Surface Collecting in the Intertidal Zone
I began collecting chipped stone (“lithic”) artifacts from the shoreline at the Rye Marshlands Conservancy in 1979, soon after I was hired as an Assistant Professor of Anthropology at SUNY Purchase. I left SUNY in 1986, but I continued visiting the Conservancy while I still lived in Westchester County until 1994, when my career as a contract archaeologist took me to northern Virginia. I was able to walk the shore only rarely after that, during occasional visits to family and friends over the following decade.
From 1979 to 1994, I would visit the Conservancy once or twice a month, usually on weekends, except during the coldest weeks of winter and the hottest days of summer. The artifacts were only visible on the surface for a few hours at low tide, so I had to check the daily tide charts to see if there would be a low tide during daylight hours on any given day. Back then, there was no GPS technology, so I recorded the location of each notable artifact (mostly, whole or fragmentary dart or arrow points) on a sketch map. The pinpoint accuracy of GPS would probably have been overkill, anyway, because once the artifacts were dislodged from wherever they had been stuck—whether in the eroding marsh bank or in the bottom sediments of Milton Harbor (I could never be sure which)—the water could have washed them back and forth across the surface repeatedly before I spotted them.
This was not a Pompeii-like situation; the points had not lain undisturbed where some Native American had discarded or lost them, for several thousand years until the day I arrived. Nevertheless, there were a few cases that suggested the water had not carried the artifacts too far away from their original location. In one particularly memorable instance, I found the mid-section of a late-stage quartz preform (an unfinished rough-out), probably for a 3,000-year old Orient Fishtail point; many months afterward, I found the tip of the preform, which refit to the other fragment perfectly along the transverse fracture. The tip lay only a few feet away from the spot where I had found the mid-section. No doubt, this accidental break was the reason the ancient toolmaker had tossed the preform away. However, I also found, on different occasions, 6,000- and 600-year-old artifacts in the same area.
Limitations of the Archaeological Record
This example illustrates the fundamental limitations of the archaeological data from the Marshlands. There is no stratigraphic vertical separation of artifacts of different age; they all lie on the same modern surface. Stone tool-making methods changed very little from the arrival of ancestral Paleoindians about 13,000 years ago until the Indians adopted metal tools and guns to replace their traditional toolkit in the mid-1600s. Thus, the chips or flakes (“debitage”) created in that process generally cannot be assigned to any particular time span or culture. The only (very rare) exception is the “channel flakes” that were incidentally produced when long flakes or “flutes” were driven off the bases of Paleoindian spearpoints; that type of debitage was no longer made after about 12,000 cal BP.
Another problem at the Marshlands is the absence of buried, sealed “features” such as fireplaces (hearths) or storage pits. When archaeologists excavate a buried hearth, we can recover charcoal, burned nutshells, and burned bone fragments. All of these organic (once living) materials can be dated using radiocarbon (14C). The same kinds of samples may be recovered from storage pits (often re-filled with debris), but the pits also sometimes contain carbonized seeds, maize kernels, or larger, identifiable animal remains that can inform us about the diets of prehistoric people. If stone artifacts are found in or very close to dated features, we can confidently apply those radiocarbon dates to those artifacts (which, being inorganic, cannot be dated using 14C).
What Can We Learn from “Projectile Points”?
Although the basic techniques for making stone tools did not change over thousands of years, the shapes of those tools did change. Archaeologists are particularly interested in the changing shapes of “projectile points.” To a lay person, these are “arrowheads.” However, we archaeologists think that the bow and arrow was adopted by indigenous peoples of North America very late in prehistory, probably after cal (calibrated) AD 500 in most regions. In the Northeast, we assume that the triangular points made after cal AD 800, and perhaps also the corner-notched points made during the preceding few centuries, really were arrowheads. When Europeans began to explore the coast in the 16th and 17th centuries the indigenous peoples (“Indians”) were using the same sort of stone triangles as the tips of arrows shot from long bows. However, the stone points made before ca. cal AD 800 were, we think, the tips of darts that were thrown with spearthrowers (or atlatls, as they were called by the Aztec of Mexico). I emphasize that we don’t really know for sure. In dry caves of the Great Basin and Southwest, and in the recently melting ice of the Yukon, the wooden shafts of these weapons are sometimes preserved, and we can tell from their butt ends if they were made to be propelled from the string of a bow or the hook of an atlatl. In the generally acidic, often wet soils of the Northeast, artifacts made of wood, bark, leather, fibers, basketry, all decayed within centuries, even decades after they were discarded. There are no preserved bows or atlatls, so we don’t know exactly when the bow replaced the spearthrower, or if they co-existed for an extended period. It is particularly perplexing that some triangular dart (?) points, made about 6,000 years ago, cannot be distinguished from the post-AD 800 arrow points.
