FOUR

Iroquoian Households

A Mohawk Longhouse at Otstungo, New York

DEAN R. SNOW

INTRODUCTION

In the seventeenth century the Mohawks were the easternmost of five Iroquois nations strung in an east-west line across what is now the state of New York. West of them resided the Oneidas, Onondagas, Cayugas, and Senecas, from east to west. They were the five nations of the League of the Iroquois, a weak confederation of former enemies who had found a way to put aside their aggression toward each other and redirect it collectively toward other native nations of the Northeast. Current evidence indicates that the League emerged sometime in the years AD 1590–1605 (Kuhn and Sempowski 2001). The legend of this process comes down to us as one of the three great components of Iroquois cosmology, the other two being the origin myth and the Handsome Lake religious revival (Snow 1994). Both oral tradition and archaeology indicate that the period leading up to the formation of the League was fraught with violence between the Mohawks and their Iroquoian relatives as well as between them and Algonquian-speaking nations in New York, New England, and elsewhere in the Northeast. We see archaeological evidence of this aggression by AD 1450. Small scattered villages began nucleating into a few large communities situated on defensible hilltop sites around that time (Snow 1995a:85–91). This pattern persisted for over a century, until the rise of the League reduced the threat of attack and towns gradually relocated back toward less protected but agriculturally more productive sites.

Figure 4.1. Northern Iroquoian village site clusters in the fifteenth and sixteenth centuries. The location of the Mohawk village site of Otstungo is denoted by a large star

In this chapter I will describe a Mohawk Iroquois household. The village site known as Otstungo (Figure 4.1) was but one of several sites that crews under my direction excavated through the course of the Mohawk Valley Project, which spanned from 1982 to 1995. This site was the most exciting and rewarding of them all. Our testing and excavation at Otstungo started in 1984 and ended in 1987 (Snow 1995a:115–142). We left the place reluctantly, pressing on with the demographic objectives of the larger research program but knowing that we would not likely work again on such an evocative place, for Otstungo was a pristine site untrammeled by modern agriculture and still laden with a primordial evanescence that captivated me, my students, and our Mohawk friends. This site revealed details of the internal operations of a typical Mohawk longhousehold (a cumbersome but convenient neologism).

The Northern Iroquoians—which included nations in Pennsylvania, Ontario, and Quebec, in addition to the five that formed the League of the Iroquois proper—were all swidden horticulturalists. They depended on a trio of key domesticates: maize, beans, and squash. Their fields were patches of forest that they cleared by girdling the trees and burning the underbrush. Without domesticated animals, the fertilizer those animals would have produced, and pesticides to control insects, Iroquoian farmers were compelled to open new fields nearly every year to make up for the declining productivity of older fields. The practice led them farther and farther afield; over the course of a decade or two their active fields were so remote from the village as to prompt relocation of the village itself. Consequently, Iroquoian villages were designed to last only a decade or two. The Mohawk Valley Project documented the relocations of Mohawk communities through more than four dozen village sites and over the course of three centuries of occupation. Because each site was occupied for only a relatively short time and because Mohawk village populations can be computed from village areas, I was able to measure demographic change from precontact times through catastrophic epidemics and colonial wars, to the final expulsion of the Mohawks from their valley during the American Revolution (Snow 1995b). That, however, is only an aside to the purposes of this chapter.

All Northern Iroquoian nations were, and to some extent still are, strongly matrilineal and matrilocal. Kinship is reckoned through the female line, and prior to the eighteenth century, residence was in multifamily longhouses governed by senior women, their sisters, and their daughters. The longhouse also was and is a symbol for the League of the Iroquois. A vast invisible longhouse stretched metaphorically across what is now New York State, with the Mohawks and the Senecas the keepers of the doors at either end. Each Iroquois village comprised a few to a few dozen of the real longhouses—spindly structures of thin posts, bent saplings, and elm bark—that inspired this image.

