PUBLICATIONS :: NYSM RECORD :: Current Research in New York Archaeology

cover Current Research in New York Archaeology: A.D. 700–1300

edited by Christina B. Rieth and John P. Hart


Christina B. Rieth and L. Lewis Johnson

Abstract. Traditional models of Late Prehistoric (A.D. 7001300) interaction in the middle Hudson Valley suggest strong ties with contemporaneous groups in southern New England. Recent research, in the form of trace element analysis of lithic artifacts from the Trapps Gap site in Ulster County, New York, questions this assumption suggesting a more diverse and complex landscape in which groups interacted. This paper will discuss whether all of the samples were procured from the same deposit, how such procurement patterns may have changed over time, and what such data might reveal about regional behavior.

Since the beginning of the twentieth century, archaeologists working in the Shawangunk Mountain Region of New York have considered the region to be an enigma. Artifacts recovered from sites along the mountain’s ridge were often considered to be foreign in origin based on the visual color of the materials. Projectile points resembling types found in Pennsylvania, Delaware, and Virginia have been found at foraging and resource procurement stations in the region, leading archaeologists to suggest that the region was a travel corridor linking groups living in the Appalachian Highlands (see Eisenberg 1978, 1982; La Porta 1996; Schrabisch 1919a, 1919b; Sullivan and Prezzano 2001). Despite the presence of seemingly “foreign materials” recovered in the region, the Shawangunk Ridge overlooks a large flat plain of the Wallkill and Hudson Rivers. Late Prehistoric settlements dating between A.D. 700 and 1300 are found across the plain and scouting parties and family groups crossing into highland areas might also account for the deposition of such materials within cave and rockshelter sites in the region (see Diamond 1995:22-23; Eisenberg 1974:41–42; La Porta 1996:73–83).

Compounding this issue is an absence of archaeometric studies which have actually examined the materials recovered from sites in the region. Instead, most studies have relied solely on visual inspection of the materials to make determinations about the local and non-local procurement of these materials. As Calogero (1992:87–90) and others (Kuhn and Lanford 1987: 57–69; La Porta 1996:73–83) point out, visual inspection of materials often leads to erroneous assumptions about the origin and procurement of lithic artifacts.

In an effort to shed light on this issue, researchers from the New York State Museum and the Department of Anthropology at Vassar College conducted trace element analysis on lithic artifacts from the Trapps Gap site, located in Ulster County, New York. The Trapps Gap site is a small rock shelter site located where Routes 44 and 55 cross the Trapps Gap. The use of this pass through the ridge extends back into prehistory. Excavation of the site by Vassar College in 2007 sought to test the hypothesis that the site was used in the prehistoric period by travelers resting for the night on their way across the ridge. This hypothesis led to a subsidiary hypothesis, reported in this chapter, that the lithic materials found in the site would reflect forays along the ridge and that we would find materials that originated from several different outcrops. The subsidiary hypothesis was tested through x-ray fluorescence analysis of flakes from the Trapps Gap site.


Geography and Natural Environment of the Shawangunk Ridge

The Shawangunk Ridge forms the northernmost point of the Shawangunk Mountains in Ulster, Sullivan, and Orange Counties, New York. The Shawangunk Mountains are an extension of the Appalachian Mountains running northward from Virginia where they start as North Mountain. The Appalachians also cross the Kittatinney Mountains near the Delaware Water Gap in New Jersey and the Blue Mountains in eastern Pennsylvania. The Shawangunk Ridge is primarily formed of Shawangunk Conglomerate, a silica based conglomerate that is resistant to weathering. Underlying the Shawangunk Conglomerate are pockets of Martinsburg Shale, which are believed to have formed more than 400–450 million years ago (Snyder and Beard 1981; Sullivan and Prezzano 2001).

The Shawangunk Ridge is bordered on the east by a large plain that extends throughout the Wallkill Valley to the Hudson River. To the west and south, other waterways including the Roundout Creek, Neversink River, and the Delaware River, wind their way across the landscape. In addition to these waterways, several small swamps and wetlands are found within and surrounding the Shawangunks and provide a vast array of aquatic and avian resources. Among these is Rhododendron Swamp, believed to have formed about 15,000 years ago from a slow draining pro-glacial lake (Snyder and Beard 1991).

Eisenberg (1991:160) suggests that these waterways were present throughout prehistory and would have helped to sustain groups living both within mountain areas and on the neighboring plain. Today, more than one hundred different species occupy the ridge, including turkeys, ducks, and falcons, as well as white-tailed deer, fox, (box) turtle, woodchuck, beaver, river otter, and black bear. Most of these species would have been available throughout prehistory and may have been sought by native groups (Diamond 1995:21; Eisenberg 1974, 1991:162; Funk 1976; Ritchie 1994; Ritchie and Funk 1973; Schrabisch 1919a:146).

