edited by Christina B. Rieth and John P. Hart
Chapter 5: A SMALL SITE IN COXSACKIE, CIRCA A.D. 1200: SOME ECOLOGICAL ISSUES CONCERNING ITS AGE AND LOCATION
Abstract. The excavation of Concentration 23B.1, a small site in Coxsackie, provides an unusual glimpse at late prehistoric, short-term backcountry settlement in the Hudson drainage. Occupying low ground near a small stream within the lake plain, Concentration 23B.1 contained several archaeological features, a varied lithic assemblage indicating different stages of lithic reduction, fragmentary Owasco-like ceramics, and twelfth–thirteenth century A.D. radiocarbon dates. Settlement pattern implications are explored in terms of the changing use of the local landscape as well as the diversity of late prehistoric settlement systems in the upper Hudson Valley.
When the colonist Pieter Bronck purchased a large tract named Koixhackung (Coxsackie) from the Mohicans in 1662, the land was described as having 252 acres (102 hectares) of land already cleared, located away from the river (Dunn 1994:284; Vedder 1985:65). In Coxsackie, the land rises up quickly more than 30 m (100 ft) above the Hudson, indicating that if Bronck’s already-cleared land was located away from the river, it was on the Glacial Lake Albany plain, which has some sandy soil but is mostly clay and clayey loam (Broad 1993). This land presumably was cleared by the Indians, and unless recently cleared with steel axes, was cleared gradually with stone axes and the application of fire. Aboriginal clearing with stone axes and fire most usually occurred over a long period of time due to the labor involved (as Mann 2006:335–337 discusses in relation to aboriginal clearing in South America). Given the extent of cleared land in the Bronck Patent in 1662, the questions arise: what plausible circumstances contribute to the history of the cleared land, and how may this history and these circumstances be related to concepts of land use, settlement and subsistence during the Late Prehistoric period (A.D. 700–1500)?
Bronck’s discovery of already-cleared land to buy from the Indians was not unusual in the experience of seventeenth-century Europeans. Day (1953) has synthesized a wide variety of early European observations of extensive clearing, and more recently, Dunn (1994) has detailed the locations of some of these lands in the upper Hudson region. Dunn (1994) also has indicated the seventeenth-century Dutch preference for land already cleared. The Bronck Patent is one of many examples of Dutch purchases of cleared Indian land. Most of these Dutch purchases were along floodplains and low river terraces, however, in contrast to Bronck’s purchase in the interior.
Following Day’s (1953) seminal article, the historians Cronon (1983) and Pyne (1997) and the geographers Denevan (1992) and Doolittle (1992) promoted views of the Native American past that see active human clearing of the forest as a primary transformative force creating the world that Eastern Woodlands Indians lived in before European contact (although forest clearing or alteration by Native Americans is not always accepted as a significant factor in pre-contact Eastern Woodlands ecology by natural scientists, such as Forman and Russell [1983:5], Pederson et al. , and Russell [1983:86]).
Some archaeologists or interdisciplinary teams of archaeologists and natural scientists have taken the study of anthropogenic forest alteration further by investigating its time depth, extent, and more precise character. In some areas, forest alteration and at least some limited clearing were important by the Late Archaic period (approximately 3,000–4,000 years ago), and are clearly demonstrable on a more extensive basis in many areas by the Late Prehistoric (A.D. 700–1500) (Chapman et al. 1989; Delcourt et al. 1998; Johnson 1996; Moeller 1996; Patterson and Sassaman 1988:128–130).
Some of the prime reasons offered to explain the extensive clearing practice include the creation of gardens and the expansion of forest edge area to benefit hunting and gathering. In addition, cyclically performed seasonal forest burning cleared out underbrush (facilitating movement through the woods) and increased browse for prey species such as deer, while preserving an open forest of large trees, particularly trees of fire-resistant species (among others, see Cronon 1983; Day 1953; Johnson 1993; Moeller 1996; Patterson and Sassaman 1988).
The perspectives of anthropogenic Native American environments (developed increasingly through the second half of the twentieth century) are widely replacing earlier notions of uninterrupted, impenetrable, and mostly uninhabitable wild forests that had shaped earlier concepts of the pre-contact, eastern North American past. However, as this view affects archaeological and other interpretations more often, it is important to eschew static reconstructions (since the ethnohistoric literature does not adequately reflect conditions of 500 years earlier), and to consider the dynamics that affected forest clearing and associated gardening, hunting and gathering on a regional basis, recognizing that these human ecological systems likely fluctuated between periods of relative stability and instability, and that fluctuations affected different habitats or larger and smaller areas differentially.
For example, much of the land clearing observed in the Hudson valley or other areas of the Northeast during the early seventeenth-century must have occurred in the several centuries prior. However, this was a period of two notably extreme climate episodes, the Medieval Warm Period (A.D. 800–1300) and the Little Ice Age (A.D. 1300–1860). Did the forest clearing, often for agricultural purposes, proceed simply despite severe climate issues such as extended droughts, excessive flooding, and unpredictably curtailed growing seasons, or did progressive forest clearing spread risk or actually facilitate subsistence productivity in the face of these environmental perturbations? Moeller (1996) points to an answer by envisioning the clearing of gardens and extension of forest edge area (enhancing species diversity and abundance for hunting and gathering) as fundamental practices increasing diversity, and thereby adding resilience to human adaptation exceeding the effect of climate change. Current evidence indicating periods of higher and lower visibility of the Hudson valley archaeological record between A.D. 700–1500 suggests that climatic perturbations may have had effects upon settlement size and land use strategies.
Addressing these issues and questions in some detail here, this chapter examines (1) fluctuations in the Late Prehistoric (A.D. 700–1500) archaeological record of the upper Hudson valley; (2) the Medieval Warm Period (A.D. 800–1300) in relation to eastern New York State; (3) recent archaeological applications of an increasingly important body of theory, Resilience Theory; (4) apparent upper Hudson regional resilience strategies; and (5) an archaeological site referred to as Concentration 23B.1. Concentration 23B.1, located 2.4 km (1.5 mi) from the Hudson River, demonstrates use of the interior portion of the Bronck Patent several hundred years before Bronck’s purchase, and reflects settlement system diversity that existed in interior areas during the Medieval Warm Period.
CHRONOLOGICAL PATTERNS IN THE HUDSON VALLEY DURING THE LATE PREHISTORIC PERIOD (A.D. 700–1500)
The Late Prehistoric period is considered here as the A.D. 700–1500 time frame, a chronological interval that spans earlier definitions of the later part of the Middle Woodland period (A.D. 1–1000) and the entirety of the Late Woodland period (A.D. 1000–1500) (Funk 1976; Ritchie 1969). The alternative periodization used in the present chapter reflects apparent cultural continuity in artifact traditions and subsistence practices spanning the traditional Late Woodland temporal threshold. It also includes the entirety of the Medieval Warm Period (A.D. 800–1300). Within the present chapter, the term Late Woodland is used sometimes in reference to older data (such as Bender and Curtin 1990), or when reference to its traditional meaning was intended in earlier publications or reports.
In publishing the first archaeological overview of the Hudson valley, Ritchie (1958) looked to the existing Mohawk Valley archaeological sequence to document and internally differentiate the Hudson Valley Late Woodland period. This fundamentally meant the recognition of major Owasco and Iroquois-like chronological subdivisions, as well as the expectation that these major periods also would be divided into shorter phases. This effort, however, was hampered by a relative dearth of Owasco-like components, especially those of the Middle and Late phases (such as the Canandaigua or Snell Phase, A.D. 1100–1200, and Castle Creek Phase, A.D. 1200–1300 [Ritchie 1958:102]). Ritchie (1958:102) in fact suggested that the “little known occupants of the Hudson Valley…barred the Owasco people and their culture” from this region. At the same time, he found ceramic evidence for a later succession of cultural phases with pottery similar to the “Oak Hill, Chance, and Cayadutta prehistoric levels” (Ritchie 1958:108). Although Ritchie noted the occurrence of such evidence, he remarked upon the lack of substantial sites (similar to Iroquois village sites), attributing the sparse archaeological record to a combination of site destruction with urban development, short occupation spans, small group size, and high mobility.
Funk (1976:300–302) continued Ritchie’s (1958) view of a general Owasco-like to Iroquois-like material culture sequence, composed of phases that would replicate the Mohawk Valley sequence in terms of diagnostic ceramic types. However, Funk referred to the major chronological divisions as Late Woodland I (A.D. 1000–1300) and II (A.D. 1300–1600) in order to de-emphasize the Iroquois cultural reference for the Algonquian Hudson Valley. Funk (1976) found greater evidence of an Owasco occupation of the Hudson region than Ritchie, although still without major sites. Nonetheless, the triangular Levanna projectile point type, strongly diagnostic of the Late Woodland period, proved to be extremely common in Hudson Valley archaeological collections, indicating a relative population peak during this time (Funk 1976:312; see also Bender and Curtin 1990). Although the evidence of Late Woodland I occupation seemed sparse and scattered, Funk (1976:300–301; 310) viewed the information at hand, such as components in Albany and Ulster counties containing Owasco series ceramic types, as sufficient justification to anticipate the future discovery of Owasco village sites in the Hudson Valley. With respect to the Late Woodland II period, Funk (1976:311) took the existing evidence to indicate “unbroken development into stages similar to the Oak Hill, Chance, and Garoga horizons of the Mohawk Iroquois” (Ritchie [1958:108] had referred to the Garoga phase or horizon as the “Cayadutta level”).