THE LONGHOUSE STRUCTURE

Although Northeastern archaeologists have been fascinated for decades by the longhouse and have peeled back the soil to reveal the post patterns left by many of them, attempts to build convincing replicas have often failed. I have examined and photographed over a half dozen serious attempts at replication and found them all wanting. There are several reasons for these failures, but the most important of them is that English-speaking scholars have too often depended on flawed translations of French descriptions of longhouses. We know from archaeology and even the inaccurate translations that longhouses tended to be about six meters wide and of variable lengths, depending on the sizes of the families they were designed and built to hold. Nuclear families were arrayed in pairs along the main axis of each longhouse, each pair sharing a hearth in the center aisle. Access to each house was typically only through doors at either end. Hearths were spaced about six meters apart, and they could number from as few as two to a dozen or more. Thus each hearth was in the aisle of a compartment shared by two families. Some archaeological longhouses exceed 100 meters in length. Quite often the end compartments were storage areas that also buffered the interior compartments against the cold during the winter months. Longhouses built in densely packed hilltop villages like Otstungo typically lack end storage compartments. We know from archaeological evidence that most of the house posts were about ten centimeters in diameter. These posts were laced together in a framework aboveground and sheathed with large sheets of elm bark. (For discussions of large social units in other cultures, see chapters in this volume by Ciolek-Torrello, González Fernández, Henderson, McCormack, and Varien.)

Figure 4.2. Northeastern American longhouse cross sections (from Snow 1997:77)

It might seem a small matter to project the aboveground structures of longhouses based on what archaeologists can see in the ground and infer from early written descriptions, but that is not the case. The most common error is to assume that long posts were set in the ground to form the outer walls and then bent over farther up to form a series of arched ribs to hold the elm-bark roofing. In practice the result is a Quonset-style structure having a half-circle cross section (Figure 4.2b) (Snow 1997:77). Neither surviving illustrations nor narrative descriptions indicate that Iroquoian longhouses were constructed this way, yet this form is the erroneous solution that has been adopted in many attempted reconstructions. A second common error is to assume that each longhouse had two long continuous benches for sitting and sleeping, one along each side of the structure’s interior. Compartment partition walls are typically assumed to have divided the long benches into six-meter segments. Finally, even those reconstructions that have correctly used straight outer walls surmounted by an arched roof composed of separate roof members typically make the walls too high and the roofs too flat (Figure 4.2c). While the foregoing discussion might seem overly negative, it is necessary because so much of the literature on the Iroquois as well as various reconstructions have perpetuated the errors I have described.

I have examined all of the available early descriptions in English and French, as well as all of the relevant early illustrations of longhouses (Snow 1997). When I corrected the mistakes of translators unfamiliar with Iroquois ethnography, a much more coherent vision of Iroquoian longhouses emerged. Passages that are misleading or make no sense in Standard English translations are clear in the original French. I will not repeat all that evidence here. Instead I offer only my main conclusions.

A standard longhouse compartment was about six by six meters in floor plan. An aisle two meters wide held a shared hearth and separated two living areas, one on either side. Outer bark-sheathed walls were 3.5 to 4.0 meters high and wall posts were ideally made of rigid but rot-resistant cedar. More flexible woods were used to build an arched roof structure above. The arch was nearly a half circle, six meters across and 2 to 2.5 meters high, giving the entire structure a height of about six meters. Bartram’s careful cross section dating to 1751, reproduced here as Figure 4.2a, is consistent with other early images and with narrative descriptions of longhouses (Snow 1997:77).

Benches for sitting and sleeping did not extend the entire length of the compartments’ outer walls. The benches were about two meters deep and two to three meters long, no more than half the length of the compartment, leaving space for storage and other activities on the floor of the living area. The benches were raised about a half meter above the floor, high enough to raise people above the damp earth but low enough to allow small children to climb up unaided. Each bench had its own roof about two meters off the ground, on which were stored pots and other household objects. There were also side walls to the little structure such that the whole unit was a kind of snug berth or cubicle in which the nuclear family relaxed and slept. Food preparation, tailoring, and other daily activities were carried out in the aisle or inside areas to one side or the other of the berths. The documents alone make it clear that simpler interpretations of interior structural features of longhouses are erroneous.

THE LONGHOUSE SOCIAL UNIT

Typically, before more improvised and expedient rearrangements brought on by catastrophic seventeenth-century epidemics came about, the women who shared fires within compartments were close female relatives. Indeed all of the women in a longhouse were typically members of the same clan segment. In the case of the Mohawks there were three such clans: the Wolf, Bear, and Turtle clans. Thus each Mohawk longhouse, regardless of the number of longhouses in the village, was identified as being occupied by one of these clans. Harmen van den Bogaert visited Mohawk and Oneida villages in the winter of AD 1634–1635 and noted that houses were identified by wooden totem signs above their doors (Gehring and Starna 1988:13).