Many chert outcrops would have surrounded the Shawangunk Ridge with many high quality cherts found to the east in the Wallkill Valley (La Porta 1996:73–83), m. Most of these outcrops can be identified as being within the Ontelaunee Formation. Bifaces made from Harmonyvale chert as well as points dating to the Lamoka and Orient Fishtail types fashioned from Crooked Swamp chert have been identified in many of the same collections, suggesting that nearby groups may have shared information about procurement of materials and the whereabouts of specific quarries within the valley.

History of Archaeological Investigation on the Shawangunk Ridge
The collecting of Native artifacts from the ridge has probably been going on since the first Europeans settled in the highlands. Max Schrabisch, the first professional archaeologist to work on the ridge, located most if not all of the sites he recorded through conversations with local boys and collectors (1936:1–3). Schrabisch located and described shelters on the ridge, and recent surveys have located a number of other shelters, which do not seem to be the ones he investigated (Eisenberg 1991; Kaplan and Johnson 2009). Numerous other shelters undoubtedly exist in the Shawangunks, most known only to local amateur archaeologists.

Five shelters in the Shawangunks have been investigated by modern professional archaeologists: the Rhododendron Swamp/Mohonk and Ski Minne shelters by Leonard Eisenberg and students from SUNY New Paltz (Eisenberg 1991:159–176), and the Trapps Gap, Bonticou, and Burger rockshelters by teams from Vassar College under the direction of the junior author (Sando and Johnson 2008). Vassar students have also re-recorded and reanalyzed the Ski Minne collection (Kaplan and Johnson 2009). Rhododendron Swamp and Trapps Gap were investigated by Schrabisch, but Ski Minne and Bonticou do not seem to have been; the Burger Rockshelter may be Schrabisch’s Minnewaska shelter.

Judging from the projectile points found at these sites, the heaviest occupation of the ridge was in the mid- to late Archaic Period (ca. 6,000–1500 B.C.). This is true of many of the region’s sites. At Rhododendron Swamp, where the occupation was on the apron rather than under the overhang and was therefore missed by Schrabisch, and at Ski Minne, the overwhelming majority of the points were Late Archaic in date. Although most of the Burger Rockshelter had been dug out by previous amateur (or professional) investigators, two Lamoka points were found by archaeologists from Vassar College in and under the stone wall in the front of the shelter, suggesting that it, too, is Late Archaic (Ritchie 1971). No diagnostic materials were found in the Bonticou shelter. Eisenberg (1991:173–174) hypothesized, following Robert Funk, that the Shawangunk Ridge was used throughout the Archaic period by hunters who followed the deer into the highlands in the late fall and winter. None of these sites contained significant amounts of pottery, suggesting a limited Late Woodland presence. At these sites, Eisenberg found 130 sherds, all of which were collected from a single pot; there were 15 sherds in the Ski Minne collection, and no sherds were recovered from the Bonticou or Burger sites.

The Trapps Gap Rockshelter is very different from these other shelters. While it did have evidence of a mid- to late Archaic occupation, it also contained a large concentration of material dating to the Late Woodland, including triangular Madison projectile points and at least 15 different vessels (Schrabisch 1936; Sando and Johnson 2008). Since large quantities of pottery are generally associated with the presence of women in Northeastern Woodlands cultures, this suggests that women, not just male hunters, may have used the Trapps Gap shelter during the Late Woodland period. As discussed below, visual examination of the lithic artifacts from the Trapps Gap site (Figure 4.1), suggest that they were made from a wide variety of materials many of which may not have been local to the area, further strengthening Sando and Johnson’s (2008)  belief that the occupants of the Trapps Gap site were transient groups originating from other areas.

Figure 4.1

Photograph of projectile points recovered from the Trapps Gap site in Ulster County, New York. (Photograph provided by L. Lewis Johnson, Vassar College.)


The Vassar College fall archaeology field schools under the direction of L. Lewis Johnson excavated the Trapps Gap site in 2006 and 2007. The site is located along the east side of the ridge near a carriage trail leading to Rhododendron Swamp. It is located between two other predominant shelters: Ski Minne Rockshelter, which is located to the west, and Mohonk Shelter, which is located to the east (Figure 4.2).

Figure 4.2

Map showing the location of the Trapps Gap site and other sites located along the Shawangunk Ridge. (Key: 1=Ski Minne site, 2=Trapps Gap site, 3=Mohonk Rockshelter site.)