In the several decades since these initial formulations were published, only a few new Late Woodland sites have been studied in the upper Hudson Valley, and those that have been studied tend to indicate small, seasonal occupations, sometimes repeated intensively. Indeed, there has been a growing sense that Late Prehistoric settlements were small, seasonal, and reflected a more mobile adaptation than the larger Owasco and Iroquois villages of central New York and the Mohawk Valley (Bender and Curtin 1990).
More radiocarbon dates are now available, allowing temporal evaluations that can be viewed independently of ceramic chronology and the traditional phase systematics. Improved chronology also benefits from more widely recognized associations between Native material culture and European material culture of known age, refining sixteenth to seventeenth-century time frames and views of the emerging and unfolding Contact experience (Bradley 2007). Better chronological control allows consideration of the potential contemporaneous use of ceramic types (such as various Oak Hill, Chance, and Garoga phase types) once typically assumed to indicate different time frames (Lavin et al. 1996), as well as the clarification of such concepts as the Garoga horizon (Bradley 2007).
Currently, an innovative way of looking at this period is to use the concept of a Late Prehistoric period spanning the late Middle and Late Woodland periods, and continuing to the Contact period, an apt historic context for studying the development of horticultural systems, settlement patterns, and land-use changes in much of the Northeast (approximately A.D. 700–1600; cf. Rieth 2002a, 2002b; various references in Miroff and Knapp 2009). Among other things, this concept allows views of periods such as A.D. 700–1100 as a continuous sequence, enhancing perspectives of long-term or continuous processes, while allowing comparison to periods that may differ in significant ways, although in the past these periods (as phases of the Late Woodland) have often been viewed as parts of a continuum.
It is proposed here that, although there was no strict hiatus in human occupation of the upper Hudson region during the Late Prehistoric period, there were alternating phases of relatively high or low archaeological visibility. The earliest of these is a period of high archaeological visibility circa A.D. 700–1000 or 1100. This period includes residence along the Hudson and large tributaries at such sites as Tufano and Black Rock (Funk 1976), Winney’s Rift (Bender and Brumbach 1986; Brumbach and Bender 2002), Site 211-1-1 (Cassedy and Webb 1999), apparently Welling (Funk 1976:300), and possibly Goes Farm (Brewer 2001). These sites have large artifact assemblages, riverine (often floodplain) settings, and frequently, substantial archaeological features such as pits and processing facilities.
This period is followed by a time of low archaeological visibility, A.D. 1100–1300. The backcountry site Concentration 23B.1 as reported later in this chapter was occupied during this second phase. Other small, backcountry sites nearby that seem to contain evidence of occupation during this period include the Bronck House and Zimmerman rockshelters (Funk 1976). Other roughly contemporaneous occupations include evidence of a ceramic tradition that incorporated both Middle Woodland and Late Woodland attributes dated between ca. A.D. 1150–1385 (1σ range, uncorrected thermoluminescence and radiocarbon dates) from the Winney’s Rift and Mechanicville Road sites (Brumbach and Bender 2002:236; also see Hartgen Archeological Associates 1983). Other possible components of this period would include assemblages with “Canandaigua” or “Castle Creek” ceramics; however, there appear to be few if any large assemblages or sites with substantial features from this time-range documented in the upper Hudson region (the briefly reported Little Wood Creek site in Fort Edward may be an exception; see ceramics mentioned by Starbuck 2004:4). Thus, to the extent that sites of this time can be adduced from existing information, they occur in riverine as well as backcountry locations. The paucity of large or well-documented sites from this time is taken here as an indication of either population decline, or some other aspect of low archaeological visibility, such as smaller group size, higher mobility, and dispersion into backcountry areas.
The following period of time—A.D. 1300 or 1400 to 1700—is one of relatively high archaeological visibility, as sites are more numerous and are often more substantial than earlier ones. These sites are located in riverine settings, and in some cases represent reoccupations of sites occupied during the earlier time period, A.D. 700–1100. During this period, Goldkrest and several other sites were occupied on Papscanee Island (Lavin 2004; Lavin et al. 1996; Sopko 2007; Sopko and Schmitt 2009). On the Roeliff Jansen Kill, Site 211-1-1 appears to have been occupied intensively once again (Cassedy and Webb 1999:87-89), apparently after a hiatus or period of low intensity site use. The Coffin site near Schuylerville was occupied at this time, after a very long hiatus (Funk and Lord 1972). Winney’s Rift continued to be occupied, perhaps more intensively at times. The most intense episode of occupation of the Goes Farm during the Late Prehistoric occurred in this period, based upon ceramics and burial features (Brewer 2001). These sites variously contain storage and processing features, and in some cases, burials.
Almost 40 years after Funk’s finding that major Late Woodland components are lacking—while Late Woodland projectile points are relatively abundant—accumulating evidence indicates that the Late Prehistoric is comprised of alternating periods of relatively high (A.D. 700–1100, A.D. 1300–1700) and relatively low (A.D. 1100–1300) archaeological visibility. Setting aside the possibility of a population decline, three aspects of this differential visibility may include: (1) variation between assemblages with different degrees of normative content or pattern; (2) differences in site size or artifact density; and (3) differential use of riverine and backcountry or upland locations. Concentration 23B.1 provides an example of all three dimensions of low archaeological visibility, since temporally diagnostic artifacts are lacking, site size is small, and the location is on clayey soil within the interior lake plain, some distance from the Hudson River. These factors make sites like Concentration 23B.1 difficult to find and recognize as Late Prehistoric components.
The Medieval Warm Period
Fagan (2008) has recently provided a broad and accessible discussion of the Medieval Warm Period. This term and the alternative, Medieval Climatic Anomaly, refer to a period of global or near-global warming from A.D. 800–1300. These terms have been used since the 1950s and 1960s when the British meteorologist Hubert Lamb introduced the notion of a “Medieval Warm Period” based upon the study of historic climate records.
The Medieval Warm Period was a largely beneficial period of warm weather and adequate rain for agriculture in Western Europe, which led to population growth, forest clearing, the proliferation of new farms, church-sponsored building, and the accumulation of wealth. The main deleterious effect in much of Europe was famine when population growth outstripped food-production. Two responses, in addition to more land clearing, included technological improvements for more effective plowing and the intensification of fishing. Warfare over neighboring food surpluses was another effect of local food shortages. Because of the growth of European population, economy, and cultural expression over this period, the Medieval Warm Period has often had the positive connotation of an ameliorating environment when used by European and American historians and archaeologists (for example, Snow 1994:21).
Much of Fagan’s (2008) book, however, is about medieval warming effects in the rest of the world, where this climatic episode was often associated with deleterious droughts and significant reductions in food supplies. Fagan (2008) chronicles the severe effects of warming upon North American populations in California, the Great Basin, and the Southwest, where the extreme droughts of the late twelfth-century led to the Chacoan and other abandonments (see Judge 1989; Larson et al. 1996). Recently, Benson et al. (2009) have considered the effect of severe twelfth to thirteenth-century droughts on the Mississippian urban center of Cahokia. They relate an eleventh-century cycle of high rainfall to rapid growth and agricultural diversification and subsequent long periods of drought to abandonment of a significant upland farming complex, social and political reorganization within Cahokia itself, and progressive depopulation of the urban center leading to abandonment.
Maps of drought severity (informed by tree ring data) indicate that the Hudson Valley was somewhat or marginally affected during the twelfth to thirteenth centuries (Benson et al. 2009:477). Other data, however, indicate that medieval warming effects may have been more adverse locally than these maps suggest. Fagan (2008) cites the study by Pederson et al. (2005) indicating environmental change in the lower Hudson region related to the Medieval Warm Period. Fagan (2008:229) notes that from A.D. 800 to 1300, the lower Hudson estuary became more saline. Increasing salinity in the lower estuary primarily was the result of decreased fresh water flow from upstream areas to the north (including the Hudson’s major western tributary, the Mohawk; cf. Boyle 1979). This suggests a long period of drought, including areas well upstream from the Piermont Marsh-lower Hudson study site.
According to Pederson et al. (2005:238), the major (and perhaps most direct) evidence for a strong expression of the Medieval Warm Period in the Northeast involves “striking increases in charcoal and Pinus domination from ca. A.D. 800 to 1300, paralleling paleorecords southward along the Atlantic seaboard.” This evidence is interpreted as an indication of significantly increased forest burning, which would result in the increased deposition of charcoal in stream and marsh sediments, and a shift of pollen frequencies toward the increased representation of fire-resistant tree genera such as Pinus. Pinus species also are early successional and tend to colonize abandoned fields and clearings that have been in transition (Asch Sidell 2008).
Pederson et al. (2005) attribute the Piermont Marsh data to the effect of natural fires greatly exacerbated by drought, although anthropogenic fire is also discussed, but viewed as less likely. However, there have long been differences of opinion about the ancient prevalence of natural wild fires versus human-set fires for land clearing or forest management (Day 1953; Denevan 1992; Patterson and Sassaman 1988). Certain vegetation changes may reflect anthropogenic change: for example, Native American gardening may account for the strong representation of Ambrosia (ragweed, a chronic garden invader) in the Piermont March samples pertaining to the Medieval Warm Period (data in Pederson et al. 2005). In eastern Kentucky, Delcourt et al. (1998) see an increase in Ambrosia pollen as consistent with other evidence of anthropogenic fire and increased gardening. In an alternative view involving the lower Hudson data on the Medieval Warm Period, Pederson et al. (2005:245) note that Ambrosia is expected to quickly colonize burned over areas. Nonetheless, recognizing a long-standing controversy based upon different paradigms, it seems reasonable to continue to consider the possible human role with respect to the lower Hudson (Piermont Marsh) data on forest burning and vegetation changes. At the same time, there appears to be substantial evidence for increased forest burning in the lower Hudson region during the Medieval Warm Period, whatever the proximal cause.