A senior woman presided over the clan segment (matrilineage) that occupied each longhouse along with in-marrying male spouses. Nuclear families tended to average five people (Snow and Starna 1989). Thus each compartment, typically containing two cooperating nuclear families, had an average population of ten people. One can reasonably compute the population of a longhouse by multiplying the number of hearths (fires) by ten. Early documentary sources such as the Jesuit Relations and Allied Documents often report the number of fires in houses or even entire villages, which makes computation of local population sizes in such cases much easier than it might otherwise be.

Longhouses built in the fifteenth century often had three to five fires, implying populations of thirty to fifty people. Women tended to bear and raise three or four children each at that time. Assuming for the moment that every woman had exactly two daughters, that they invariably married and bore children, and that generation time was twenty years, one can envisage the development of a “longhousehold” made up of forty people and containing a senior woman about seventy years old, her two daughters, her four granddaughters, sixteen greatgrandchildren (both boys and girls), and various in-marrying men. Presumably such a household would split into two smaller ones with the death of the matriarch, a repackaging that probably would have occurred most often at the next village relocation. The two daughters would then preside over the construction of two new longhouses of their own, and the birth of a new generation of babies would push the total population of each new longhouse toward forty or fifty. This ideal scenario was probably not played out often in the real world of Iroquois life, of course. Some women were childless while others had only sons. Marriages sometimes failed, sons sometimes died far from home, and so forth. Some households flourished and grew while others shrank, their occupants forced to merge with other less fortunate households and to invent fictive kin ties that would allow them to operate as single households. But overall the archaeology of Iroquois sites suggests that things tended to balance out. There are many examples of villages in which most of the longhouses had three to five compartments warmed by an equivalent number of fires.

Something happened in the late fifteenth and sixteenth centuries to change the dynamic equilibrium just described. Small villages came together to form towns that could be more easily protected, and at the same time newer long-houses were built to sizes much larger than earlier ones. An Onondaga house at the Schoff site was built to a record length of a 122 meters (Tuck 1971:95). Apart from that unique example, longhouses with twelve or more compartments came to be much more common (Snow 1994:74; see also chapters by Ciolek-Torrello and Varien, this volume, for discussions of architectural change in the American Southwest). A longhouse having twelve fires would have housed about 120 people, based on the above stated average of ten people per fire. Because each of twelve compartments would have had two in-marrying men from other clans (one for each nuclear family), ninety-six of the household’s residents were related women and their children belonging to the same clan. Given the nature of lineage growth I described above, it is unlikely that such a large household could have been presided over by a single individual, usually a senior woman, the common ancestor of over ninety women and children. At the rate that even the most fortunate Iroquois lineage grew in that era, such a size could not be reached within the lifetime of one woman. In other words, any twelve-compartment longhouse must have been occupied by women and children whose common ancestor was no longer living, which implies that the household matriarch of the late sixteenth century presided over not just her daughters, granddaughters, and great-grandchildren but over at least a few sisters and perhaps some cousins as well. We do not know how such a matriarch was chosen from what was probably a handful of eligible senior women.

We do not know the details of how this social innovation worked, but its existence seems hard to deny (Snow 1994:74). The most economical hypothesis available to explain the change is that it was a consequence of the prolonged occupation of more densely packed Iroquois villages and towns during this period of heightened warfare. The Otstungo site, on which this chapter focuses, might have been occupied for as many as seventy-five years. Prolonged village occupations and less frequent relocations would have meant that there were fewer opportunities for longhouseholds to split, even after senior matriarchs died and supervision passed to their daughters.

Support for the prolonged occupation hypothesis comes from linguistic research. Prior to the sixteenth century all Iroquois children called their aunts by the same term they used for their biological mothers, regardless of which of the five Iroquois languages they spoke. Sometime before 1635 they all began to distinguish once again between their biological mothers and other women in their household belonging to their biological mother’s generation (Mithun 1988). We know this fact because Mohawk terms recorded by Van den Bogaert in that year indicate that the shift had already occurred in that language (Gehring and Starna 1988:54). The specific linguistic solution was different for each of the Iroquois languages, but the general principle was the same across Iroquoia. The best current explanation is that the growth of residential matrilineages had simply generated too many “mothers” for a single term to serve (Snow 1994:74).