All three sites are located on the edge of the Shawangunk Ridge at an approximate elevation of 300 m above sea level (Eisenberg 1991:161). The Trapps Gap site is located between the deep, narrow Rondout Valley on the west and the open Wallkill Valley to the east. This position was likely an important factor in the site’s occupation and would have afforded its occupants a clear view across the Wallkill Valley with which to track animal migrations across the plain and other native groups upon approach.

The Trapps Gap site has three predominant occupations dating to the Middle Archaic (6,000–3,500 B.C.), Late Archaic (3,500–1500 B.C.), and Late Woodland (A.D. 800–Contact) periods as represented by Genesee, Brewerton, Lamoka, Snook Kill, and Madison projectile points (Ritchie 1971). Incised pottery sherds, resembling those found on Munsee sites in the Lower Hudson Valley, further suggest occupation of the site during the Late Woodland period. Seventeenth-century pipe fragments were also recovered suggesting a European presence at the site. The Late Archaic and Late Woodland period occupations are the most substantial and produced more than half of the lithic artifacts recovered from the site.

The artifact catalog from the Trapps Gap site consists largely of pieces of debitage recovered from living floor and feature contexts. Visual inspection of the materials suggests that much of it originated from quarries outside of the Shawangunk Mountains. Debitage fragments include pieces of locally available light and dark grey Onondaga chert and blue and green Normanskill chert as well as pieces of quartzite, and brown, white, semi-translucent grey and red chert. Several of these artifacts showed signs of intentional use-wear and may have been refashioned for use as expedient tools.

The range of materials recovered from the Trapps Gap site resembles assemblages recovered from nearby sites such as Mohonk Rockshelter. According to Eisenberg (1991:165–170), the artifact assemblage from Mohonk Rockshelter contained a wide diversity of materials, including black, light and dark grey, white, blue-green, brown, and red-colored cherts from deposits dating to the Archaic and Late Woodland periods. Pieces of jasper and chalcedony were also recovered. At the Ski Minne site, located to the west of Trapps Gap, artifacts recovered from the Middle Archaic, Late Archaic, and Late Woodland contexts also included cherts of various colors, including blue-grey, white, and light and dark grey (Kaplan and Johnson 2009).

The identification of lithic materials by visual inspection alone has long been practiced in Northeast archaeology. Calogero (1992:87–90), however, argues that visual inspection of materials alone often results in erroneous identifications of the origin of such materials. In a study of lithic materials from Connecticut, the vast majority of the materials were misidentified as to their origin when only visual inspection was used. The causes for the misidentification of materials included surficial weathering of materials as well as variation in how individual researchers identified materials. When other archaeometric and petrographic techniques were used, the inherent geologic features of the artifacts could be determined linking them to a specific region or possible outcrop.

The potential identification of “non-local materials” in utilized flakes and debitage, both of which are not traditionally traded, raises interesting questions about the movement of groups across the ridge (Brennan 1979). While Brennan (1979) does not speculate about the mechanism and activities that caused groups of people to move across the Shawangunk Ridge, Eisenberg (1991) suggests that the unique environmental characteristics of the feature, combined with the range of food and utilitarian materials that might have been accessible along the feature, could partly explain such movement.

In an effort to test whether all of the materials were from the same outcrops or from many different deposits, debitage recovered from living floor and feature contexts at the Trapps Gap site was subjected to x-ray fluorescence analysis. The remainder of this paper addresses the results of this research. Possible scenarios for the recovery of these materials and their role in understanding the settlement patterns of the region are provided.


X-ray Fluorescence Analysis

X-ray fluorescence analysis was used to collect trace element data on lithic samples from the Trapps Gap site. Assays were performed at the University at Albany, State University of New York. The instrumentation used functions on the basis of energy dispersive spectroscopy (EDS). During the decay of atoms, X-rays are emitted and the instrument directs the emitted radiation toward a target ring composed of Tin (Sn), Iron (Fe), and Copper (Cu). When exposed to the X-rays from the secondary target, the atoms of the sample fluoresce, or get excited. These x-rays are, in turn, emitted back toward the x-ray fluorescence instrument where they are detected by a silicon-lithium (Si-Li) detector.

As shown in Figure 4.3, the x-axis in the energy spectrum is used to identify individual elements based on their particular energies, while the y-axis is used to determine the amount of energy emitted by a particular element in the sample. The peak energy positions are compared to a database of known elemental x-ray fluorescence energies; when the peak is positively correlated to a known energy or wavelength, the element in question is identified. The size of the peak is determined by the amount of x-ray radiation (in count units) received by the detector at the defined wavelength or energy. This in turn indicates the elemental concentration of the sample, where larger proportions of an element would produce proportionately larger peaks.