Further north, above the Mohawk-Hudson confluence, a study of core samples from Collins Lake (an oxbow on the lower Mohawk) indicates a long period of increased storminess from ca. A.D. 1150 to 1590 (White and Rodbell 2010). However, it is unclear how increased storminess may have been related causally to the Medieval Warm Period—for example, whether there may have been less rainfall but more violent storms—since the indicated time frame spans the last century or two of the Medieval Warm Period and the first half of the Little Ice Age. Nonetheless, these data strongly indicate potential issues for floodplain agriculture as early as the twelfth century, due to increasingly severe floods and erosion, and perhaps for agriculture in general due to factors such as increased wind and hail damage.
Although the evidence is circumstantial, the possibility of a delay of 300 to 400 years between forest clearing and intensive seasonal occupation of the Goldkrest site on the upper Hudson may be related to floodplain dynamics during this period (since earlier instability of the floodplain seems to have significantly inhibited human use). At Goldkrest, channel migration and over-the-bank deposition eventually led to a somewhat more elevated and stable ground surface by ca. A.D. 1400 (based on discussions and data in Lavin  and Schuldenrein ).
It is important to consider that climate effects may vary within and between regions. For example, Mullins et al. (2011) have retrieved diverse paleoenvironmental data from Cayuga Lake in New York’s Finger Lakes region that indicate an increase in local rainfall rather than drought during the Medieval Warm Period. This appears to conflict with near-continental scale, mapped information depicted by Benson et al. (2009:477). However, while noting that Benson et al. (2009:474) also record increased rainfall in the Cahokia region during the eleventh-century, providing at least a partial, potential similarity to the trend in the Cayuga Lake data, it would seem that much more information on spatial and temporal variation is needed to understand medieval climate change in the Northeast. In any event, it seems wise not to assume uniform conditions during major episodes of climate change, but to be aware of issues of environmental perturbation and instability while accumulating data.
Considering the major trend of the available information, the Medieval Warm Period had several possible adverse effects upon subsistence and settlement strategies in the Hudson Valley. These include unstable floodplain sites affected by seasonal submergence, devastating erosion, and burial under loads of sediment. Potential deleterious effects also include propensities for crop damage due to droughts, floods, high winds, and hail. When it happened, increased May-June flooding may have delayed planting, perhaps adversely. Uncontrolled wild fires may also have become a greater threat during the Medieval Warm Period (although the threat of uncontrolled fire would have been greatest in places where dead wood accumulated on the ground, especially in blow-downs associated with strong storms [Day 1953; Pyne 1997]).
Resilience in Social and Ecological System Interaction
Recently, several archaeologists have discussed Resilience Theory and the related concept of panarchy as a basis for understanding linked changes in social and ecological systems (Delcourt and Delcourt 2004; Nelson et al. 2006; Redman 2005; Redman and Kinzig 2003; Thompson and Turck 2009; for a critique of sorts, see Jones 2005). A short definition of resilience is the “ability to recover rapidly, as from misfortune” (Webster’s Dictionary 1996). Panarchy refers to the interaction of ecosystems at nested spatial and temporal scales and, in the anthropological context, to the operation of human ecosystems at different spatial and temporal scales (Delcourt and Delcourt 2004). In human ecosystems, regional processes of social and political organization, interaction, and information exchange integrate spatial scales, while the ability to remember the past (such as social memory), or even anticipate the future integrates or connects temporal scales:
These nested hierarchies may have a stabilizing effect due to the fact that they provide the memory of the past and of the distant to allow recovery after change occurs. They may also have a destabilizing effect when dynamics across scales become “overconnected” or “brittle” allowing small scale transformations to “revolt” and explode into larger scale crises. (Redman and Kinzig 2003:1)
Resilience Theory sees change between stable and unstable system states as normal and expects ecosystems to follow cyclical patterns of growth, conservation (of resources, institutions, or practices), release (or “revolt”), and reorganization. Growth and conservation are associated with system stability; release and reorganization are associated with system transformation. During periods of instability, when conservation fails and release occurs, systems change quickly, and then reorganize to adapt (Nelson et al. 2006; Redman and Kinzig 2003; Thompson and Turck 2009).
Resilience Theory encourages thought about diversity and flexibility as important components of observed regularities, such as responsiveness to seasonal cycles of resource availability or abundance (Thompson and Turck 2009:256–257). In order to visualize resilience as a process affecting small scale societies, it is possible to imagine the failure or partial failure of agricultural subsistence strategies due to droughts or floods, leading to a subsistence-settlement system release into smaller, more mobile co-residential groups subsisting on wild resources to a significantly increased extent (see Nelson et al. 2006). This could occur on temporal scales ranging from a year to a multi-decadal drought, as occurred during the Medieval Warm Period in parts of North America. Subsistence and residential diversity and flexibility, often as they are evoked by social memory and the preservation of traditions, may condition successful subsistence and settlement strategies at local scales and in age-old habits or remembered places, technologies and behaviors.
Resilience, seen as the potential to respond effectively, allows recovery when small scale transformations lead to large scale crises. In some dramatic cases, such as the sequences of Medieval Warm Period drought and abandonment (or irrigation problems and geographic contraction) seen in Chacoan, Mimbres, and Hohokam societies of the American Southwest, human and social ecosystems have been seen as extremely stressed and described as collapsing, but also subsequently reorganizing, often with significant change in geography, demography, cultural institutions, and sociopolitical organization (Cordell 1997:399-441; Diamond 2005; Fagan 2008; Nelson et al. 2006; Redman and Kinzig 2003). For example (and apropos of the thought-experiment above), Nelson et al. (2006) have examined the reorganization process in the Mimbres region of New Mexico. While drought was an external problem, the prior decline of diversity in social and ecological units was seen as a general, contributing factor leading to transformation. One way of looking at this is to view human communities in the region as too over-committed to agriculture to withstand the Medieval Warm Period droughts. Increasing commitment to agriculture and larger communities had reduced social and ecological diversity by localizing population and making food crops more central to subsistence. While large communities were in fact depopulated as a result, small communities shifting toward the use of more wild resources and smaller residential sites characterize the release and reorganization phases of the “Mimbres collapse” adaptive cycle.
In some of the archaeological literature on the Resilience Theory adaptive cycle, there is a sense that the release phase involves processes such as fragmentation into smaller co-residential groups, dispersion of population, increased mobility, and diversification of the resource base. This model has been hypothesized for hunter-gatherers shifting from collector to forager strategies (Thompson and Turck 2009:258), but may also pertain to mixed hunting and gathering and agricultural strategies, especially those transitioning toward increased hunting and gathering (Nelson et al. 2006).
Once agriculture had become a significant food source in a region such as eastern New York, environmental perturbations may have similarly favored resilience strategies leading to reorganization. Trends toward smaller site size and dispersion into the backcountry (which would equate at least in part with subsistence shifts toward increased hunting and gathering and the diversification of food resources), would also lead to lower archaeological visibility, since smaller sites located in settings not viewed as typical for the period are less likely to be found or investigated by archaeologists. Thus, it is possible to hypothetically connect the degree of archaeological visibility generalized for the upper Hudson region to the dynamic cycle proposed by Resilience Theory, and the climatic vicissitudes surrounding the Medieval Warm Period, as shown in Table 5.1.
Dynamic Cycles of the Upper Hudson Late Woodland.
In Table 5.1, the Medieval Warm Period is divided into earlier and later sub-periods related to perceptions of archaeological visibility and an as yet vague notion of the severity of contemporaneous climate change. In doing so, the apparent, near continent-wide, extended droughts of the twelfth and thirteenth centuries are assigned greater significance than climate change during the earlier portion of the Medieval Warm Period, although with sufficient water, warming during the ninth to eleventh centuries may have greatly favored the expansion of agriculture.
The assumptions made regarding settlement and subsistence are based on the limited data from these periods. Any interior dispersion of population is based in part on negative data, i.e., the paucity of well-documented interior sites and the assumption that this is due to low archaeological visibility. However, the record of interior occurrences of Late Woodland (or Late Prehistoric) sites based upon institutional records (Bender and Curtin 1990) is discussed further below, and seem to strengthen the case for regular use of interior, backcountry and upland settings.
Since linked social and ecological systems are considered as nested at different temporal and spatial scales, this model currently is presented as characterizing only the Late Prehistoric period in the upper Hudson region, and not necessarily as characteristic of adjoining regions, or of the Northeastern Late Prehistoric period in general. However, because global or hemispheric-scale climatic events affect very large areas, adjoining regions so affected are also likely to contain contemporaneous evidence of social-ecological stress and responses. These effects and responses may parallel those observed in the upper Hudson region or they may show regional differences. Depending upon inter-regional interactions, some systemic changes may be linked at a broad regional scale.