When the first people arrived, they encountered an amazing menagerie of giant mammals such as mastodont, stag-moose, and ground sloths. By 12,700 cal BP, all of these “megafauna” were extinct, probably in part because Paleoindians had hunted them using fluted points. From that time until European Contact, the newly established Eastern deciduous forests contained an unchanging suite of relatively large mammals that included white-tailed deer, elk, and black bear. The large predators were wolves and cougars, which humans would not usually have considered edible. In some periods when grasslands expanded, bison may have briefly intruded into the region, but there is no evidence for this among identifiable bones (which, like other organic materials, are not often well preserved in the Northeast). Some “projectile points” may have functioned occasionally or primarily as knives, but most of them were designed to penetrate the hides of hunted mammals. Given that this was their primary function, and the same suite of animals was available as prey for 12,000 years, why would the shape of dart points have changed at all after that time? And if they did change, shouldn’t those changes have followed an evolutionary path toward ever-increasing efficiency?
We must not ignore another likely use of dart points, to kill other people in warfare. The oldest skeleton in North America with an embedded stone point is Kennewick Man, who lived in Washington State about 9000 cal BP. In the Northeast, the earliest direct evidence of warfare dates from only about 5800-4500 cal BP—but that may only reflect the small sample of preserved ancient bones. Male skeletons buried on Frontenac Island in Cayuga Lake in western New York exhibit unmistakable signs of violence, such as a Lamoka dart point embedded in a rib, and a point tip stuck in a skull.
Experiments have shown that a wooden spear or dart with a fire-hardened tip is as capable of penetrating hides as a stone-tipped dart. What then, is the advantage of a stone point? If it shatters after entering the animal’s body, the sharp fragments can slice through blood vessels as the animal moves, and the bleeding will hasten death.
Basically, a dart point needs to have a sharp tip, to be strong enough not to break on contact, but fragile enough to snap after entering the target. Thus, we don’t expect much deviation in shape at the tip. At the base, the point has to be attached securely to its wooden shaft, or to a detachable foreshaft. There are basically two ways to do this: the pointed or convergent base can be plugged into a round hole, or the thinned base can be set into a narrow slot. Some sort of glue, such as pine tar, can be used to secure the attachment, or the base can be tied on using sinew or thin string. In that case, usually a notch has been removed from the side or corner of the point’s base to hold the string, and the notch has been ground smooth so it won’t cut through.
Surprisingly—given these limited possibilities for variability as dictated by functional constraints-- the shapes of dart points changed repeatedly in the Northeast over the course of 12,000 years, and not in a way that can be interpreted as a trend toward greater efficiency. For example, corner-notched points were used about 10,500 years ago, 6,000 years ago, and again 1400 years ago.
Around 3,000 years ago, people made Orient Fishtail points at the Marshlands. These were not notched. At about the same time, people in central and western New York were making Meadowood points, which were side-notched. I have found only eight Meadowood points at the Marshlands, as opposed to 97 intact or nearly complete Fishtail points. The Meadowood points from the Marshlands are all made of chert; the great majority of all Meadowood points in the Northeast are made of Onondaga chert, a variety that outcrops in western New York. In sharp contrast, most of the Orient points from the Marshlands are made of quartz, although the craftsmen (they probably were men, although that is not an easily testable assumption) also used quartzite, chert, and various metamorphic rocks. When an archaeologist sees these sorts of clear contemporaneous regional differences, we are pretty comfortable attributing them to distinct ethnic/cultural groups.
However, we do not have a well-developed set of theories to explain the observed changes of point styles over time. Usually, a recognizable style lasts for about 500 to 1500 years until it is replaced, often by a type that appears radically different in shape and often also in the preferred raw material. For example, about 4000 years ago, broadspears made of grainy metamorphic rocks or chert replaced small stemmed points that were made predominantly of quartz. I should note that we can place each of these styles (or types) in a time-frame because nearly identical specimens have been found within, or very near, features containing organic material that could be radiocarbon-dated. I can say, therefore, that the Orient Fishtail points at the Marshlands are about 3000 years old, because they have been closely associated with dated fireplaces on Long Island and in New Jersey and eastern Pennsylvania.