Figure 4.3

XRF spectra showing channel counts for lithic artifact (Artifact # 35) recovered from the Trapps Gap site. The horizontal (x) axis represents the channel number while the vertical (y) axis represents the counts of each element per artifact.

The techniques used to examine the artifacts from the Trapps Gap site allowed proportional data about the relationship of the elements to be collected. As discussed in Kuhn (1986), proportional data allow ratios of trace elements to be measured in terms of “the number of characteristic x-rays observed in fixed time.” Following bombardment of the sample with a Cadmium 109 radioisotope for one hour, fluorescent x-rays were emitted whose energies are characteristic of the elements present in the artifact. The energy levels of each element were recorded numerically, allowing concentrations of specific elements to be determined. Sixteen elements were measured including rubidium (Rb), strontium (Sr), iron (Fe), lead (Pb), vanadium (V), potassium (K), zinc (Zn), nickel (Ni), barium (Ba), yttrium (Y), titanium (Ti), scandium (Sc), manganese (Mn), copper (Cu), and zirconium (Zr). Vanadium (V) and potassium (K) were removed from the final analysis due to low detection limits. Although the number and types of elements used in this study are sufficient for distinguishing between different samples, in most cases, additional samples and elements are needed to identify the precise location of a specific outcrop. Since this was beyond the scope of this project, the results are only discussed in terms of differences in the samples.

The recorded energy levels of each flake were collected and stored for analysis using the computer program AXIL (Van Espen n.d.). AXIL was applied to measure peak area counts and standard deviations for each element. Differences in peak counts were recorded in tabular and graphic formats. Based on studies by Kuhn (1985) and Kuhn and Sempowski (2001) using similar instrumentation, eight trace element ratios were selected as being the most potentially reflective of trace element variability between samples. These ratios are as follows: Iron/Strontium (Fe/Sr); Zirconium/Strontium (Zr/Sr), Rubidium/Strontium (Rb/Sr), Iron/Rubidium (Fe/Rb), Iron/Zirconium (Fe/Zr), Rubidium/Zironconium (Rb/Zr), Yttrium/Strontium (Y/Sr), and Yttrium/Rubidium (Y/Rb). Multivariate statistics, including principal components analysis, cluster analysis, and discriminant function analysis were applied to assess the degree of homogeneity between samples.

Trapps Gap Sample

Fifty-nine lithic samples from the Trapps Gap site were examined during this study. The samples were recovered from different units and were designated by the level in which they were found. When selecting samples for this study, the size and condition of the lithic artifacts were factored. Samples smaller than one centimeter were avoided because analysis would be difficult with current instrumentation. Consideration was also given to samples that were typologically identifiable as chert, that lacked surface evidence of weathering, and that had evidence of human modification.

In addition to the samples from the Trapps Gap site, eleven raw-material samples from New York chert outcrops were included to see how archaeological samples compared to archaeological chemical results (Figure 4.4, Table 4.1). Six of these (samples 1–6) were personally collected; the remaining samples came from collections curated at the New York State Museum. Five of these eleven samples originate from collections located within a 15-mile radius of the site. Four of the remaining samples were collected from deposits located between 15 and 50 miles east and west of the site. The remaining two samples were recovered from deposits located between 50 and 100 miles from the site near Middleburgh and Hoosick Falls, New York, respectively.

Figure 4.4

Map showing location of chert sources.

Table 4.1

Summary of Lithic Samples Collected.

Analysis and Results

Five lithic groups were identified and are here referred to as Groups A–E (Table 4.2). Principal components analysis showed that 89.9 percent of the total variance in the data set could be explained by the first four components (Table 4.3). Figure 4.5 illustrates the spatial relationship between the groups. Group A is the largest group with 30 artifacts clustering within this group (Table 4.1 and Table 4.2). This group includes cherts visually identified as belonging to different outcrops based on their light grey and dark grey color. Also in this group is raw material sample 2, which was recovered from Allard’s Corner’s, New York, and raw material sample 8, which is from Napanoch, New York. The loosely clustered members of Group A (Figure 4.4) suggest that several different deposits may have been used by the occupants of the Trapps Gap site. The second half of the Late Woodland (A.D. 1000–1300) is often characterized by increased sedentism, territoriality, and warfare. If this were true, we should expect tighter clusters suggesting that the region’s occupants may have continued to practice a semi-sedentary settlement pattern during this time period. Alternatively, warfare may not have been excessive, allowing prehistoric groups to move across clan and tribal territories to gather needed resources (Rieth 2002).

Table 4.2

Grouping of Lithic Materials from the Trapps Gap Site.

Table 4.3

Principal Components Analysis for Samples.