Strategies Affecting Resilience in the Upper Hudson Region
The upper Hudson region as defined here includes the Hudson drainage from about the mouth of Catskill Creek in Greene County to the Town of Queensbury in Warren County, New York. In the upper Hudson region, strategies favoring Late Prehistoric human ecological resilience fall into three major categories: hunting and gathering, agricultural risk, and anthropogenic change.
Hunting and Gathering
An increase in the importance of hunting and gathering in the reorganization of subsistence and settlement has been cited in a Southwestern U.S. example (Nelson et al. 2006), and imagined as pertaining to the present situation discussed later in this chapter. While easy to imagine, is there currently any evidence for this in the upper Hudson region?
At one level, the answer is no, in the sense that floral and faunal analyses pertaining to this question have rarely been performed, and will be somewhat limited in the future by poor faunal preservation in this region. At another level, however, the answer is apparently yes, because there is artifact spatial-distributional data supporting this answer in a 1990 study of institutional site file information from the upper Hudson region (Bender and Curtin 1990). Conclusions may be seen as preliminary, since a great deal of archaeological survey has been performed since these data were compiled. However, Bender and Curtin (1990) made considerable note of the variable quality of the data set, and interpreted it with due caution. Relying upon temporally diagnostic artifacts such as projectile point and pottery types, as well as the locations and soil associations of all recorded archaeological sites within the region at that time, this study showed several important correlations: (1) floodplain soils are highly productive for corn crops; (2) Late Archaic and Late Woodland sites follow similar trends in relation to soil capacity for corn production, except for the most productive soils, at which point Late Woodland site frequency increases dramatically in comparison to Late Archaic site frequency; and (3) Late Woodland sites are relatively concentrated on soils with the highest corn productivity (13 percent of Late Woodland sites occur on these soil types, which occupy 2 percent of the region [Bender and Curtin 1990:91]). Meanwhile, Bender and Curtin (1990:87) had found that compared to most other periods, the Late Woodland period had a relatively high frequency of sites that were occupied for the first time. This is taken as an indication that a new or unique settlement strategy was in operation, one that probably pertained to horticultural intensification involving soils with high corn production capacity.
At the same time, approximately 75 percent of sites occupied during the Late Woodland had been occupied previously. Some of these multicomponent sites are floodplain locales and others are in the backcountry. Summarizing their perspective of these intersecting dimensions of land attributes and use, Bender and Curtin (1990:91) stated that “Late Woodland settlement was semi-sedentary and still incorporated aspects of earlier hunting and gathering adaptations. This type of settlement system would require the continuation of strategic procurement locales in habitats not related to horticultural activities.”
This can be considered to be a substantial mapping-on to earlier hunter-gatherer land-use patterns (although those patterns themselves are somewhat more complex than this equation implies, and does not recognize variation in Late Prehistoric land use strategies). While this may be viewed synchronically as part of the diversity of a mixed agricultural-hunting-gathering subsistence strategy, a diachronic perspective on the changing intensity of backcountry land-use is needed to more fully relate site location data to propositions of relative archaeological invisibility, or population dispersion in relation to climate change. The data do, however, indicate the plausibility of social memory of backcountry subsistence activities and site uses as a factor favoring resilient shifts toward increased hunting and gathering.
Spreading environmental risks to crop production over space is a fundamental way of making agricultural systems more resilient. Dunn’s (1994, 2000) discussions of the ownership or control of gardens (or lands available to garden) by Mohican clan leaders shows how a close relationship between social organization and land use spread this risk in the upper Hudson region. Records of land sales indicate that clans controlled land in different drainages, including lands along both sides of the Hudson and on different tributaries. This obviously could allow harvests in some places despite losses in others. Although the antiquity of this system is not known, it is a system that may have developed in extent and complexity during the Late Prehistoric period, as more land was cleared for agricultural production, and perhaps as clans became larger, stronger, and more influential over larger areas. Thus, while this system provided for a high degree of resilience during the very Late Prehistoric and Contact periods (and may have effectively aided reorganization through crisis periods of the contact experience), it is not clear how applicable it was to buffering climatic perturbations of the Medieval Warm Period.
Nonetheless, during the Contact period, some of the Mohicans’ cleared lands were located away from the river, as indicated in the description of the land sale to Pieter Bronck. The large amount of land cleared, 252 acres (102 hectares), may have required a significant investment of time and labor, perhaps executed over several generations. It is highly likely that these cleared lands in the Bronck Patent were located on clayey soils that retain water during dry periods better than sandy and silty soils. The use of cornfields in locations with different soil qualities (and perhaps a different mix of wild resources in surrounding woods and forest edges) adds diversity to floodplain-dominated agricultural systems.
Anthropogenic change in this case refers to Native American land-use practices involving land clearing and controlled burns in the woods to enhance hunting by clearing out underbrush and increasing forage for animal prey. Fire was used in both cases, accompanied by the use of stone axes for land clearing. The antiquity of these practices and their geographic extent at any particular time in the remote past are not known, but evidence of land clearing during the Late Archaic has been reported from Tennessee and Kentucky (Chapman et al. 1989; Delcourt et al. 1998), and burning of forest understory during the Terminal Archaic has been reported from western Massachusetts (Johnson 1996). These practices are well documented as occurring in the Hudson Valley and other parts of the Northeast during the Contact period (Cronon 1983; Day 1953; Denevan 1992; Engelbrecht 2003; Mann 2006; Patterson and Sassaman 1988; Pyne 1997). Possibly reflecting these processes, at Concentration 23B.1 (discussed below), the soil below the plow zone is infused with small bits of charcoal and melted silica spheres, while the surrounding terrain has elevated frequencies of heat-altered artifacts (Curtin 2008; McKnight 2010).
Moeller (1996) has provided an important perspective on this for the Late Woodland period. He has argued that systematic alteration of the environment by land clearing is both an ancient practice and an important aspect of Late Woodland ecology and adaptation. Increasing forest edge area, Moeller (1996:63) observes, is a significant way to increase hunting and gathering potential because of the ecological diversity created in these situations. This practice probably was going on for so long that it seems that it should be considered a traditional practice more often than a Late Woodland innovation. If this is indeed the case, then the traditional backcountry landscape that Late Prehistoric people in the upper Hudson continued to use for hunting and gathering already contained cleared, open areas and forest edges.
Forest edge is most effectively created by expanding existing openings, since clearing new patches produces little edge area at first, and thus is slower and more labor intensive (Doolittle 1992:392–393; Mann 2006:335–336). Therefore, clearing probably was accomplished progressively over time as part of traditional uses of specific landscapes. The very large clearings reported during the Contact period (as described by Dunn 1994, Engelbrecht 2003, and others) would seem to imply a significantly long period of progressive forest clearing prior to contact.
Moeller (1996:65) has also observed that the effects of land clearing and field maintenance by burning exceed “those of climatic change over the short term.” That is, the anthropogenic landscape provided significant biodiversity and potential garden plots that could be tapped with some reliability when subsistence alternatives were needed. Moreover, Native American creation and expansion of patches of biodiversity can reasonably be assumed to have been an ongoing process prior to and during the onset of the Medieval Warm Period. The clearing and traditional uses of open areas in the forest for species diversity, enhanced hunting, and agricultural fields would have been a significant factor adding resilience to the Late Prehistoric adaptation. Making clearings in the backcountry would have created opportunities to increase social and ecological diversity by favoring the dispersion of small, mobile groups as a subsistence-settlement alternative to larger, horticulturally based settlements along the rivers.
The following discussion of a recently investigated backcountry site, Concentration 23B.1, provides an example of a portion of upper Hudson prehistoric land use that is relevant to considerations of small groups, mobility, possible subsistence diversity, and flexible settlement strategies during the Late Prehistoric period. Moreover, based upon radiocarbon dating, there is good reason to believe that Concentration 23B.1 was occupied during the later part of the Medieval Warm Period, when adverse climatic effects may have been most severe. This site is located within the interior portion of the Bronck Patent, and thus plausibly involved with land uses such as the extension of forest clearing during the Late Prehistoric period, or the actual use of soil or biotic resources associated with the cleared area. The surrounding area contains earlier sites reflecting a long tradition of hunting and gathering. Based upon size, location and apparent age, this type of site is predicted by a theory of social-ecological system release and reorganization associated with prolonged episodes of climate-related stress during the Late Prehistoric period.
Concentration 23B.1 (Figure 5.1) is a small, Late Prehistoric site of 225 sq m. It sits in a broad, shallow swale along an unnamed drainage leading to Coxsackie Creek. The Hudson River is ca. 2.41 km (1.5 mi) directly to the east, or ca. 4.82 km (3 mi) to the northeast if the route taken follows the winding Coxsackie Creek to its mouth. In this respect, Concentration 23B.1 is oriented to upland food resources rather than riverine, and it has no access to significant floodplain (the narrow floodplain along Coxsackie Creek in this area is classified as poorly drained to very poorly drained Wayland silt loam (Broad 1993). However, the site was more directly connected to other Indian communities to the south, as it is only about 0.80 km (0.5 mi) east of a major, north-south pathway, the traditional Indian footpath following the base of the Kalkberg Ridge (Figure 5.2). In early historic times this trail linked the lands around Catskill Creek with those to the north (a portion of this trail is shown on a colonial-era map included in Dunn 1994:195).
Concentration 23B.1 location in Coxsackie (Greene County highlighted in map inset).