The shift presaged the effects of demographic collapse occasioned by the first smallpox epidemics in 1635. Mohawk village populations fell by 63 percent in less than a year. Four Mohawk villages were abandoned and four new replacement villages built over the course of a few months. I am compelled to infer that for longhouses of more-or-less standard size to have been constructed and occupied in the new villages, the social fragments forming the surviving 37 percent of the earlier population must have been consolidated and repackaged in a number of improvised ways (Snow 1995a:300–304).

Archaeological evidence from the Mohawk Valley Project indicates that from AD 1635 on, longhouses became standardized at three to four fires each and they no longer expanded and contracted to accommodate variably sized and changing matrilineages. This was also a period in which all Iroquoian nations attempted to offset the population losses caused by repeated waves of smallpox and other epidemic diseases by taking in prisoners and refugees as adoptees. The symbolically potent longhouse still stood, but it became the durable container for increasingly fictive matrilineages, not the flexible container for durable, evolving matrilineages it once had been.

This new role for the longhousehold persisted for nearly a century, until the economic and military realities of the eighteenth century caused the Mohawks to move to dispersed settlements of one- and two-family houses built along European lines. The best example of this new format from the Mohawk Valley is the site known as Caughnawaga, which was occupied from AD 1679 to 1693. Here the Mohawks constructed twelve longhouses, each having three or four fires. The orderly layout of the longhouses in two rows within a square palisade resembles a modern mobile-home court than it does the much more random settlement patterns of sixteenth-century sites (Snow 1995a:434).

The above brief description of the longhouse as social unit is not new, with the exception of the hypothesis regarding the emergence of much larger households and the probable collective leadership of them by sets of sisters. The previously published information in this section (Snow 1994, 1997; Snow and Starna 1989) provides the context that sets the stage for the interpretation of the Otstungo longhouse, a site that has revealed details of the internal operations of a typical Mohawk longhousehold. It is to that discussion that I turn now.

OTSTUNGO LONGHOUSE 1

The Otstungo site is located on a cliff overlooking Otstungo Creek, a tributary of Otsquago Creek, which is itself a tributary of the Mohawk River. The site is about six kilometers southwest of the Mohawk River in the town of Minden, Montgomery County, New York. It was founded as early as AD 1450 and might have been occupied continuously until AD 1525. The period of its occupation thus almost certainly included AD 1492, the year of the first voyage of Columbus. This circumstance led to the inclusion of Otstungo in a National Geographic magazine article on four North American Indian settlements that were occupied on the eve of Columbus’s voyage (Bruchac 1991).

The setting of Otstungo is dramatic. The peninsula on which Otstungo is sited is elongated, about 180 meters long and 50 meters wide, oriented northwest-southeast. The southwestern and northwestern approaches are protected by Otstungo Creek and a twenty-four-meter cliff. The northeastern side is protected by a less steep but still protective slope. Only the narrow southeastern side is easily approached, and this was protected by an artificial ditch and palisade (Figure 4.3). The living area of the village measures 7,572 square meters. Because of the location of the creek and its steep banks, it seems likely that the active fields tended to be located to the south and east of the village.

Figure 4.3. Contour map of the Otstungo site, New York, showing thirty-meter grid, hearths, house grid, and defensive ditch (contour interval = 1 m)

The site was first described by Ephraim Squier, who visited it in 1848 (Squier 1849:59–60). Collectors and amateur archaeologists visited it many times in the years following. Research teams under my direction first tested the site in 1984. Our preliminary work suggested that although the site had been logged repeatedly, there was no evidence that it had ever been plowed. This presented a unique opportunity for us to investigate one or more relatively undisturbed longhouse floors inasmuch as no other Mohawk sites offered such intact preservation.

The topography of the site suggested that most longhouses would have been oriented roughly northwest-southeast. Beginning in 1985 we conducted detailed magnetometer surveys to locate lines of hearths by nonintrusive means. We probed subtle magnetic anomalies in search of confirming fire-reddened soil and eventually mapped nine lines of hearths that we took to be evidence of at least that many longhouses. The dots on Figure 4.3 show the distributions of hearths. One line was selected for detailed excavation and the centerline of the excavation grid was laid down along the axis of the hearths as shown on Figure 4.3. The excavation grid was oriented northwest-southeast, roughly 45 degrees off cardinal directions. This decision gave us excavation units that mapped closely to the original longhouse floor plan and enabled us to acquire unprecedented detail about life inside a single household.