Group B contained 13 artifacts from the Trapps Gap site (Tables 4.1 and 4.2, Figure 4.5). In addition, the grouping also contained raw material samples 1, 3, and 6. Each of these samples was recovered from within less than 80 km (50 miles) of the site, with sample 1 recovered from within 24 km (15 miles) of the site. The grouping contained several artifacts recovered from the same test unit suggesting that many of these artifacts may have been produced from a single core or during a single knapping episode. The relatively tight clustering of data points also suggests that a limited number of different outcrops are represented by these materials.

Figure 4.5

Discriminant function analysis results showing five groups identified within the Trapps Gap lithic sample.

Group C contained seven lithic artifacts and one raw material sample (Tables 4.1 and 4.2, Figure 4.5). The sample consists of raw material Sample 5 which was procured from a source located at Hopewell Junction on the east side of the Hudson River. Although the sample was recovered from a deposit located more than 50 miles away, it is still within an intermediate distance of the site and could have been procured by groups living on the Wallkill plain that is located east of the site. This group contains highly variable artifacts that visually exhibit both grey and black colored cherts. Also, these materials contain two artifacts with potlids, suggesting that heat-treatment may have been needed to enhance the knapping of the materials.

Group D contained six pieces of debitage and one raw material Sample 9, which was recovered from Hunter, New York (Table 4.1 and Table 4.2, Figure 4.5). The sample from Hunter, New York, is located within an intermediate (24–80 km) radius of the site. Unlike groups A–C, the artifacts recovered from this group consist mostly of medium and dark grey chert bifacial thinning flakes recovered from three different units spread across the living floor of the Trapps Gap site. None of these artifacts showed signs of retouch and it is unlikely that they were used as expedient tools by the site’s occupants. The fairly loose clustering of the materials in this group suggests that the artifacts may have come from two or more different outcrops.

Group E consists of three different lithic samples as well as raw material sample 4, which was recovered from an outcrop near Middleburgh, New York (Table 4.1 and Table 4.2, Figure 4.5). This outcrop is located more than 80 km away from the Trapps Gap site. The scattered nature of the samples in this group suggest that they were procured from several different outcrops scattered a significant distance from the site. This is further supported by the fact that the samples do not cluster near Groups A–D, suggesting some differences in the concentrations in the elements contained within the materials.

In summary, this study suggests that the lithic artifacts from the Trapps Gap site can be grouped into five different lithic groups based on their trace element composition. These groups are identified as Groups A–E and contain variable characteristics. Several groups of artifacts clustered with raw material samples 1–9, suggesting that some of these samples may have been recovered from the eastern side of the Shawangunk Ridge in the Wallkill Valley. Since the focus of this study was not to identify where each individual lithic artifact was procured, future research is needed to test outcrops on the west side of the Shawangunks and compare how such outcrops influence the groups identified as a result of this research. The remainder of this chapter discusses what the results of the archaeometric analysis might mean in terms of interpreting the land use patterns of the Shawangunk Ridge and the adjacent Roundout and Wallkill Valleys.


The Shawangunk Ridge has attracted archaeologists since the end of the nineteenth century. Schrabisch described early explorations of the ridge in Mountain Haunts of the Coastal Algonquian (Schrabisch 1919a) and Indian Rockshelters of the Shawangunk Mountains (Schrabisch 1919b). In these publications, he described the archaeological sites of the region as being composed mostly of small rock shelters and camps, many of which are located in areas that are “quite small and uncomfortable … [with] … shelters sporadically at the foot of rock masses that become detached … an example … is furnished by a station [near] Lake Minnewaska, on Shawangunk Mountain, Ulster County, N.Y.” (Schrabisch 1919a:141).

The presence of the Trapps Gap site along the ridge supports Schrabisch’s (1919a) assumption that the Shawangunk Ridge was populated throughout prehistory. Few large Late Woodland settlements have been identified in the Shawangunk Mountains; instead most large sites are located in adjacent river valleys to the west or more open areas to the east on the Wallkill floodplain. According to Schrabisch (1919a:146), “at certain seasons, neighboring tribes would join in hunting trips to the mountains, which were a kind of game preserve, held in common by a group of tribes, and where they would stalk the deer and secure other quarry, valued for food and peltry.”

The trace element composition of the lithics from the Trapps Gap site suggests that the artifacts were recovered from several different deposits with at least five different groupings suggested. Many of the raw material samples clustered within these groups, suggesting that one or more local outcrops many have been exploited. The largest number of artifacts (43 percent) clustered with materials recovered from Napanoch and Allard’s Corners, suggesting relations with groups to the east and west of the Shawangunk Ridge.

