A more complete version of this report is available through the Society for Pennsylvania Archaeology: Prehistoric Occupations at Fishbasket Vol 69: No 1 (Spring 1999).

“Fishbasket” is a local place name referring to a tract of wide primary terraces bordering both banks of the Redbank Creek near its confluence with Town Run approximately 2 km (1.4 miles) above the town of New Bethlehem Pennsylvania. The first known documentation of the site name appears in the late nineteenth century during a flurry of formal historical publications about western Pennsylvania and the individual counties. The histories of Clarion and Armstrong Counties both refer to "Fishbasket" or "Old Town" as it is also known as located at the junction of Town Run and the Redbank Creek.

The History of Armstrong County (1914:2) records that "The Indians had several villages in this section when the settlers came, even occupying them some years afterward. Besides the famous one at Kittanning, there was another "Old Town" in Redbank Township, known to be a large settlement in 1770. In the same publication reference is made to this site as "a place called by the Indians "Fishbasket" from the immense schools of fish that were found there". The History of Clarion County (1887:60) also notes that "Old Town" was a village of the Shawnee. The Pioneer History of Western Pennsylvania (McKnight 1905:36) extends the last purported occupation well past the first white settlements into this area by noting a very late village of approximately three hundred Indians here briefly in 1816.

Between 1976 and 2000, the author has conducted independent long-term excavations at two adjacent Late Woodland sites on opposite banks of the Redbank at this locale. Fishbasket (36Ar134) was designated to the site on the Southern bank in Armstrong County, while Fishbasket North (36CL93) was referred to the location directly across the creek in Clarion County.

Indians Fishbasket by Larry Smail

FISHBASKET NORTH (36CL93)

Fishbasket North is located today in the center of a large cornfield on a long sandy rise which is situated between an existing springlet and the creek. A total of 5250 square meters have now been excavated at this location exposing portions of 3 distinct and isolated occupations.

Locus 1

Locus 1 was designated to a single long-term intensive occupation enclosed by a weak palisade line 54 meters (180 ft) in diameter that was constructed with posts averaging 10 cm (4 in)in diameter and spaced up to 62 cm (25 in) apart. No indication exists of additional uprights, a supportive embankment or complex protective entryways and while numerous wide gaps exist in the post alignment it has been impossible to determine specific entrances.

Positioned in relative nearness (within 7.2 meters (8 ft) ) around the interior periphery of the palisade, at least 11 identifiable round houses encircle a large open plaza. House diameters average 16.2 meters (18 ft) and were built with a framework of poles averaging 8.5 cm (3 ½ in) in diameter and spaced up to 30 cm (12 in) apart in an inward sloping manner. Interior features usually include a centralized hearth and an occasional basin or small bell shaped storage pit. Interestingly, the only other fired features at this site are numerous small basin shaped smudge pits found scattered across the village each containing large amounts of charred corncobs probably burned to produce smoke as a means of warding off the summer insect swarms.

Feature Type Profiles

Many interesting theories have been presented regarding the utilization of post enclosed pits including: meat smoking, dry food storage, dog pens, sweat lodges, and even survival huts. While all may be somewhat valid, they do not appear to fit the 25 examples found here as oval pits averaging 1.75 meters (69 in)in diameter with flat bottoms up to 22.5 cm (9 in) deep. These structures used the same diameter posts as noted in house construction. All are in close proximity, but not attached to houses, with entryways placed in an open avenue of convenience rather than to utilize the prevailing wind as needed to operate a meat-curing smoker. Ten were lined with grass and none were found to be associated with an interior or exterior associated hearth. A plausible application for these huts is their use to cache large stores of frozen meat procured in early winter when permanent low temperatures would easily sustain a frozen preservation throughout the cold months. Snow banked over these small domes would help create insulating igloos keeping the interior temperatures well below freezing even on occasional sunny days. These pits usually contained non-stratified deposits indicative of rapid refilling once abandoned. Many were destroyed by fire – possibly an easy method of neutralizing a rancid vermin infested structure produced by spring warming of raw flesh residue. Reutilization of the same pit is frequent apparently relating to a series of annual destruction and rebuilding events.