Excavation continued through all or parts of three field seasons, from 1985 to 1987. We began with a traditional three-meter grid, chosen because it approximated the ten-foot grids used previously in Mohawk archaeology. These were divided into 1.5-meter quads for purposes of record keeping. We soon discovered that the quality of the provenience information we were getting justified even smaller excavation units, so the quads were further subdivided into 75 centimeter subquad units. Personal computers and spreadsheets were not available at this time so the cumbersome manual system of squares, quads, and subquads was used through the entire three seasons. I later translated all the records into a single, much simpler computer file in which each of 864 subquads is a cell in a simplified grid. Figure 4.4 shows the alphanumeric coordinate system on the x- and y-axes. I have included confirmed hearths to allow the reader to see the connections between this and other figures. The excavation actually extended beyond the northwest end of Longhouse 1, across an open area and into the southeastern end of Longhouse 2, but my discussion in this chapter is restricted to the contents of the six compartments of Longhouse 1.

There were some serious impediments to our excavation. Figure 4.5 shows the roots of trees that hampered standard cell excavation. Shallow shale bedrock was often encountered and the clay soil overlying it was difficult. Artifact recovery varied even among adjacent excavation cells because of almost daily differences in light, temperature, and humidity. The expertise of my student excavators varied by ability and length of experience. Some of the variations in artifact concentrations discussed below must be attributed to these variables in our ability to recover evidence. Figure 4.5 also shows the distribution of artifactual debris recovered from the floor of Longhouse 1, each dot representing ten objects. Despite variations in recovery success, clusters of objects of different kinds, all of which are aggregated in Figure 4.5, are clear enough to draw inferences about activity areas.

Figure 4.4. Excavation grid showing excavated cells (gray), the cell coordinates, and confirmed hearths of the Otstungo site, New York

Post molds were easily confused with root stains, decayed shale, and animal burrows in the shallow Otstungo soil. Every suspected post mold was cross-sectioned and deliberately subjected to as much general argument among my assistants and me as was needed for confirmation by consensus. Many features that probably were indeed post molds did not survive this rigorous process. The result is that the pattern of 110 post molds is less complete than it might otherwise have been (Figure 4.6). However, it is also less confused by the large number of fraudulent post molds that muddle so many published long-house post patterns. Gray ash was found coating the floor across some parts of Longhouse 1; this too was mapped. Apart from post molds and ash thirty-seven features in the floor of Longhouse 1 were uncovered (Figure 4.5 and Table 4.1).

DIVISION OF LABOR

We know much about Iroquoian division of labor from written sources. Sagard (1968), for example, provides considerable detail. In some places he talks about the daily activities of the Hurons, implying that he is talking about men much of the time and mentioning women and girls only when he refers to their specific activities. Fortunately he is repetitive and more explicit in other places, indicating that in the village some activities are primarily or exclusively men’s while others are primarily or exclusively women’s. For example, he explains that women and girls make cordage and that men use it to make nets and snares (Sagard 1968:101). Things are different outside the village, for in all-male groups men must cook for themselves, an exclusively female activity in the village. Within a complete and functional village household, however, the standard division of labor is clear.

Figure 4.5. Longhouse 1 plan showing the probable outline of the house, tree roots, excavation limits, confirmed hearths, the probable locations of six residential compartments, and artifactual debris of the Otstungo site, New York (each dot represents ten objects)

Figure 4.6. Longhouse 1 plan showing ash (light gray), hearths, compartment outlines (A–F), and features (numbers key to Table 4.1), Otstungo site, New York

Table 4.1. Features from the Otstungo Longhouse 1, New York

Table 4.2 lists those activities that Sagard clearly assigns to both women and girls on the one hand and men and boys on the other (see discussions of gender divisions in household labor in other cultures in chapters in this volume by Gonlin, Gougeon, Henderson, and Wiewall). Missing from the two lists are those activities explicitly attributed to both (snowshoeing, gambling, feasting, and dancing) and those activities that are of uncertain assignment (painting, house construction). Many other contemporary sources support Sagard’s generalizations and their applicability to the Mohawks in particular. Some of these references are of special importance to the conclusions I draw below. Megapolensis tells us in 1644 that men smoke and that they make their own pipes (Snow, Gehring, and Starna 1996:45). In his long 1654 narrative, Pierre Esprit Radisson mentions tobacco smoking several times but attributes the activity to only men (Snow, Gehring, and Starna 1996:69). Adriaen Cornelissen van der Donck similarly describes smoking as a male activity (Snow, Gehring, and Starna 1996:120). Paolo Andreani, who passed through Mohawk country in 1790, explicitly says that men “stay at home lying about, smoking pipes and playing games while the wives ruin themselves at hard labor in the field” (Snow, Gehring, and Starna 1996:322).