View of the Kalkberg Ridge from the western side of the Greene Business and Technology Park study area. An Indian trail (the Catskill Path) ran north-south at the base of the ridge.
The Lake Albany plain is relatively wide in this section of the Hudson Valley, and is bounded on the west by the Kalkberg Ridge (part of the Helderberg formation) and on the east by the trough of the Hudson River. The predominant soil type is the Kingsbury and Rhinebeck series of somewhat poorly drained silt and clay (Broad 1993). The Kingsbury-Rhinebeck series underlies and surrounds Concentration 23B.1. The site’s aspect is to the north, shading it somewhat during summer and exposing it to cold winds during the winter. Personal experience working at this locale during the late fall and early winter indicates that exposure to wind may have made occupation of this site undesirable during the cold season.
Relatively high quality, workable stone is abundant in Coxsackie. Normanskill chert is available from tilted shale exposures that outcrop through the lake plain floor. A prominent source of Normanskill chert, Flint Mine Hill, is located about 4.82 km (3 mi) to the south of Concentration 23B.1. In addition, Helderberg cherts occur in the escarpment forming the valley wall on the west, possibly as close as 0.80 km (0.5 mi) to Concentration 23B.1. Normanskill chert is the most abundant type in local artifact assemblages, while Helderberg cherts such as Kalkberg often are strongly represented minority types. Eastern Onondaga chert also is present in local artifact assemblages, but not nearly as common as the Helderberg cherts. Exotic stone types such as Pennsylvania jasper, purple-weathering argillite, western Onondaga chert, or Flint Ridge chalcedony are much less common (Cassedy 1992).
Local chert is so abundant that its procurement sometimes has been thought of as a reason that ancient Native Americans visited Coxsackie (Parker 1924). More recently, however, settlement system analyses and chronological studies have indicated at least seasonal residence in the local area on a sustained basis over a very long time (Curtin 2008; Curtin et al. 2008; Funk 2004). From this perspective, the abundant chert appears to be an important resource within an area rich in other environmental resources due to the Hudson River, local creeks and wetlands within the wide lake plain, and varied upland forest resources. The combination of resources favored sustained settlement in the local area over several thousand years.
In 2002, supported by the Greene County Industrial Development Agency (IDA), Curtin Archaeological Consulting intensively sampled Concentration 23B.1 as part of the Phase 3 data recovery project conducted for the Greene Business and Technology Park. Concentration 23B.1 was the only Late Prehistoric site identified during this work. Most of the other sites were either small Early-Middle Archaic, short-term residential sites, or had unknown periods of occupation. The surrounding landscape of some 81 hectares (200 acres) appears to have been occupied only sporadically after the Middle Archaic. Late Archaic–Early Woodland sites are few and appear to be logistical field camps. Most of the surrounding area appears to have been used for food or other resource procurement from the Late Archaic onward. Concentration 23B.1 appears to represent a reoccupation of this area after a settlement hiatus of perhaps 2000 years (Curtin 2008). During this period of settlement hiatus, continued low intensity use of the area now containing the Green Business and adjoining Kalkberg Commerce Parks is indicated by the isolated recovery of Early and Middle Woodland projectile point types such as Rossville, Fox Creek Stemmed, and Jack’s Reef Corner-Notched.
The excavation of Concentration 23B.1 was conducted primarily within a large block and adjoining, short, hand-excavated trenches consisting of 129 1-x-1-m excavation units (Figures 5.3, 5.4, and 5.5). Two 1-x-1-m units were deployed in a low artifact density area east of the excavation block, while a backhoe trench removed the topsoil in an area surrounding the excavation block in order to explore for outlying features.
Concentration 23B.1 excavation plan.
Concentration 23B.1, view of excavation block in progress, looking northwest toward Kalkberg Ridge.
Concentration 23B.1, view of excavation in progress, looking east.
The block excavation found a plow zone about 35–40 cm thick overlying either the subsoil or a thin (4–8 cm) dark brown soil deposit containing artifacts, charcoal, and melted silica spheres. This stratum was called the brown organic zone, and since patterns of plowscars seen outside of it appeared not to penetrate through it, the brown organic zone is considered part of the natural stratigraphy, rather than an older plow zone remnant. Figure 5.6 shows the extent of the brown organic soil and the total artifact density. These distributions correspond rather closely, suggesting that the dark stain is associated with more intense levels of human activity. The area of higher artifact density largely reflects the density distribution of chert debitage in the excavated sample (all strata combined).
Concentration 23B.1, distributions of total artifact density and brown organic stratum. Artifact densities equal 0–41 artifacts: unshaded; 41–108 artifacts: lighter shading; >108 artifacts: darker shading.
A complex pattern of historic and prehistoric features was found in the block excavation (Figure 5.7). The historic features consist of numerous post molds that contained badly corroded nails. These tended to form linear alignments, as if they marked intersecting or reconstructed fence lines. The longest line of historic post molds was exposed over a distance of 11 m (36 ft), within which the post molds were usually spaced about 2 m (6 ft) apart.
Concentration 23B.1, plan view of archaeological features.
In addition to the historic post molds, several post molds were found that did not contain nails, including a group of apparent prehistoric post molds that formed an elliptical pattern 3 m long and 2 m wide. This pattern is interpreted to represent a small shelter or windbreak (Figure 5.8).
Concentration 23B.1, detail of inferred shelter area.
Other features included small pits, a basin-shaped hearth, charcoal stains, and patches of burned earth. Notable among these is Feature 5, a hearth or shallow pit. It has been radiocarbon dated by Beta-Analytic Inc. and Geochron Labs to about 750–780 radiocarbon years (circa A.D. 1200), which (using the CALIB 6.0.2 program) appear to calibrate to the early to mid-thirteenth century (although multiple intercepts of the calibration curve indicate several possibilities [Table 5.2]). The original calibrated date provided by Beta-Analytic Inc. (using INTCAL98 Radiocarbon Age Calibration) was cal. A.D. 1260, with a 1σ range of cal. A.D. 1210–1280.
Concentration 23B.1 Radiocarbon Dates (Obtained from Feature 5).
Feature 6 was a small pit containing highly fragmented pottery and small flecks of charcoal. Its vasiform shape and concentrations of pottery near its walls (Figure 5.9) suggest that it may have been a buried pot rather than a pit per se (or perhaps it was an otherwise undiscerned, larger pit containing a pot).
Concentration 23B.1, west profile of Feature 6.
Feature 13 was a lithic debitage concentration measuring about 1 m by 35–40 cm, and occupying a rough crescent shape. Its contents were significantly different from the general debitage assemblage recovered from Concentration 23B.1. While the debitage assemblage from the site in general showed a relatively low percentage of large flakes (about 26 percent of flakes were larger than 1.5 cm, in a total sample of 3,101), the percentage of large flakes from Feature 13 was relatively high (about 62 percent of flakes were larger than 1.5 cm, in a total sample of 351). Thus, while most of the lithic reduction conducted at the site yielded relatively small flakes, apparently pertaining to late-stage bifaces and tools, Feature 13 is associated with a significantly earlier stage of lithic reduction yielding relatively large flakes. This inference is supported by the high number of decortication flakes in Feature 13, where 13 decortication flakes were found, compared to only 3 found in the larger assemblage outside of Feature 13.
A varied artifact assemblage was recovered, including pottery and various lithic tools, cores, and bifaces. The pottery assemblage is highly fragmented. One flat-lipped rim sherd was identified; no rims with other shaped lips were found. The flat lip shape and the surface treatments of the body sherds (interiors plain, exteriors plain, cord-marked or smoothed-over cord), and wall thickness (range 5.0–11.9 mm, with 75 percent between 9.0 and 10.9 mm) are consistent with pottery of the twelfth and thirteenth-centuries as observed at other Hudson Valley sites and in nearby regions (Funk 1976:31–36, 98–112; Rieth 2002a:222; see Ritchie 1969 and Ritchie and MacNeish 1949 for general discussions of Late Woodland ceramic attributes).
It has already been noted that most of the lithic debitage falls into small size grades, indicating a predominance of late-stage lithic reduction, but that Feature 13 is an exception to this trend, representing a relatively early-reduction-stage work area. In addition to abundant evidence of chert tool manufacturing and maintenance, a varied group of scrapers, choppers, hammerstones and other tools, as well as cores and bifaces was found. The diverse stone tool assemblage (Table 5.3) provides evidence that a wide range of activities was performed at the site. At the same time, it is notable that projectile points and point fragments are scarce in this assemblage, and none of the distinctive, Late Prehistoric, triangular Levanna points were found. The most distinctive point recovered is highly atypical of Late Prehistoric lithic assemblages: a stemmed point of Cresap or Adena form, made from an exotic light-colored stone (apparently Flint Ridge chalcedony) (Converse 1994; this chapter, Figure 5.10). Cresap and Adena points have been found elsewhere in the Hudson Valley, including the Zimmerman Rockshelter very near this site (Funk 1976:34–37, 108–109, 277–278). Although possibly a stray find at Concentration 23B.1, the base has been reworked into a graver tip, suggesting rather that this particular artifact has a potentially long, multi-use history, and one that perhaps involved loss or caching during the Early Woodland period, rediscovery, and transformation for a new use-life during the Late Prehistoric.
Concentration 23B.1 Stone Tool Assemblage.
Concentration 23B.1, stemmed projectile point with base reworked into graver.