Generally located interior to the house ring, bordering and often intruding onto the plaza were eighteen large deep storage pits. These features averaged 112.5 cm (44 ½ in) in diameter and were up to 80 cm (31 ½ in ) deep with straight or slightly insloping profiles. Some had been plastered with a thick layer of prepared clay, and three contained large flat stone slabs used as a “basement” pavement. Flotation of these pits often produced quantities of maize, occasionally beans and a variety of nut hulls. The contents of these pits were probably drawn upon communally. By opening only one cache at a time, the other still sealed features could be maintained intact longer in order to insure a better state of non-disrupted preservation. This is also implied by their use when abandoned as village-wide garbage receptacles that frequently contain large concentrations of artifacts from which ceramic sherds are often refitted with others found in features specific to multiple identifiable households. While wigwam shaped houses and the use of freestanding post enclosed storage pits share an association with the Monongahela of southwest Pennsylvania, deep cylindrical storage pits are definitely not sites of that culture. These subterranean root cellars were distinctly used among the northern Allegheny Valley Iroquois, Owasco and Clemsons Island cultures.

Four human internments were encountered within the village confines. Two adults and one child were buried outside of, but nearby possibly related houses, while a third adult was located within a house. All of the adults were tightly flexed on their left sides in a nearly uniform body orientation in an east – west alignment with the face to the south possibly indicative of a ceremonial custom. Pathophysiological determinations were impossible due to extensive bone destruction by the soils high acidity. Accompanying grave goods were not encountered with the adult burials. During the excavation of each however, a thin red stain containing small granules of hematite and a single faceted hematite core was recovered in the grave fill several inches above each of the bodies. The occurrence of ceremonially deposited hematite was conspicuously absent within the child burial. It did however contain a single grave offering of a small limestone tempered cordmarked pot decorated with cordwrapped stick impressions below the rim.

While most of the village area adjacent to the stockade was a crowded jumble of houses, storage structures, racks and subsurface pits, the large open central plaza was doubtlessly where group social activities were held. This nearly featureless plot is broken by a dominant round house pattern and a remarkable group of pits. The house is located on the northwest plaza margin and is notably out of position to the ring of other dwellings. With a diameter of 8.4 m (28 ft) it is the largest structure on the site and is constructed differently with straight or slightly outsloping postmolds. It has an apparent doorway open to the south. A central firepit and other internal features that would normally be associated with a family-oriented lodge are absent and seem to indicate that this building was possibly used for communal purposes. Adjacent to this house and at the geographic center of the village was a large area of fire reddened subsoil that encompassed a cluster of six features grouped around a large postmold. This postmold measured 35 cm (14 in) in diameter and was apparently hand dug to a depth of 40 cm (16 in). It still contained several large stones that served as support wedges. The surrounding features were small non-fired pits, of which three contained a large plow truncated ceramic pot.

Typical artifacts from this village include Madison and a minority of Waterford, Speidel and Serrated triangular projectile points along with numerous T-base and ovate base drills, several large triangular knives, spokeshaves, bifacial abraders and ryan notched knives, which are earlier period projectiles, and knives found and reworked by the latter occupants. All chipped stone tools were either made of Onondaga or Huronian glacial pebble cherts from the Allegheny River basin or Vanport Siliceous Shale (aka Jefferson County chert), which is found as a bedded sedimentary formation in the headwaters of the Redbank. Other artifacts include several celt fragments, partial discoidals, notched netsinkers, and siltstone hoes. Items of personnel use and adornment consisted of undecorated obtuse angle and platform pipe fragments of stone and clay, pendant fragments, and beads of stone and clay.

Ceramics were abundant at this site with limestone tempering predominate. Stylistically most of these vessels have elongated globular bodies with slight necks and rims that vary between straight, slightly flared and occasionally extremely outflaring. Surface finish often varies on the same vessel relative to the amount of burnishing of randomly applied cordmarking prior to the pot being fired. As a rule the neck is almost completely smoothed while some remnant or occasionally unaltered cordmarking remains on the shoulder and body. Decoration is present on only a small percentage of sherds usually consisting of cordwrapped stick impressions, punctuations, or limited zonal incising executed obliquely or transversely on the neck or rim. Appendages are found as rounded castellations often formed as double, attached lugs both decorated and undecorated and at least one example of a strap handle.