Table 4.2. Iroquoian division of labor according to Sagard (1968)

Apart from Sagard, sources are generally silent about who does the chert knapping. Bows and arrows are mentioned, but most early narratives about the Mohawks postdate the introduction of firearms, so the manufacture of knapped points and knives does not get mentioned. Women were responsible for hide preparation and food preparation, so I infer that they would have engaged in at least some scraper and knife resharpening. However, Sagard is quite clear in attributing the making of chert-tipped arrows and chert knives to men (Sagard 1968:98, 323). Careful reading of his original French confirms that this assignment is not a consequence of mistranslation.

DEBITAGE AND PIPES

Given the clarity of Sagard’s narrative, it is reasonable to hypothesize that Iroquois men were primarily responsible for the manufacture of chipped-stone tools. This implies that concentrations of chert debitage should identify male work areas within the longhouse. Figure 4.7 shows the distribution of debitage, each small dot representing a single flake.

Figure 4.7. Artifact distributions at Otstungo site, New York; black diamonds show the locations of smoking-pipe fragments

We know from various sources cited above that men made pipes for smoking tobacco. If men were the primary makers of pipes and primarily responsible for knapping, and if they carried out both activities in specific areas of the longhouse, then pipe fragments should cluster with debitage in male work areas. The distribution of pipe fragments within Longhouse 1 is shown by black diamonds on Figure 4.7. Pipe fragments tended to be located in or around hearths or in or very near dense clusters of debitage. These distributions support the hypotheses that pipes and chipped-stone tools were both made and used primarily by men and that certain areas within longhouse compartments were male work areas.

One cluster of pipe fragments was found in a small storage pit or cubbyhole (Feature 1) in Compartment B. When mended, these turned out to be pieces of a single striking pipe depicting a human effigy mask (Figure 4.8). The pipe was featured in the 1991 National Geographic article on Otstungo (Bruchac 1991:69) and is on the book jacket of the current volume. I infer Feature 1 and others like it to be small personal storage cache pits. Although this one appears to have been used by a man, we cannot be certain that all eight of those identified were used by only men.

ASH

The distribution of ash on the floor of Long-house 1 can be seen on Figures 4.6, 4.7, and 4.8. These carpets of ash were concentrated in the end compartments and to a lesser extent in the adjacent ones. I infer that the ash came largely from the cleaning of hearths and that spillage was cumulatively greater as people carrying away the ash moved toward the end doorways.

Figure 4.8. Human effigy mask smoking pipe from Otstungo site, New York (drawing by Gene Mackay; Snow 1995a:130)

Smaller patches of ash near the sides or corners of Compartments A, B, D, E, and F might mark places where cooks added ash to soaking corn, a common technique for breaking down the kernels. Thus, I tentatively infer these patches to mark food-preparation areas. All contemporary documentary sources make it clear that women were primarily responsible for food preparation in the village, so it is reasonable to conclude that these were largely female activity areas.

Figure 4.9. Artifact distributions at Otstungo site, New York; rim sherds are shown as small dots with each dot representing a single sherd

POTTERY

Sagard is quite clear in telling us that women made and used pots (Sagard 1968:102, 109). The distribution of sherds in Longhouse 1 is informative even though this artifact class was more subject than others to variations in recovery during excavation, as mentioned above. Three tendencies can be noted from the distribution shown in Figure 4.9. First, the sherds of broken pots were often gathered and dumped against the wall outside the longhouse. Such concentrations can be seen outside Compartments A, D, E, and F. Second, sherds tend to concentrate around hearths, where breakage was presumably high and some fragments were quickly trampled into the earthen floor. Third, larger sherds tended to concentrate inside the longhouse but away from the center aisle in places where we can infer sleeping berths were located. Sherds that were kicked under the berths often escaped trampling and cleanup.