The spatial distribution of several artifact classes provides insights into the organization of activities at Concentration 23B.1. Pottery occurs in two main clusters: one on the north side of the possible shelter and concentrated around Feature 8, an area of burned earth; and a second on the south side of the possible shelter, including a spread 3 m to the west (Figure 5.11). The radiocarbon-dated hearth Feature 5 and pottery-rich Feature 6 occur in this southern area, which also contains the majority of the pottery outside of Feature 6. Fire-cracked rock is not very common at this site, but tends to occur west and south of the possible shelter, generally in the southern area of pottery concentration (Figure 5.12). The utilized flake distribution is somewhat different than the pottery and fire cracked rock, being centered more in the area of the possible shelter, and extending north to the general vicinity of the early-stage lithic reduction area (Figure 5.13). Thus, the spatial distributions of pottery and fire cracked rock, both associated with cooking, overlap to a considerable extent, while the distribution of utilized flakes both covers some of these areas and diverges more strongly to the north, including the early stage lithic reduction area at Feature 13. These distributions indicate that cooking to some notable extent occurred outside and south of the inferred shelter, while the use and discard of utilized flakes appear to be more strongly centered in and immediately around the shelter.
Concentration 23B.1, pottery density distribution (right), with feature distribution for comparison (left).
Concentration 23B.1, fire cracked rock distribution (right), with feature distribution for comparison (left).
Concentration 23B.1, utilized flake distribution (right), with feature distribution for comparison (left).
Analysis of flotation samples from Features 5 (1950 ml) and 6 (2000 ml) by McKnight (2008) showed abundant carbonized white oak and other deciduous wood samples but no subsistence remains. Gall and stem fragments found in Feature 5 are described as possible tinder used as fire starter (McKnight 2008:176). Uncarbonized seeds and oak flower fragments from these samples are considered relatively recent and intrusive. The brown organic stratum contained small flecks of carbonized, deciduous wood charcoal and melted silica spheres resulting from the burning of silica-rich plants, such as grasses. Recently deposited uncarbonized seeds of several native species also were present (McKnight 2010).
The lack of subsistence remains is perhaps not unusual because the sample sizes available were small, and in any event, the presence and preservation of cultigens or other food remains may be fortuitous in many circumstances. Considering a wide variety of Northeastern contexts, even when corn (for example) is present at an archaeological site, it often is not abundant (see discussions and data in Asch Sidell 2008; Cassedy and Webb 1999; Chilton 2002, 2008; Largy et al. 1999; Miroff 2002). That is to say, it does not necessarily follow that because corn or other subsistence remains were not found at Concentration 23B.1, corn was not grown there or people did not consume food there.
Concentration 23B.1 appears to represent an episode of reoccupation within a landscape that had been used relatively intensively during the Early and Middle Archaic periods, and sporadically afterward by logistically organized parties. Two 14C dates from the same feature processed by different labs provide a consistent radiocarbon age estimate of approximately A.D. 1200, or perhaps approximately cal. A.D. 1250. Flotation samples have identified hearth wood as white oak and other deciduous species, while the soil below the plow zone contains small pieces of wood charcoal and melted silica from burning plants, perhaps grasses. The artifact assemblage indicates a varied range of camp activities, but has no substantial evidence that the site was a specialized base for hunters. Despite the abundance of chert in Coxsackie, the debitage assemblage in general suggests the presence of a curated artifact assemblage. The use of a curated, highly transportable stone assemblage suggests that people arrived at this site after traveling from a previous occupation some distance away. At the same time, there is some situational evidence for early-stage lithic reduction, perhaps directed to the production of an assemblage of expedient, utilized flake tools. This site probably was occupied seasonally during late spring, summer, and/or early fall; however, its features and artifacts represent a substantial episode of short-term land use inland from the Hudson River, rather than ephemeral evidence of a random, wandering, or relatively insignificant visit.
SUMMARY AND CONCLUSION
The investigation of Concentration 23B.1 has provided a rare glimpse into life at a seasonal, inland site of the twelfth to thirteenth centuries A.D. within the Hudson drainage. This small site contained a modest, seemingly insubstantial shelter, plus cooking facilities and small storage or processing pits. A deeper, unplowed stratum immediately below the plow zone provides evidence of burning, perhaps during an early stage of site use, since it contains charcoal flecks and melted silica spheres of the sort that may be associated with burned grass. The artifact assemblage is diverse but contains very few projectile points or point fragments, suggesting that little hunting may have been conducted from this site. Different aspects of the debitage assemblage seem to indicate early-stage chert reduction oriented to the generation of expedient flake tools, plus the frequent repair and maintenance of late-stage, formally manufactured tools. The highly fragmented ceramics provide a weak temporal-cultural signature. This sort of weak signature, exacerbated by the paucity of projectile points, may be a factor contributing to low archaeological visibility during this period, particularly in backcountry settings.
A review of literature on the Late Prehistoric period in the Hudson Valley indicates that the best-understood sites occur along the Hudson or in other riverine settings such as along Fish Creek and the Roeliff Jansen Kill, and that the beginning and end of the Late Prehistoric period are better represented in the archaeological record than the middle of this period, including the twelfth to thirteenth centuries. The general paucity of archaeological sites of the twelfth to thirteenth centuries includes large sites as well as small; thus the alternative argument that Concentration 23B.1 was a satellite camp of a larger village cannot be advanced, at least not at this time. It may well be a rather ordinary seasonal residential site of this period. Importantly, Concentration 23B.1 provides information demonstrating occupational continuity rather than hiatus during this time frame, in support of other weak indications of occupation elsewhere in the region (Funk 1976:300–302).
The present paper also reviews information on the possible effects of the Medieval Warm Period on Hudson drainage regional settlement and subsistence systems. This is an area of incipient research, yet one that has already shown marked environmental effects of the Medieval Warm Period in the lower Hudson region. Other data from the lower Mohawk area indicate an increase in storminess beginning in the twelfth century, possibly one of the most droughty times of the Medieval Warm Period based upon southwestern and midwestern U.S. evidence. More information on the issue of the possible regional occurrence of drought is desirable; nonetheless, the mapping of twelfth to thirteenth-century droughts provided by Benson et al. (2009:477) extends marginally to the Hudson valley, while Pederson et al. (2005) consider the lower Hudson (Piermont Marsh) indications of Medieval Warm Period drought to be consistent with other data from southward along the mid-Atlantic coast. On the other hand, data from central New York indicate increased rainfall in the Cayuga Lake catchment during the Medieval Warm Period (Mullins et al. 2011).
In accordance with this information, a hypothetical sequence of social-ecological system changes is proposed: A native agricultural system had become significant during the early part of the Late Prehistoric (A.D. 700–1000 or 1100). However, as time passed, this system was increasingly put at risk by a variety of possible climate change effects, including increased storminess, flooding, erosion, drought, wind, and hail. The adverse effects of climate change were perhaps most pronounced during the twelfth to thirteenth centuries, as they were throughout much of North America. This is consistent with evidence from this time frame of increased storminess in the Mohawk drainage, increased incidences of fire and erosion in the lower Hudson drainage, and increased salinity in the lower Hudson estuary.
Native people of the Hudson Valley may have responded to these problems with great resilience in terms of settlement pattern and subsistence changes. The larger or more sedentary communities noted during the early part of the Late Prehistoric period may have broken into extended family segments and become more mobile, using a wide variety of sites, and using the backcountry more intensively. Sites occupied previously by hunter-gatherer ancestors continued to be used, although efforts to farm on the floodplains were probably part of the overall strategy, and probably partially successful. In addition, however, the cultivation of clayey upland soils may have provided further diversity in the food production system, since these soils would retain water longer than sandy or silty soils during droughts. This argument is relevant to the interpretation of Concentration 23B.1 and other Late Prehistoric sites that may be found within the lake plain setting of Coxsackie. For example, although archaeologists often focus on well drained floodplain soils in referring to native corn horticulture, Figure 5.14 is a recent (2010) photograph showing modern corn growing on somewhat poorly drained clayey soil in Coxsackie. The Kingston and Rhinebeck series soil at the site pictured is the same as the soil type at Concentration 23B.1, and presumably these soils were suitable for cultivation during the Late Prehistoric period.
View of corn growing on somewhat poorly drained Kingsbury and Rhinebeck series soil in Coxsackie, August 2010.
The adjustments just cited were probably made in the context of backcountry land use involving a long-term process of land clearing that increased forest edge and related biodiversity, and may have offered attractive open sites for gardens and small settlements. That is, the long-term effect of continued clearing of forest would have been to make hunting and gathering more productive, and to diffuse risk within the agricultural system. Clearing in the backcountry of Coxsackie would have made clayey soils available for planting. This context, or some aspect of it, is a plausible setting for the seemingly unusual archaeological site, Concentration 23B.1.
Several regions of North America, including much of California, the Southwest, and the central Mississippi Valley, saw such severe drought effects during the Medieval Warm Period that subsistence and related economic patterns were significantly disrupted, social and political organization were altered (in many cases fragmented and reorganized), and residential sites, farm complexes, and large portions of regions were abandoned. These effects in these regions have been seen as elements of societal collapse, although recent scholarship—and particularly application of Resilience Theory—is also investigating the strategies that societies have used to cope with mismatches between environmental productivity and social scale during extreme episodes of climate change.