The presence of more than 20 ceramic vessels and several dozen expediently chipped stone tools at the site suggests that the occupants were not merely hunting parties who cached resources as they followed deer and other migratory animals across the landscape, but may have also included women (and possibly children). This is important because historically, models of upland land use have focused on the use of these areas as hunting venues for males. Few studies have focused on women’s roles in the upland much less those of family groups. Recently, Versaggi (1996) and others (Rieth 2002, 2008, 2009) have suggested that women’s forays into upland areas were equally important to those of men and often served for collecting plant and animal resources that would have been needed to make baskets, mats, and other household items such as specialty foods and medicines.

The recovery of expedient tools from the site is in line with gendered models of tool use among women. Gero’s (1991; see also Gero and Conkey 1991) study of “gender lithics” and women’s roles in stone tool production in various hunter-gatherer groups shows that while expedient tools are made and used by both genders, women appear to make and use these tools most often. Tie in the use of expedient tools to plant procurement and small game collecting/processing, and the evidence begins to mount for interpretations of this and other upland sites as locations of female foraging and gathering.

Our interpretation of the Late Woodland use of the uplands assumes that the site’s occupants were logistically organized around an established residential base camp located in lowland or valley areas (Binford 1982). Evidence from the Upper Susquehanna and Schoharie Valley suggest that this assumption is appropriate for many parts of New York (Rieth 2009:1–18; Versaggi 1987). This model proposes that groups established residential base camps near areas of abundant aggregated food resources and moved their base camps to different locations across the landscape as resource availability changed. Consequently, the foraging radius, equivalent to the distance that could be traversed in less than a day, existed around the residential base. According to Versaggi (1987), within this radius, people collected, gathered, hunted and performed some processing activities as part of daily treks outside the perimeter of the camp. Those tasks that required longer-distance travel resulted in the creation of special-purpose, single- or multi-night encampments located beyond the foraging radius of the base camp.

The recovery of lithic materials from both sides of the Shawangunk Ridge support Schrabisch’s (1919a) aforementioned belief that the region was a communal hunting and resource procurement area for neighboring tribes. Temporally, however, the absence of Early and Middle Woodland occupations at many of these rockshelter sites suggests that during these periods, the Shawangunks may not have been widely used for resource procurement and/or hunting. What, then, would have caused native groups to resume collecting along the Shawangunk Ridge during the Late Woodland Period? 

Diamond’s suggestion that groups may have fled into upland areas to seek familiar shelter or collect in safer lands might explain the presence of Late Woodland and Contact materials in this area. As Iroquoian and Algonquian groups to the north and east reorganized village settlements and as competition for fertile land for growing crops increased, groups living on the margins of these areas may have retreated to upland areas such as the Shawangunk Ridge as they looked for new places to settle and forage food and non-food based materials.

Following Snow (1994:1–6), native perceptions of land use and ownership were very different from those of European groups, with territorial boundaries and resource exploitation zones fluctuating depending upon socio-cultural alliances between groups. Ashmore and Knapp (1999:2) suggest that many archaeological sites, such as isolated finds, farming stations, rockshelters, and resource procurement stations do not fit into the traditional definition of sites. Attention to these types of sites is important since they remind us how complicated the past is and stress the inter-relationship “among people and such traces, places, and features in space and time” (Ashmore and Knapp (1999:2).

Future research should be focused on comparing the patterns observed at the Trapps Gap site with other sites along the ridge to determine whether the lithic procurement patterns identified are unique to this site or represent wide-scale patterns across the ridge. Funk argued that, “due to its geographic setting, the region [and the region surrounding Trapps Gap pass] provides an unusual opportunity to test hypotheses concerning the range of various cultural manifestations and changing land use patterns. The mountains … presented natural barriers to the movement and communication on the part of aboriginal groups” (Funk 1976:7–8). For groups to traverse these barriers and communicate with groups on opposite sides of the ridge would have been difficult at best. In addition, navigating such a corridor during inclement weather and through unfamiliar terrain would not have made the journey any easier.

Finally, this research suggests that trace element analysis provides important information about the source of raw materials that may not be gained by visual inspection alone. Variation in chert outcrops has been underemphasized in studies of raw material use in the Northeast. As Wray (1948) and others (Calogero 1992:87-90; Hammer 1976:39–62; Kuhn and Lanford 1987:57–69) have suggested, variation in the color and texture of chert recovered from lithic outcrops can in some cases be very great and obscure attempts to accurately pinpoint source areas without a more extensive analysis of trace element composition. As the above discussion demonstrates, chert debitage that visually looked different based on color often grouped within the same trace element group. In order to truly understand how the materials from the Trapps Gap site fit into larger land use patterns, trace element analysis studies of the lithic artifacts from nearby sites such as Mohonk Rockshelter and Ski Minne should be undertaken to determine if the site’s occupants used many of the same outcrops. Only then can we understand the changes in the occupation of the ridge over time.