The irregular black shape in the lower right corner of Compartment F is a large milling stone. The distribution of ash near it also suggests that this was a cooking area. That inference is further supported by the concentration of sherds in the same area.

Figure 4.10. Artifact distributions at Otstungo site, New York. Food-bone fragments are shown as small dots; each dot represents a single fragment

Unfortunately, if one looks just at the distribution of sherds as shown in Figure 4.9, one cannot easily distinguish between berth areas where sherds were easily kicked and forgotten and food preparation areas where sherds were produced by breakage in large numbers but then were incompletely cleared by subsequent cleaning. However, if average sherd size is considered, the distinction becomes clear. Sherds found in high traffic areas were small and those lost under berths were larger.

BONE

The distribution of bone fragments reveals a different and interesting pattern (Figure 4.10). Large concentrations were found in and around hearths. The locations are where one should expect bone fragments because raw food was added to the pots here and people eating from them tended to discard bone fragments into the fire. Other concentrations were found in some, but not all, areas already inferred to have been food-preparation areas. A few concentrations, one very dense, were found outside the house, further indication that refuse was dumped there. A concentration of bone refuse was found inside the house along the left wall of Compartment B. The ground outside the house sloped up and away from the house in this area and the builders had cut into the hillside to flatten the floor of the house. It is possible that the bone refuse in this part of Compartment B resulted from fragments washing downhill and under the house wall into the interior.

Figure 4.11. Artifact distributions at Otstungo site, New York; distribution of scrapers (black squares) and projectile points (black triangles)

CHIPPED STONE TOOLS

We found 96 cores, 34 scrapers, 10 knives, 168 projectile points, and 30 other bifaces in Longhouse 1. It is common in North American archaeology to hypothesize that women used most of the scrapers and that men used most of the projectile points. This hypothesis receives strong support from Sagard’s documentation of Iroquoian division of labor (Table 4.2). As can be seen on Figure 4.11, the points do not cluster in any obvious way, suggesting that their distribution is not as informative as chert debitage in indicating male work areas. It is likely that post-manufacture use, breakage, and discard of points randomized their deposition.

The most interesting observation regarding the distribution of scrapers is that many of the thirty-four recovered were found outside the longhouse beyond Compartment F and therefore outside the area covered by Figure 4.11. This suggests that much of the work done with scrapers, hide preparation, and the like was done outside the house. As is the case for projectile points, the distribution of scrapers does not tell us much about who might have been using them.

The distribution of ground-stone tools did not supplement interpretation of other patterns because there were too few stone tools to show any clear pattern.

SHELL AND TEETH

We recovered 1,634 shell fragments from the floor and immediate exterior of Longhouse 1. These were mostly mussel-shell fragments, which I infer to have been mostly food debris. We found no marine quahog (Venus mercenaria) or whelk (Buccinum undatum) shell fragments, which were popular a century later. These sticks were used to make wampum beads in the seventeenth century, but there is no evidence that they were already being so used at Otstungo. The mussel-shell fragments were distributed mainly around hearths and in areas already identified by ash and food bone fragments as probable food-preparation areas.

We recovered ninety animal teeth, some of which were food debris while others probably were used as pendants. Still others, especially beaver incisors, were used as cutting tools. Their distributions did not provide any insight into household activities.

CRYSTALS

We found twenty quartz crystals, the so-called “Herkimer Diamonds,” for which the Mohawk Valley has been known since pre-Columbian times. The Mohawks valued the crystals and many found their way in trade to other nations in the region. The Mohawks called themselves the “Kanyenkehaka,” which is often translated “People of the Place of the Flint.” But local flint (more properly “chert”) was not of particularly good quality, and it is more likely that the term referred to quartz crystals (Snow 1994:86). The crystals were highly charged symbolically, and their distribution in Longhouse 1 is therefore of special interest. All but one of the crystals was found inside or just outside the exterior door of Compartment F (Figure 4.12). The lone exception was found outside longhouse Compartment D.