In the Northeast, there are no indications that the effects of the Medieval Warm Period were as extreme as in the Southwest or within the Cahokia polity, in part because droughts do not appear to have been as severe in the Northeast (see Benson et al. 2009:477), but also because northeastern societies, organized on a smaller scale and with significantly less administrative hierarchy, could respond more flexibly to adversity caused by extended droughts, severe floods, and violent storms. Nonetheless, this flexibility in the Northeast probably did involve higher mobility and segmentation into smaller co-residential groups. Chilton’s (1999, 2002, 2008) concept of mobile farmers appears to be very compatible with these concepts of long-term, Late Prehistoric subsistence-settlement dynamics.
Over the last several years, a number of archaeologists working in New York have examined upland and backcountry sites with an eye to understanding their importance within settlement and land-use strategies (Abel 2000; Diamond 1996; Miroff 2002; Rieth 2007). In addition, Rieth (2009) has looked at a larger landscape containing early Late Prehistoric sites to see how spatial and temporal patterns shifted between residential and limited activity uses. In the upper Hudson Valley, the significant questions raised by the location, features, and artifact assemblage of Concentrations 23B.1 include how did this site fit into a land-use pattern conventionally thought to be dominated by river and floodplain-oriented residential sites? To be sure, a broad range of sites within well-defined temporal and spatial scales will need to be studied in order to determine whether shifts over time occurred between settlement patterns containing relatively large, relatively visible settlements, and patterns composed of small, less visible sites.
Concentration 23B.1 has a number and variety of archaeological features, fairly abundant pottery, a diverse stone tool assemblage, and an abundance of stone waste material, all suggestive of a seasonal residence. If this is indeed a residential site, some of the questions for future study become: How common are small Late Prehistoric residential sites in the upper Hudson region? How do they vary internally as a group? How do their locations, assemblages, and features vary over time, and with respect to the developing body of evidence on ancient climate change?
I am grateful to the Greene County Industrial Development Agency, its former Executive Director Alexander “Sandy” Mathes, and its Interim Executive Director Rene Van Schaack for the generous support of the archaeological investigation of Greene Business and Technology Park and Kalkberg Commerce Park; encouragement of the broad dissemination of this information; and arranging for a series of public presentations on this work in Coxsackie. I also thank the New Baltimore Conservancy for inviting me to present the results of the investigations, and Robert Hallock, President of the Greene County Historical Society, for providing important information about Pieter Bronck and the Bronck Patent. Derek Almstead served as field director of the excavation of Concentration 23B.1. Tim Lloyd regularly assisted in guiding the 23B.1 excavation. Much of the artifact cataloging and analysis was conducted by Jon Vidulich and Sarah Vidulich. Graphics were produced by Kerry Nelson and Sarah Vidulich, assisted by Curtin Archaeological GIS Coordinator Andrew Farry. I also thank Christina Rieth and John Hart for encouraging the writing of this paper and for providing helpful comments, Roger Moeller for providing a significant view of Late Woodland ecology, and three anonymous reviewers who helped me to correct, clarify, and support the information provided here. I alone bear responsibility for any errors of fact or interpretation.
Abel, T. 2000. The Plus Site: an Iroquoian Remote Camp in Upland Tompkins County. North American Archaeologist 21:181–215.
Asch Sidell, N. 2008. The Impact of Maize-based Agriculture on Prehistoric Plant Communities in the Northeast. In Current Northeast Paleoethnobotany II, edited by J. P. Hart, pp. 29–51. New York State Museum Bulletin 512. The University of the State of New York, Albany.
Bender, S. J., and H. J. Brumbach. 1986. Winney’s Rift: A Late Woodland Village Settlement in the Upper Hudson River Valley. The Bulletin and Journal of Archaeology for New York State 92:1–8.
Bender, S. J., and E. V. Curtin. 1990. A Prehistoric Context for the Upper Hudson Valley: Report of the Survey and Planning Project. Department of Sociology, Anthropology, and Social Work, Skidmore College, Saratoga Springs, New York. On file, New York State Office of Parks, Recreation, and Historic Preservation, Peebles Island State Park, Waterford, New York.
Benson, L. V., T. R. Pauketat, and E. R. Cook. 2009. Cahokia’s Boom and Bust in the Context of Climate Change. American Antiquity 74:467–483.
Boyle, R. H. 1979. The Hudson River: A Natural and Unnatural History. Norton, New York.
Bradley, J. W. 2007. Before Albany: An Archaeology of Native-Dutch Relations in the Capitol Region, 1600–1664. New York State Museum Bulletin 509. The University of the State of New York, Albany.
Brewer, F. I. 2001. The Goes/Van Derzee Farm Site, Albany County, Bethlehem, New York. The Bulletin, Journal of the New York State Archaeological Association 117:36–48.
Broad, W. 1993. Soil Survey of Greene County, New York. United States Department of Agriculture, Soil Conservation Service, in cooperation with Cornell University Agricultural Experiment Station, Catskill, New York.
Brumbach, H. J., and S. Bender. 2002. Woodland Period Settlement and Subsistence Change in the Upper Hudson Valley. In Northeast Subsistence Settlement Change A.D. 700–1300, edited by J. P. Hart and C. B. Rieth, pp 227–239. New York State Museum Bulletin 496. The University of the State of New York, Albany.
Cassedy, D. F. 1992. Native American Interaction Patterns and Lithic Acquisition Strategies in Eastern New York and Southern New England. Unpublished Ph. D. dissertation, Department of Anthropology, Binghamton University, Binghamton, New York.
Cassedy, D., and P. Webb. 1999. New Data on the Chronology of Maize Horticulture in Eastern New York and Southern New England. In Current Northeastern Paleoethnobotany, edited by J. P. Hart, pp. 85–99. New York State Museum Bulletin 494. The University of the State of New York, Albany.
Chapman, J., H. R. Delcourt, and P. A. Delcourt. 1989. Strawberry Fields, Almost Forever. Natural History, 9 September:51–58.
Chilton, E. S. 1999. Mobile Farmers of Pre-Contact Southern New England: The Archaeological and Ethnohistoric Evidence. In Current Northeast Paleoethnobotany, edited by J. P. Hart, pp 157–176. New York State Museum Bulletin 494. The University of the State of New York, Albany.
Chilton, E. A. 2002. “townes they have none”: Diverse Subsistence and Settlement Strategies in Native New England. In Northeast Subsistence Settlement Change A.D. 700-1300, edited by J. P. Hart and C. B. Rieth, pp 289–300. New York State Museum Bulletin 496. The University of the State of New York, Albany.
Chilton, E. A. 2008. So Little Maize, So Much Time: Understanding Maize Adoption in New England. In Current Northeast Paleoethnobotany II, edited by J. P. Hart, pp. 53–59. New York State Museum Bulletin Series 512. The University of the State of New York, Albany.
Converse, R. W. 1994. Ohio Flint Types. The Archaeological Society of Ohio, Columbus.
Cordell, L. 1997. Archaeology of the Southwest, second edition. Academic Press, San Diego.
Cronon, W. 1983. Changes in the Land: Indians, Colonists, and the Ecology of New England. Hill and Wang, New York.
Curtin, E. V. 2008. Phase 3 Archaeological Data Recovery Report, Greene Business and Technology Park, Town of Coxsackie, Greene County, New York. Curtin Archaeological Consulting Inc., Ballston Spa, New York. Submitted to Greene County Industrial Development Agency, Coxsackie, New York. On file, New York State Office of Parks, Recreation, and Historic Preservation, Peebles Island State Park, Waterford, New York.
Curtin, E. V., K. L. Nelson, and J. Schreyer. 2008. Strategies for Investigating and Interpreting Small Prehistoric Sites and Low Density Artifact Distributions: Examples from the Hudson Drainage. In Current Approaches to the Analysis and Interpretation of Small Lithic Sites in the Northeast, edited by Christina B. Rieth, pp. 41–61. New York State Museum Bulletin 508. The University of the State of New York, Albany.
Day, G. M. 1953. The Indian as an Ecological Factor in the Northeastern Forest. Ecology 34(2):329–346.
Delcourt, P. A., H. R. Delcourt, C. R. Ison, W. F. Sharp, and K. J. Gremillion. 1998. Prehistoric Use of Fire, the Eastern Agricultural Complex, and Appalachian Oak Chestnut Forests: Paleoecology of Cliff Palace Pond, Kentucky. American Antiquity 63:263–278.
Delcourt, P. A., and H. R. Delcourt. 2004. Prehistoric Native Americans and Ecological Change: Human Ecosystems in Eastern North America Since the Pleistocene. Cambridge University Press, New York.
Denevan, W. M. 1992. The Pristine Myth: The Landscape of the Americas in 1492. Annals of the Association of American Geographers 82(3):369–385.
Diamond, J. 2005. Collapse: How Societies Choose to Fail or Succeed. Penguin Books, New York.
Diamond, J. E. 1996. Terminal Late Woodland/Early Contact Period Settlement Patterns in the Mid-Hudson Valley. Journal of Middle Atlantic Archaeology 12:95–111.
Doolittle, W. E. 1992. Agriculture in North America on the Eve of Contact: A Reassessment. Annals of the Association of American Geographers 82(3):386–401.
Dunn, S. W. 1994. The Mohicans and Their Land 1609–1730. Purple Mountain Press, Fleischmann’s, New York.
Dunn, S. W. 2000. The Mohican World 1680-1750. Purple Mountain Press, Fleischmann’s, New York.
Engelbrecht, W. 2003. Iroquoia: The Development of a Native World. Syracuse University Press, Syracuse.
