occupation of the probable house. In Section 7, the till floor stepped down several cm, forming some kind of a pit or trough into which higher strata dipped.

Next in the upward progression, was a layer of gravelly soil 6 cm thick where pure gravel, and 30 cm where very soily. It consisted of almost pure angular gravel in places, elsewhere of very stony brown soil. It was not confined to filling in an old house pit. This till-like gravel appears to be spoil dug out from another structure pit and spread over the western part of the site. It may have come from the Tsunami House No 1. If that were the case, we have a construction sequence of House 2 and then House 1. But the preference for flaked chert in House 1 could be taken to mean that that house is the older one, positing increased use of ground slate through time.

The gravelly layer was overlain by a 40-50cm thick, less stony layer of artifact-bearing charcoal-streaked dark brown soil. A 5cm-thick layer of tsunami sand, after which the site is named, appeared near the top of this deposit. Usually, this layer was very consistent, even, and undisturbed. This suggests that the site was abandoned at the time of the tsunami, unlike the case for the area excavated in 2001.

Then there was 1 to 2 cm of topsoil under the 1912 volcanic ash. The shallow position of the tsunami layer near the top of the soil sequence further indicates that there was little or no occupation of the area investigated in 2003 after that event. But, in the section with the step-down or trough into till, Section 7, the tsunami layer takes a major plunge. It drops as a broken or discontinuous layer, and then, close to the base of the site (close to till), it thickens to a 10-15 cm depth. This disposition suggests that flood water had entered a structural pit where a much greater amount of sediment was able to settle out of the deeper water column.

It appeared at the end of the 2001 season that deposits displayed in the end-wall of the excavation were part of the fill of the Tsunami House 1. Orange and yellow volcanic ash found there could have been on the floor of either House 1 or 2. A step-down in the floor at the west wall, at first thought to be a feature of House 1, actually could have been the first floor layer of House 2, which had been built at a slightly lower depth than House 1. A post hole in the till floor of the Second house is very close to the edge of the limit of the 2001 floor, showing that the floors of the two houses essentially met and may even have overlapped. This is a distinct possibility if the two were joined, or if construction of one cut into the other house which by then had been abandoned. The presumed beam that spanned two rock piles exposed during the two seasons could have been part of the architecture of the juncture of both houses, if they were contemporary.

Exposure of the west end of House 1 may have been more complete than originally was thought to be the case. Simply stated, recovery of the edge of a house-pit and soils formed outside the house had been expected. They were not seen. But in retrospect, it is possible that the excavation of 2001 extended to the second house, which was filled with culturally derived deposits and thus presented a matrix similar to that of the House 1 fill. And if the two houses were joined by a wide opening, there would be no house-pit edge between the two. The beam capping this opening probably was 136-139 cm above the House 2 floor. This is the height of the boulder and rock pile noted earlier.

The following complex series of events is suggested for the area excavated in 2003: Excavation to till, in most places, of a pit for a house; accumulation of soil and refuse from occupation of the house, No.2, and then, when the house was abandoned, material from the caved roof, slumped walls, and from use of the pit to hold spoil from the construction of other house(s); excavation of another house pit, No. 3, immediately to the west, which was revealed in sections 6, 7 and Z, or at least intercepted, though not fully understood. The base of Section 7 and Section Z, which is thought to be on the House 3 floor, is 25-30 cm lower than the House 2 till floor. Spoil from this excavation further raised the site deposit that lay atop the remains of House 2. Then a tidal wave deposited about 6 cm of sediment over the site. At House 3, the structure was still open and standing, and possibly occupied. There, sediment-laden water entered the structure and left a thick deposit on the floor. Then the site was abandoned except for minor reoccupation over the locus of House 1. Some of these details are an interpretation, but there definitely was more than one house of Kachemak age at the site, and two houses were flooded by a tidal wave.

Not all the collections (218 cataloged items) are from the occupation of House 2. Many pieces were dumped into the depression of the abandoned house or are from refuse later spread over the area. This refuse came from occupation elsewhere on the site including, probably, Houses 1 and 3.

Conclusion
Evidence can be interpreted that construction of the Tsunami House No. 1 clipped the edge of an already-abandoned House 2. But the evidence also can be interpreted to show that the two are contemporary and joined by a wall opening at least two meters wide. But House 2 was built at a slightly lower depth. Artifacts suggest that the two are roughly contemporary, but chert flaking was favored in House 1. This should mean that House 1 is older, although other evidence suggests that House 2 is the older one. We note, too, that the House 1 penchant for stone slabs on the floor is poorly reflected in House 2. If the two were connected, it is possible that there were activities or activity areas peculiar to each.

Also, it appears that most of House 2 had been filled in and smoothed over, except in the one area described earlier as a pit, at the time of the tsunami, while House 1 was still a depression, and possibly a standing open house at that time, judging from the disposition of the tsunami deposit. This would indicate a younger age for House 1. House 3 overlay the western edge of House 2, and was at least 2.5 meters west of House 1. The limited excavation produced no evidence that those houses were connected. The topographic depression west of House 1, which was the objective of renewed investigation, actually appears to be a surviving expression of House 3. The occurrence of two-room Kachemak houses needs further investigation.

CATALOG 2003 AFG-215 TSUNAMI SITE, LITNIK