In this chapter we have summarized the results of a project conducted between researchers from the New York State Museum and Vassar College designed to examine the sources of lithic artifacts recovered from the Trapps Gap site, Ulster County, New York. The results of this research suggest that the site’s occupants procured materials from a variety of outcrops, with the lithic materials from the site clustering into at least five different groups. Most of the materials recovered grouped with known deposits located within between 24 km and 80 km of the site.

At this time, it remains unclear whether these patterns represent more widespread patterns of land use across the ridge or represent phenomena unique to the Trapps Gap site. If regional land use patterns are changing during the Late Woodland Period to more fully incorporate the occupation of rockshelter sites as temporary refuge areas for groups foraging across the ridge, archaeological excavations should be able to locate these occupations and the role that such sites played in the settlement and subsistence patterns of the region’s occupants. Future research should focus on detecting the frequency of such occupations across the ridge and the range of outcrops that were used by Late Woodland groups to the east and west of the ridge.


Archaeological investigations at the Trapps Gap site were completed by the Vassar College Archaeological Field School in 2006 and 2007 with assistance from the Daniel Smiley Research Center at Mohonk Preserve. The collections are temporarily curated at the Vassar College Archaeology Laboratory. Collections from the site are curated at the Daniel Smiley Research Center at Mohonk Preserve. Trace element analysis of lithic materials was conducted at the Accelerator Laboratory at the University at Albany, State University of New York. We also thank John Hart for his comments on an earlier version of this paper as well as co-organizing the symposium and volume for which this paper is presented. We thank the anonymous reviewers for their comments on this chapter. All errors and omissions are our own.

References Cited

Ashmore, W. and A. B. Knapp. 1999. Archaeologies of Landscape Contemporary Perspectives.

Blackwell Publishers, Oxford.

Binford, L. R. 1982. The Archaeology of Place. Journal of Anthropological Archaeology 1:5–31.

Brennan, L. A. 1979. Propositions Concerning the Early Archaic in New York. The Bulletin, Journal of the New York State Archaeological Association 75:1–14.

Calogero, B. L. A. 1992. Lithic Misidentification. Man in the Northeast 43:87-90.

Diamond, J. E. 1995. The Catskill Rockshelter, Town of Olive, Ulster County, New York. The Bulletin, Journal of the New York State Archaeological Association 110:16–25.

Eisenberg, L. 1974. Excavation at Old Fort II. The Bulletin, Journal of the New York State Archaeological Association 60:29–43.

Eisenberg, L. 1978. Paleo-Indian Settlement Pattern in the Hudson and Delaware River Drainages. Occasional Publications in Northeastern Anthropology No. 4.

Eisenberg, L. 1982. A Preliminary Analysis of Datum: A Multi-component Site near the Hudson River, Ulster County, New York. Man in the Northeast 24:1–36.

Eisenberg, L. 1991. The Mohonk Rockshelter: A Major Neville Site in New York State. In The Archaeology and Ethnohistory of the Lower Hudson Valley and Neighboring Regions: Essays in Honor of Louis A. Brennan, edited by H. C. Kraft, pp.159–176. Occasional Publications in Northeastern Anthropology, No. 11. Bethlehem, Connecticut.

Funk, R. E. 1976. Recent Contributions to Hudson Valley Prehistory. New York State Museum Memoir 22, The University of the State of New York, Albany.

Gero, J. M. 1991. Genderlithics: Women’s Role in Stone Tool Production. In Engendering Archaeology: Women in Prehistory, edited by Joan M. Gero and Margaret W. Conkey, pp. 163–193. Blackwell Publishers, Oxford.

Gero, J. M., and M. W. Conkey (Editors). 1991. Engendering Archaeology: Women and Prehistory. Blackwell Publishers, Oxford.

Hammer, J. 1976. Identification and Distribution of Some Lithic Raw Materials from New York State. Man in the Northeast 11:39–62.

Kaplan, J., and L. L. Johnson. 2009. Ski Minne Rockshelter and the Prehistory of the Shawangunk Ridge. Paper presented for the Annual Meeting of the New York State Archaeological Association, Syracuse, New York.

Kuhn, R. D. 1985. Trade and Exchange among the Mohawk Iroquois: A Trace Element Analysis of  Ceramic Smoking Pipes. Unpublished Ph.D. dissertation, Department of Anthropology, University at Albany, State University of New York, Albany.

Kuhn, R. D. 1986. Interaction Patterns in Eastern New York: A Trace Element Analysis of

Iroquoian and Algonkian Ceramics. The Bulletin and Journal of the New York State

Archaeological Association 92:9-21.