The distribution of crystals in Compartment F corresponds to a concentration of pipe fragments in the same area (Figures 4.7 and 4.12). This suggests that the crosshatched area shown on Figure 4.12 approximates the location of the berth of a senior male. Pipe fragments and crystals would have fallen easily through the platform of sticks that made up the bottom of the berth. Another concentration of pipe fragments without associated crystals across the aisle suggests that the berth of a less senior couple was located there. A personal cache pit (Feature 12) was found in this second probable berth area. Ash, food bone, potsherds, the milling stone, and a storage pit (Feature 3) all suggest that both of the open areas below the two berths (on this representation) were food-preparation areas. Curiously, one crystal was found near the milling stone. While one might hypothesize that senior families would live near the center of a longhouse like this one, perhaps in Compartment C or D, the evidence of the crystals does not support such a hypothesis.

SUMMARY AND CONCLUSIONS

One can use the distributions of pipe fragments, debitage, sherds, bone, and ash to map the probable locations of berths in the other compartments of Longhouse 1. The example of Compartment F, however, is sufficient to demonstrate archaeological support for the inferences I have drawn from documentary sources regarding the internal structuring of longhouseholds. There were identifiable work areas on one or both sides of sleeping berths, and each berth took up only about half the area allocated to each family.

Figure 4.12. Distribution of quartz crystals (black stars) and other features in Compartment F at Otstungo site, New York; crosshatched rectangles indicate probable berth locations

Longhouse 1 was probably occupied for a long time compared to other Iroquois longhouses, yet spatial patterns persisted in the distributions of key artifact types. Support for the length of occupation can be found in the details of the hearths. Hearths were begun as surface fires, but repeated cleanings gradually turned them into basins twenty centimeters deep or more. I have referred to the deepest of them as “roasting pits” in Table 4.1. There were two such pits in Compartment B (Figure 4.6, Features 2 and 16), one intruding into the other. Feature 2 was clearly old when Feature 16 was begun, and Feature 16 also had a long life. There was also a deep hearth (Feature 5) and a secondary shallow one (Feature 19) in that compartment. All of this evidence points to a long occupation.

We recovered 27,162 sherds from Longhouse 1 and other excavations at Otstungo. Of those, 2,071 were rim sherds, resulting in a ratio of about 13 to 1. I used 864 rim sherds from the longhouse floor to plot the distribution pattern shown on Figure 4.9. I used rim sherds from Compartments A and B to determine how many vessels were present, mending rim segments whenever possible. Only seventy-four could be joined to others, and these represented no fewer than twelve vessels. This suggests that the 864 rim sherds represent at least 140 vessels, almost a dozen per family. I can only guess at production and breakage rates, but a minimum of a dozen vessels broken, discarded, and missed by cleaning per family implies a long occupation. These numbers are much higher than those for longhouses on other Mohawk sites that have closely dated occupations of a decade or two. I have estimated the occupation range of Otstungo to be AD 1450–1525 based on radiocarbon dates (Snow 1995a:133–138). It is reasonable to infer that Longhouse 1 was occupied for much or all of that period.

Male and female activity areas can be identified in Otstungo Longhouse 1. We can also identify the compartment in which quartz crystals were stored and from which they were presumably distributed to destinations outside the long-house. I infer that this was the compartment occupied by the senior nuclear family in the household, probably a woman and a man of considerable standing in the community as a whole and perhaps beyond.

The Otstungo site has provided us with a unique view of the internal structure of a Mohawk longhousehold. No other Mohawk site and few other Northern Iroquoian sites offer such an undisturbed record of the organization of daily life in a multifamily dwelling. Perhaps sixty Mohawks lived in Longhouse 1 by the end of its occupation around AD 1525. Many more Mohawks than that walked through the longhouse as guests of the project during the years it was being excavated. The longhouse was almost certainly occupied in AD 1492, when Columbus first touched America far to the south. At that time there was an immense white pine at the southern entrance to the site, undoubtedly an important symbol to the Mohawk villagers, whose descendants still revere the tree. The pine, which was gone by a century ago, measured 2.2 meters in diameter, larger than any surviving white pine in New York. Its multiple trunks were cut into thirty-six logs, not one of them shorter than three meters.

While loss of the white pine is a sad point in the recent history of the site, Otstungo seems safe for the future. The site remains under the stewardship of John Schuyler and his family, who have owned it and the surrounding land for many years. Several other longhouses remain virtually undisturbed at Otstungo and are available for further research as new techniques emerge. I go back there whenever I can. My son and I had the honor of hearing the Iroquois Thanksgiving Address recited there by Chief Jake Swamp one perfect summer evening, and for me it remains one of the most evocative places on the face of our continent.

REFERENCES CITED