While similar to Monongahela trends, the Fishbasket Ceramics maintain a distinct local flair that makes the sample ostensibly distinct enough to merit a new classification that includes Fishbasket Plain, Fishbasket Cordmarked, Fishbasket Cord Impressed, Fishbasket Incised and Fishbasket Punctuate varieties.

This locus produced a series of four dates positioned between 1039 AD + 162 years and 1100 AD + 70 years. Abandonment possibly resulted from a prehistoric submergence as evidenced by multiple features and numerous postmolds containing a fill of sterile sand. Just such and event was recreated on July 19, 1996 when a major flood inundated the site with 3.4 meters of swift running water causing both extensive erosion in some areas and deposition of up to 50 cm of new sediment on others.

Locus 2

In 1984 a second village referred to here as Locus #2 was accidentally encountered during the perimeter expansion of Locus #1. This occupation had not been previously identified by surface survey and the heavy stockade line suddenly encountered veering in the opposite direction was quite unexpected. While the investigation of this locus remains incomplete it presents a settlement pattern dramatically different from Locus #1.

This village maintained a well-constructed and heavily defensible palisade approximately 49 meters (161) in diameter built with large 25 cm (9 in) posts anchored up to 60 cm (24 in)deep into the subsoil. While no evidence for a supportive earthen embankment exists, the walls appear to have been bolstered by large angled braces spaced approximately 3.6 meters (12 ft) apart and up to 3 meters (10 ft) from the palisade wall.

One distinct entry was located on the southeastern section of the palisade facing the largest level and probably most open expanse on the terrace. This elaborate opening was built using a 3.6 meter (12 ft) long screen to create a constricted access requiring single file entry through a narrow channeled corridor less than .9 meters (3 ft) wide. Individuals passing through would also have traveled beneath an elevated rectangular platform about 4.5 meters (15 ft) long and 1.8 meters (6 ft) wide constructed over two upright walls perpendicular to the inside of the palisade. This bastion was probably sufficiently high to overlook the palisade wall, and would have been large enough to comfortably support several individuals. Additional areas of concentrated post placements suggest several additional non-entry platforms at intervals along the palisade interior. Such bastions are not unknown for Western Pennsylvania. Richard George (1982) noted a similar covered entryway configuration at the Gnagy site which is a Drew phase Monongahela village in Somerset County Pennsylvania.

By projecting the identified roundhouses into unexplored areas, a total of six or seven were probably present. They are similar in size and form to those found at Locus 1, but appear to be positioned to create an increased margin of safety by being up to 6 meters from the palisade wall.

In addition to the intensified protective stance of this village, an apparent cultural shift in feature usage also appears with the discontinuance of deep storage pits, large post enclosed storage pits and possibly the absence of burials within the village confines. Most of the features are very small and less than 6 cm deep. These include three freestanding post-enclosed storage structures that are dramatically smaller and much narrower than those at locus 1. Three very large roasting pits located interior to the house ring and well away from any structures may indicate a utilization of different food resources or processing methods. Each of these pits contained over 180 Kg (400 lbs) of intensively fired river cobbles, but were totally devoid of charred bone or cultigens suggesting that whatever was roasted or dried over these beds was completely prepared in another location.

Artifacts were remarkably absent as a result of the shallow features being almost totally eradicated by plowing and possibly a short occupation span. Recovered from only 11 features 3 chipped stone tools included Madison style triangular projectile points and T-drills along with quantities of limestone tempered pottery sherds, which include rims decorated with cordwrapped stick impressions or punch and drag incising executed uniformly below the rim.

Two dates of 1321 AD + 68 years and 1283 AD + 126 years were obtained for this occupation. Fire burnt subsoil and charred post remains made it conspicuous that a large section of the northern palisade and at least 1 house had been destroyed by fire. Rebuilding is not evident, possibly indicating an event of deliberate aggression or intentional firing likely at the time village of abandonment.