Before about 1970, archaeologists would have confidently ascribed new artifact styles to the arrival of new populations that migrated into the region. Since then, we have become very hesitant to invoke migration and population replacement. Instead, archaeologists now prefer to talk about local innovation and adaptation to changing environments by resilient indigenous populations. However, recent research on ancient genes has shown that, in Eurasia, migrations have been frequent, and many near-total population replacements have occurred. For example, Neolithic farmers migrating westward from Anatolia largely replaced the Mesolithic hunter-gatherers of Europe between ca. 8500 and 6000 cal BP. Similar research on ancient genomes in the US and Canada has been derailed because Native Americans have been very reluctant to allow sampling of the DNA of both living people and the ancient skeletons to which they now control access.
I participated in a 2017 study of the mitochondrial DNA extracted from bones of the late prehistoric Beothuk and the 4500-year-old Maritime Archaic people of Newfoundland. Both of these groups can be broadly labeled culturally and biologically as “Indians” in contrast to “Eskimo” or “Inuit.” Around 3500 cal BP the Maritime Archaic people died out or abandoned the island and a new population of ancestral Paleo-Eskimo, arriving from the north, took their place. Later, about 2000 to 1500 years ago, another Indian population, presumably the ancestors of the Beothuk who occupied Newfoundland when Europeans arrived in the 16th century, replaced the Eskimo. However, these ancestral Beothuk were not the genetic descendants of the Maritime Archaic population. Similar population replacements probably occurred elsewhere in the Northeast and Mid-Atlantic regions, and these events could well be reflected by the evident transitions in the archaeological record—particularly, abrupt point style changes.
What Were Indigenous People Doing at the Marshlands?
In total, I found more than 600 whole or fragmentary points and late-stage preforms along the shoreline of the Marshlands Conservancy. Apart from the points, I also found relatively small numbers of other shaped stone tools: endscrapers, saw-toothed knives, and drills. These were far outnumbered, however, by unfinished, sometimes broken, early-stage preforms, including quartz cobbles that had been discarded after just a few flakes were removed. Those early-stage preforms showed that some people visited this location to collect workable pebbles that they began to chip into stone tools. However, pebbles of chert, slate, and other materials would not have been common here, and accordingly, preforms made of those stones are rare. So, there must have been some other reason that this site was occupied.
We have no evidence for subsistence practices; whatever nutshells, bones, etc., that might have been discarded after meals have long-ago deteriorated and/or washed away. I never found any concentrations of fire-cracked rock that would indicate former fireplaces. The many points obviously indicate hunting—but hunting of what? Terrestrial game animals, mainly deer? Or, big fish? John White painted the Indians of the North Carolina coast spearing fish in 1587.
Indians fishing in coastal North Carolina (John White 1587) https://www.britishmuseum.org/collection/object/P_1906-0509-1-6
Any informed speculation about these behaviors would require some understanding of the local topography at any given time. That is why I discuss the reconstruction of sea level in Long Island Sound at such length in the accompanying context. It seems that, before about 2,000 years ago, the shore of the Sound was actually pretty far from the Marshlands. Why then, did men discard so many points here over thousands of years, and with particular frequency around 3400 years ago? It is worth emphasizing that 600-plus points is an extraordinary quantity. I have been on full-scale excavations of sites where we felt very lucky to uncover even 10 or 15 points.
Note that I have referred deliberately to men as making and discarding points, and using them to hunt. Might not women have been involved in these activities, too? Headlines were made in November, 2020, when it was reported that a young woman had been buried in Peru 9,000 years ago, accompanied by her hunting weapons. Many archaeologists were skeptical of this interpretation. In the recently extant foraging societies that ethnographers have studied, women may participate in collective group hunts, mainly for small game such as rabbits, but big game hunting is for men only. One of my favorite quotes in the archaeological literature is from David Clarke, an English scholar who cautioned that analogies drawn only from societies known to ethnography or history unduly limited our perception of the past. Clarke (Models in Archaeology, 1972:41) advocated testing of more imaginative hypotheses against the archaeological record: “Only the latter procedure could possibly reveal to us that in Neanderthal society, say, the women did the hunting, clustered in bands of up to 300, each served by only a handful of males.” This is a nice thought experiment, but I suspect that the division of labor observed in recent foraging societies also existed in the great majority of ancient societies of similar scale and mode of subsistence.