Fagan, B. 2008. The Great Warming: Climate Change and the Rise and Fall of Civilizations. Bloomsbury Press, New York.
Forman, R. T. T., and E. W. B. Russell. 1983. Evaluation of Historical Data in Ecology. Bulletin of the Ecological Society of America 64:5–7.
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.
Funk, R. E. 2004. An Ice Age Quarry-Workshop: The West Athens Hill Site Revisited. New York State Museum Bulletin 504. The University of the State of New York, Albany.
Funk, R.E., and P. Lord, Jr. 1972. Two Stratified Alluvial Sites in the Upper Hudson Valley. Pennsylvania Archaeologist 42:27–61.
Hartgen Archeological Associates Inc. 1983. Mechanicville Road (209) Site: Results of Mitigation Procedures, Waterford, Saratoga County, New York (C-36-644). Hartgen Archeological Associates Inc., Troy, New York.
Johnson, E. S. 1996. Discovering the Ancient Past at Kampoosa Bog, Stockbridge, Massachusetts. University of Massachusetts Archaeological Services, Amherst.
Jones, E. L. 2005. Review of Prehistoric Native Americans and Ecological Change: Human Ecosystems in Eastern North America Since the Pleistocene by P. A. Delcourt and H. R. Delcourt. PaleoAnthropology, May:10–12.
Judge, W. J. 1989. Chaco Canyon-San Juan Basin. In Dynamics of Southwest Prehistory, edited by Linda S. Cordell and George J. Gumerman, pp. 209–261. Smithsonian Institution Press, Washington.
Largy, T. B., L. Lavin, M. E. Mozzi, and K. Furgerson. 1999. Corncobs and Buttercups: Plant Remains from the Goldkrest Site. In Current Northeast Paleoethnobotany, edited by J. P. Hart, pp. 69-84. New York State Museum Bulletin No. 494. The University of the State of New York, Albany.
Larson, D. O., H. Neff, D. A. Graybill, J. Michaelsen, and E. Ambos. 1996 . Risk, Climatic Variability, and the Study of Southwestern Prehistory: An Evolutionary Perspective. American Antiquity 61:217–241.
Lavin, L. 2004. Mohican/Algonquian Settlement Patterns: An Archaeological Perspective. In The Continuance – An Algonquian Peoples Seminar, edited by S. W. Dunn, pp. 19–28 New York State Museum Bulletin 501. The University of the State of New York, Albany.
Lavin, L., M. E. Mozzi, J. W. Bouchard, and K. Hartgen. 1996. The Goldkrest Site: An Undisturbed, Multi-Component Woodland Site in the Heart of Mahikan Territory. Journal of Middle Atlantic Archaeology 12:113–129.
Mann, C. C. 2006. 1491: New Revelations of the Americas Before Columbus. Vintage Books, New York.
McKnight, J. W. 2008. Flotation-recovered Plant Remains from Concentration 23B.1, Dunn and Prescott Properties, Greene Business and Technology Park, Town of Coxsackie, Greene County, New York. Appendix E in Phase 3 Archaeological Data Recovery Report, Greene Business and Technology Park, Town of Coxsackie, Greene County, New York, Volume 2, assembled by Edward V. Curtin and Kerry L. Nelson. Curtin Archaeological Consulting Inc., Ballston Spa, New York. Submitted to Greene County Industrial Development Agency, Coxsackie, New York. On file, New York State Office of Parks, Recreation, and Historic Preservation, Peebles Island State Park, Waterford, New York.
McKnight, J. W. 2010. Analysis of a Single Flotation Sample from Non-Feature Contexts within Concentration 23B.1, Dunn and Prescott Properties, Greene Business and Technology Park, Town of Coxsackie, Greene County, New York. On file, Curtin Archaeological Consulting Inc., Ballston Spa, New York.
Miroff, L. E. 2002. Upland Land Use Patterns during the Early Late Prehistoric (A.D. 700–1300). In Northeast Subsistence Settlement Change A.D. 700–1300, edited by J. P. Hart and C. B. Rieth, pp 193–208. New York State Museum Bulletin 496, Albany.
Miroff, L. E., and T. D. Knapp (Editors). 2009. Iroquoian Archaeology and Analytic Scale. University of Tennessee Press, Knoxville.
Moeller, R. W. 1996. Some Thoughts on Late Woodland Ecology. In A Northeastern Millennium: History and Archaeology for Robert E. Funk. Journal of Middle Atlantic Archaeology 12:61–66.
Mullins, H. T., W. P. Patterson, M. A. Teece, and A. W. Burnett. 2011. Holocene Climate and Environmental Change in Central New York (USA). Journal of Paleolimnology 45:254–256.
Nelson, M. C., M. Hegmon, S. Kulow, and K. G. Schollmeyer. 2006. Archaeological and Ecological Perspectives on Reorganization: A Case Study from the Mimbres Region of the U. S. Southwest. American Antiquity 71:403–432.
Parker, A. C. 1924. The Great Algonkian Flint Mines at Coxsackie. Researches and Transactions of the New York State Archeological Association IV(4). Rochester, New York.
Patterson, W. A. III, and K. E. Sassaman. 1988. Indian Fires in the Prehistory of New England. In Holocene Human Ecology in Northeastern North America, edited by G. P. Nicholas, pp. 107–135. Plenum Press, New York.
Pederson, D. C., D. M. Peteet, D. Kurdyla, and T. Guilderson. 2005. Medieval Warming, Little Ice Age, and European Impact on the Environment During the Last Millennium in the Lower Hudson Valley, New York. Quaternary Research 63:238–249.
Pyne, S. J. 1997 . Fire in America: A Cultural History of Wildland and Rural Fire. University of Washington Press.
Redman, C. L. 2005. Resilience Theory in Archaeology. American Anthropologist 107:70–77.
Redman, C. L., and A. P. Kinzig. 2003. Resilience of Past Landscapes: Resilience Theory, Society, and the Longue Durée. Conservation Ecology 7(1):14 [online] URL: http//www.consecol.org/vol7/iss1/art14.
Rieth, C. B. 2002a. Early Late Prehistoric Settlement and Subsistence Diversity in the Southern Tier Region of New York. In Northeast Subsistence Settlement Change A.D. 700–1300, 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. 2002b. Introduction. In Northeast Subsistence-Settlement Change, A.D. 700–1300, edited by J. P. Hart and C. B. Rieth, pp. 1–10. New York State Museum Bulletin 496, The University of the State of New York, Albany.
Rieth, C. B. 2007. The Catskill I and II Sites: Two Early Late Prehistoric Upland Camps in Eastern New York. The Bulletin, Journal of the New York State Archaeological Association 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. 1958. An Introduction to Hudson Valley Prehistory. New York State Museum and Science Service Bulletin No. 367, The University of the State of New York, Albany.
Ritchie, W. A. 1969. The Archaeology of New York State, second edition. Natural History Press, Garden City, New York.
Ritchie, W. A., and R. S. MacNeish. 1949. The Pre-Iroquoian Pottery of New York State. American Antiquity 15:97–124.
Russell, E. W. B. 1983. Indian Set Fires in the Forests of the Northeastern United States. Ecology 64:78–88.
Schuldenrein, J. 1997. Summary Geoarchaeological Observations at the Goldkrest Site, CNG TL-470, East Greenbush, New York. In Stage III Archaeological Investigations: The Goldkrest Site, CNG TL-470, East Greenbush, New York, Appendix 7. Archeological Research Specialists, Meriden, Connecticut. Submitted to Consolidated natural Gas Transmission Corporation, Clarksburg, West Virginia. On file, New York State Office of Parks, Recreation, and Historic Preservation, Waterford, New York.
Snow, D. R. 1994. The Iroquois. Blackwell Publishers, Oxford.
Sopko, J. 2007. The Outskirts of Albany: Recent Discoveries of Dutch and Mahican Contact and Late Woodland Sites on Papscanee Island. Abstracts of Papers Presented at the Colloquium: Soldiers, Cities and Landscapes: Papers in Honor of Charles L. Fisher, New York State Museum, Albany, December 1, 2007.
Sopko, J., and V. Schmitt. 2009. PIN A031.03.701 Empire Rail Corridor, Port of Rensselaer Southern Connector Road with Route 9J, Alternate C, Cultural Resources Site Examination Report for Papscanee Creek Site 3 (NYSM Site 11694), Town of East Greenbush, Rensselaer County, New York. Cultural Resource Survey Program, New York State Museum, The State University of New York, Albany.
Starbuck, D. R. 2004. Rangers and Redcoats on the Hudson. University Press of New England, Hanover, New Hampshire.
Thompson, V. D., and J. A. Turck. 2009. Adaptive Cycles of Coastal Hunter-Gatherers. American Antiquity 74:255–278.
Vedder, J. Van Vechten. 1985 . Official History of Greene County, New York. Greene County Board of Supervisors, Catskill, New York. 1985 facsimile ed. Hope Farm Press, Cornwall, New York.
Webster’s Dictionary. 1996. Webster’s II New Riverside Dictionary, revised edition. Houghton Mifflin Company, Boston.
White, I. R., and D. T. Rodbell. 2010. A ~ 1000 Year Record of Sedimentation in Collins Lake as Evidence of Local Storminess and Flooding on the Mohawk River (New York). http://minerva.union.edu/garverj/mohawk, accessed April 14, 2010.