Locus 3

In 1987 the excavation area was again expanded to encompass yet another distinct occupation of a third palisaded village which has now been completely excavated. This village is similar in size to Locus 2 with the enclosure measuring 31.81meters (109 ft in diameter. Enclosed within are several identifiable round houses and 30 features.

Notably absent from this locus are the large post enclosed storage pits which appear to have been replaced by deep large bell shaped storage pits which average 57.5 cm (22 in) in diameter by 82.5 cm (32 3/1 in)deep but one, the largest feature yet excavated at these sites measured 1.72 meters (68 in) in diameter by 117 cm (46 in)deep. Most of these are clay lined and several were found with the clay caps intact.

Pottery at this locus resembles the Fishbasket Plain and Cordmarked varieties from Locus 1 and 2 but also found was a substantial quantity of Grit-tempered Owasco like ceramics including Carpenter Brook Cord-on-Cord horizontal Levanna Cord on Cord, and Seiferd Heringbone.

An unusual pot from this locus was dated at 1270 AD + 50 years. It appears to be a stylistic copy of an early Iroquoian or Owasco vessel, but was locally made with limestone tempering and lacking any form of decoration

FISHBASKET (36AR134)

The Fishbasket site located directly across the creek for the most part mirrors the structures, features and artifacts recovered from the Fishbasket North site with the exception that here the multiple occupations are superimposed upon each other which make the intersettlement relationships much more difficult to discern. Approximately 2/3 of the estimated 5850 square meter occupational area of this site is restricted by adjoining properties and a service road bordering the creekbank. Our excavations here were conducted early from 1976-1982 and totaled approximately 1278 square meters.

The dominant occupation at this site consists of at least 2 partially explored palisade lines, 12 round houses, and 159 features that include large freestanding post enclosed storage pits and deep cylindrical storage pits paralleling the Locus 1 occupation at Fishbasket North.

Human remains include 1 distal half of an adult human femur found within a household firepit and a single well preserved tightly flexed adult male burial that again included no associated artifacts but hematite included within the fill as both as pigment and a single faceted core.. Although the absence of accompanying grave articles is common to adult burials throughout this area, the inclusion of hematite probably represents a tradition carried forward from the Middle Woodland Period. While utilitarian artifacts made of hematite are frequently found in Monongahela site inventories, faceted hematite nodules are rarely encountered. A survey of Monongahela mortuary patterns conducted by Christine Davis in 1984 failed to note a single instance of funerary association. Hematite however is fairly common as both faceted cores and powder covered objects accompanying Iroquois related burials through the early historic period.

Within this plaza again was a grouping of 14 shallow basin shaped pits centered around a large 40 cm posthole. Seven of these contained large pots each exhibiting a uniquely different stylistic and decorative motif within the earlier described Fishbasket ceramic series. This practice has never been previously identified through archaeology at any other location known to the presenter and strongly suggests their internment as part of a community ceremonial feasting function such as the Green Corn Ceremony which was practiced by many of the eastern native American groups at the beginning of the fall harvest.

Also at this site was an aberrant occupation consisting of a single longhouse measuring 12.6 meters (41 ½ feet) long and 5.4 meters (18 feet) wide oriented in a North to South alignment with an obvious entry in the southern end. Pottery sherds recovered from the interior hearth and bell shaped storage pits suggest and Owasco affiliation. No other structures or features discovered here seem to relate to this single isolated component. While Clemsons Island /Owasco ceramics have been reported by George and others at some Monongahela villages in Southwestern PA, the presence of longhouses at each of these sites represents the southernmost penetration of such structures yet reported into the Allegheny River valley.

Historic Cabin

Near the western edge of the excavation an intriguing feature was found that appears to be the remains of a cabin. It consisted of a nearly rectangular pit measuring 3.6 m (12 ft) by 3 m (10 ft), that was dug 15 cm (6 in) deep into the subsoil. Five large round upright posts had been placed at each corner for upright supports, and on the southwest wall to presumably frame a doorway. The soil of this feature was very compact from apparently having served as a living floor. This floor was overlain by a deep 5 cm (2 in) layer of large charcoal fragments, possibly pointing to a destruction of this building by fire. While notably absent were nails, or household activity items including fragments of glass or glazed ceramics, the artifacts relate to the late 18th or early 19th century and possibly point the historic Shawnee or an early European occupation.