In that case, all those points from the Marshlands represent male behavior. What, then, were the women doing? One activity that we do have some evidence for is harvesting shellfish. When I first visited the Marshlands, the only visible trace of any shell dump (midden) was a thin, sparse deposit of broken oyster shells. It usually hosted a lush growth of poison ivy, which deterred me from examining it too closely, but I did see a few small sherds of Late Woodland incised pottery in the exposed shoreward face (these were studied by Birgit Morse). I was told by Wilbur Clark, who had surface-collected and excavated in the area for decades before my arrival, that a large shell midden had once stood in the area, but had eroded away.
So, at least in the Late Woodland (after about cal AD 900) women collected shellfish. They probably also did so for several thousand years before that, but we don’t know exactly when oysters first colonized the Sound or the ancient precursor of Milton Harbor, and any older middens would surely have been located close to the ancient shorelines when they lay far to the southeast. Women probably made most or all of the pottery found at interior locations in the Conservancy; these pots are mostly of Late Woodland age, but a few sherds may be about 3,000 years old. Before clay pots were adopted, women doubtless made cooking and storage vessels out of perishable organic materials, such as leather, bark, and gourds. All of these have vanished from the archaeological record. They also presumably harvested nuts, mainly in the fall, and numerous other edible and medicinal plants. I have constantly wondered for decades, what was it that drew indigenous people to this particularly spot repeatedly, over thousands of years? Was it a favorite hunting ground for deer? Or, perhaps, was there a persistent grove of productive walnut trees, which the women set about harvesting while the men sat around repairing or replacing their stone tools?
Inferences from an Imperfect Record
In view of all the limitations and the loss of evidence that I’ve discussed, what then can we do with the imperfect archaeological record of the Marshlands?
Mostly, the points tell us about chronology. There is only one basal fragment that might have broken off a 12,500-year-old Paleoindian point. Apart from that, there is no evidence that anyone occupied this area until about 9500 years ago, and there are only two points of that age. The Orient Fishtail points that dominate the collection date from about 3400 to 3000 years ago.
Do the changing numbers of points over time tell us anything about the number of people living in the area? Did more people occupy the Marshlands at 3200 cal BP than at any period before or since? It’s hard to say. We don’t know how many points each man made or obtained each week or month, if this production rate varied seasonally, or how often the darts were lost or discarded. If bands of the same size visited the site at all periods, those that often stayed there a few weeks longer probably would have discarded more artifacts. At the end of the prehistoric period, Native men made triangular arrowheads from about cal AD 800 to 1650. The Dutch and English settlers left us no detailed accounts of tool production by the Indians. They did occasionally report how many Native people were living in a particular location, or at least how many armed men they could field for battle. However, by the time we have those numbers, successive epidemics of European-introduced diseases had swept through the Native villages, killing perhaps 90% of them. Therefore, we cannot even say how many people are represented by the Late Woodland triangles.
It is also impossible to know how representative the sample of surface-collected artifacts may be. How many artifacts lie buried under the marsh or in the underwater sediments? My impression was that the per-visit yield was already diminishing more than 20 years ago, but what has been exposed since then?
The Late Woodland people who discarded the triangular arrow points may have lived at the Marshlands, or nearby, for longer periods than earlier occupants. They grew maize, and they would have needed to watch the field to protect the crops. They could also store the maize in pits or above-ground containers. In 1871, historian Charles Baird surmised that the first English colonists in Rye were most interested in buying land on Peningo Neck and Manursing Island because the Indians had already cleared the land there for maize cultivation.
Earlier Native peoples, before the adoption of farming, were more mobile. They shifted their camps several times during each year to take advantage of seasonally available resources. How far did they move? In what direction? How extensive was the local band’s territory? Both ethnographic observations of foraging peoples around the world, and mathematical simulations, suggest that foragers must maintain social networks of about 500 or more people in order to survive. What was the geographic extent of the network that included the seasonal inhabitants of the Marshlands? With which neighboring bands did they most often exchange valued items and mates? We may be able to track these networks over time by determining the distant sources of the distinctive non-local rocks of which the artifacts of each period were made. This is why I’ve reached out to Dr. Philip LaPorta, a pre-eminent archaeological geologist, to visually and chemically analyze the lithic materials of the Marshlands artifacts. This project is now underway, with the help of a generous grant from the Friends of the Marshlands.
Stuart Fiedel, Ph.D.
Milton Harbor was mostly dry land until about 2500 years ago, when the small stream valley was engulfed by the rising waters of the Sound — ©Rory Mulligan