Kuhn, R. D., and W. A. Lanford. 1987. Sourcing Hudson Valley Cherts from Trace Element Analysis. Man in the Northeast 34:57–69.

Kuhn, R. D., and M. Sempowski. 2001. A New Approach to Dating the League of the Iroquois. American Antiquity 66: 301–314.

LaPorta, P. C. 1996. Lithostratigraphy as a Predictive Tool for Prehistoric Quarry Investigations: Examples from the Dutchess Quarry Site, Orange County, New York. In A Golden Chronograph for Robert E. Funk. Occasional Publications in Northeastern Anthropology, No. 15, edited by C. Lindner and E. V. Curtin, pp. 73-83. Occasional Publications in Northeastern Anthropology, volume 15. Archaeological Services, Bethlehem, Connecticut.

Rieth, C. B. 2002. Early Late Prehistoric Settlement and Subsistence Diversity in the Southern Tier Region of New York. In Northeast Settlement and Subsistence, edited by J. P. Hart and C. B. Rieth, pp. 209-226. New York State Museum Bulletin 496. The University of the State of New York, Albany.

Rieth, C. B. 2008. The Catskill I and II Sites: Two Upland Camps in Eastern NewYork. The Bulletin, Journal of the New York State Archaeological Society 123:27–35.

Rieth, C. B. 2009. Reevaluating Scale in the Eastern Woodlands: The View from Eastern New York. In Iroquoian Archaeology and Analytic Scale, edited by L. E. Miroff and T. D. Knapp, pp. 1–18. University of Tennessee Press, Knoxville.

Ritchie, W. A. 1971. A Typology and Nomenclature for New York Projectile Points. New York State Museum and Science Service, Bulletin 384. The University of the State of New York, Albany.

Ritchie, W. A. 1994. The Archaeology of New York State. Purple Mountain Press, Fleischmans, NewYork.

Ritchie, W. A., and R. E. Funk. 1973. Aboriginal Settlement Patterns in the Northeast. New York State Museum Memoir 20. The University of the State of New York, Albany.

Sando, A., and L. L. Johnson. 2008. The Upper Trapps Gap: Two Years of Excavation at a Prehistoric Rockshelter. Paper presented at the Annual Meeting of the New York State Archaeological Association,Syracuse, New York.

Schrabisch, M. 1919a. Mountain Haunts of the Coastal Algonquian. American Anthropologist 21:139–152.

Schrabisch, M. 1919b. Indian Rock-shelters in the Shawangunk Mountains. The Historic Wallkill and Hudson River Valleys 26:43–58.

Schrabisch, M. 1936. Archaeology of Southeastern New York. Manuscript contained in the New York State Museum Anthropological Collections, Albany.

Snow, D. R. 1994. Iroquoians and Europeans: Disunited Nations in the Early Contact Period. In Proceedings of the 1992 People to People Conference, edited by C. F. Hayes, III, C. Cox Bodner, and L. P. Saunders, pp. 1–7. Rochester Museum and Science Center, Research Records No. 23. Rochester, New York.

Snyder, B., and K. Beard. 1981. The Shawangunk Mountains: A History of Nature and Man. Mohonk Preserve Inc., New Paltz.

Sullivan, L. P. and S. C. Prezzano. 2001. The Concept of Appalachian Archaeology. In Archaeology of the Appalachian Highlands, edited by L. P. Sullivan and S.C. Prezzano, pp. xix-xxxiii. University of Tennessee Press, Knoxville.

Van Espen, P., K. Janssens, and I. Swenters. n.d. Version 3.0 AXIL X-Ray Analysis Software.

Department of Chemistry, University of Antwerp, U.I.A.

Versaggi, N. M. 1987. Hunter-Gatherer Settlement Models and the Archaeological Record: A Test Case from the Upper Susquehanna Valley of New York. Unpublished Ph.D. dissertation, Department of Anthropology, Binghamton University, State University of New York, Binghamton.

Versaggi, N. M. 1996. Hunter-Gatherer Adaptations in the Upper Susquehanna: Are the Uplands Part of the Picture? Paper presented at the Conference on Integrating Appalachian Highlands Archaeology, New York State Museum, Albany.

Wray, C. F. 1948. Varieties and Sources of Flint Found in New York State. Pennsylvania

Archaeologist 18:25–45.

« previous   |   next chapter »

Museum Open Tuesday - Sunday: 9:30 am to 5 pm | Carousel Hours: 10 am to 4:30 pm
Office of Cultural Education | New York State Education Department
Information: 518-474-5877 | Contact Us | Image Requests | Terms of Use
Sign Up for Email Updates
Join us on Facebook See us on YouTube See us on Flickr