ASSOCIATIONS

The sites comprising the Fishbasket complex represent at unique ceramic producing multi-locus that appears to share a blending of traits and traditions acquired from neighboring cultures while also maintaining some very distinct attributes. The recent nearby discoveries of at least three additional late woodland villages containing Fishbasket series ceramics indicate that this temporally different but probably related grouping of villages is extensive and unparalleled elsewhere within the Redbank watershed. The importance of this locality to village placement appears to be its strategic positioning on a principal ford of the Frankstown-Venango path on the Redbank. This location is also the probable intersection point of the Punxsutawney-Venango and Kittanning-Venango paths. Having a permanent settlement situated adjacent such an accessible well traveled route would have been extremely advantageous to any group wishing to promote contact or control within the valley and region..

This pattern of apparently isolated villages is also found at other fords of the Frankstown-Venango path. Immediately to the south 36In86 is located at the ford of the confluence of the Little and Big Mahoning Creeks, and the Kimmel site is near the next ford crossing at Crooked Creek To the north it crosses the Clarion River at the State Road Ripple site.

The recovery of a geometric serpentine bead within a feature at the Fishbasket site provides a possible intriguing connection to these early trailways. Currently two additional specimens have been uncovered in archaeological contexts that are nearly uniform in size, shape and material. These beads are so unique that they suggest manufacture by the same individual or related group. One was associated with burial #5 at the Fort Louden site in Franklin County dated to 1220 AD by James Herbstritt (Herbstritt 1982), while the other was recovered at a contemporaneous Shenks Ferry component at the Locust Grove site in Lancaster County. Both of these sites are positioned on direct extensions of the Frankston-Venango trail, but are separated from Fishbasket by 157 and 267 Kilometers. The implication of these nearly duplicate unusual beads having been transported so far apart along this paths main route, as well as the presence of ceramics and settlement traditions indicative of distinct cultural influences lends credence to a long prehistoric existence of this major trailway system.

ACKNOWLEDGEMENTS

The many years of labor for this undertaking was carried out by a volunteer group of family, friends and local interested volunteers mostly members of the SPA chapter 22. Several, two-week field schools were also conducted for local students from the Redbank Valley High School under the teacher leadership of Mike Masler.

As the first recipient of the Lousi A. Brennan Publications Award, I would like to take this opportunity to publicly thank both the Eastern States Archaeological Federation and the Society for Pennsylvania for their support and publication of the 1999 monograph on these sites. I would also like to express my appreciation to the Society for Pennsylvania Archaeology for providing partial grant funding to process the C-14 samples referred to within this paper.

PHMC Marker dedicated in July, 1997

2006 EXCAVATIONS

Numerous members of the Northfork Chapter 29 and other volunteers participated throughout the summer of 2006 helping to excavate and document numerous features and postmold patterns as part of an additional 3800 square foot investigation of the Fishbasket Site (36AR134) along the Redbank Creek in Armstrong county. The area explored was a continuation of the 1976-1980 excavations conducted out by Ken Burkett and is currently designated for destruction by impending construction of a sewage pumping station sometime of the next several years.

The largest feature documented was a post-enclosed storage structure that had apparently been burnt. The remaing subsurface feature contained a large amount of partially burnt posts and framework supports along with quantities of pottery and stone projectile points.

Large Post-Enclosed Storage pit

The fall issue of Westelvania Magazine contains a nice published article on the Fishbasket Site Excavations.

Excavation Overview	Excavation Overview
Removing plow disturbed topsoil	ED Kaufman Excavating a Feature
Brian Fritz Conduting Deep Soil Profile Testing	Ken and Brady Burkett Screening

Prehistoric Artifacts were numerous and included triangular projectile points, drills, pottery sherds, a clay pipe and rough stone tools including a netsinker, hoe, hammerstones and anvilstones.