Network Morphology and Colonial (In)Visibilities

Access copy — please do not cite or circulate

Slides

Women are difficult to find archivally, indigenous women even more so. In many of the records I work with documenting Haudenosaunee Iroquois women, women’s presence was often noted even as they were erased as individuals. European observers recorded the presence of specific male Haudenosaunee speakers, and that they were accompanied by a mass of “women, children, and horses besides.” Today I’m going to briefly walk you through the morphology of three reconstructed indigenous social networks from the eighteenth century to think about what network analysis can say about how much and what Europeans saw and recorded of the indigenous communities they interacted with.

[Slide 1, with the document]

Indigenous communities were not completely visible to European observers. Many of the women I study appear only once or twice in the archive, in documents like this baptismal register where they’re briefly named and then never mentioned again. My paper today is part of the very, very early stages of a project thinking about how—somewhat paradoxically—network analysis can make visible the active process of erasure in the settler colonial archive that has written indigenous and other women out of the historical record.

[Slide 2: Map]

Some very brief background: the networks I’m going to show you today represent connections in what’s now New York during the seventeenth and eighteenth centuries, primarily in Haudenosaunee communities. The Iroquois Confederacy includes the Mohawk, Oneida, Cayuga, Onondaga, Seneca, and Tuscarora nations, which continue to reside in New York, Canada, Wisconsin, and Oklahoma. Up through the American Revolution, they maintained control of their ancestral lands with very little incursion from white squatters, which had a significant role in shaping the recording of the networks I’m going to show you today. Haudenosaunee communities mainly practice matrilineal descent through clans, meaning that a man’s social standing and social connections were largely dependent on the clan and social connections of his mother and wife.

All of this is, of course, only visible archivally through the lens of European observers, and thus our view of indigenous networks in this period is a record of what was visible to the person who documented them. That seems kind of ridiculously obvious, but it’s worth emphasizing. Europeans often recorded indigenous women without names of their own, and identified them as the wife, mother, or daughter of a named man, distorting our view of indigenous social connections. I argue, and the problem I’m trying to tackle in this new project, is that I think network analysis doesn’t tell us about women’s lives—it tells us about how women’s lives were visible to those around them. This is an especially important distinction to make in the study of indigenous and other marginalized or subaltern women, because the two are often conflated.

[Slide 3, Network 1: Munsee-Ulster]

This is a very sparse network. In this and in all the other networks I’m going to show you today, men are blue and women are red, and they’re sized by degree.

This network represents connections between Munsee people who traded at Ulster NY in the late seventeenth century—people who traveled together, bought things for each other, or paid each others’ debts. The Munsee are a non-Iroquois Algonquian speaking group, and I show this to you as the first network as a point of comparison for the later Haudenosaunee networks. This is a network of economic social ties, and it mirrors much of what the scholarship says about the impact of colonialism in indigenous communities. The largest figures in this network are well-connected male sachems who had significant diplomatic connections to one another and to white colonial leaders through land sales and treaty conferences. Women in this network mostly appear as marginal figures identified as sachems’ wifes, like the Sawanossie’s wife here. This is in line with the scholarship that suggests that as the Munsee and other eastern indigenous groups lost land and sovereignty in the seventeenth century, women’s leadership roles in their communities declined. The scholarship also suggests that internal ties for many of these groups became more attenuated over time, with internal social connections broken by the impact of debt and death from epidemic disease and male leaders assuming more centralized control in place of earlier consensus governance. We see suggestions of this here with these long chains of people who have only one or two connections, and who are connected to the main network through these prominent male sachems.

However, there’s a rather notable exception in this network with the prominence of a woman named Kisay. Kisay has her own little subnetwork here which is relatively unconnected to the more well documented connections between known male sachems and includes relatively more women than the rest of the network. Her subnetwork is also relatively more dense than the rest of the network, suggesting that the Dutch trader who recorded her connections saw more of her connections than he did of the rest of the people in this network. Kisay, also notably, appears in no other archival records for this period, at least so far as I’ve been able to find.

So this tells us three things:

1. Network morphology does seem to conform to what we’d expect from traditional historical methods, with a relatively sparse network for a group experiencing significant colonial pressure.

2. Indigenous women like Kisay who aren’t visible in the archive of intercultural diplomacy are both visible and prominent in the archive of mundane, daily connections like buying and selling.

3. Some parts of indigenous networks were more observable to Europeans either because those communities were more integrated into colonial supervision or because those connections were unavoidable.

[Slide 4, Network 2: Iroquois-Albany]

This is another economic network, built with the same kinds of connections of people buying gifts, traveling together, and paying debts for each other. This and the previous network are roughly contemporaneous, and are geographically close, about eighty miles apart. Unlike the previous network, this one is only Haudenosaunee people.

The most striking thing about this network to me is how many more women there are. Like the Munsee network, this one is relatively sparse, but there is no Kisay in this network. This network is built from an account book kept by trader Evert Wendell, who preferred to deal with Haudenosaunee women under their husbands’ accounts, but who ended up having to make separate accounts for women who insisted on conducting their own exchanges separately from their husbands. Fully half of Wendell’s accounts were with women. I think there’s no Kisay in this network because the power relations are completely different. Kisay was exceptional in her network because, for some reason, she was economically unavoidable for the Ulster trader. In this network, Haudenosaunee women in general were economically unavoidable for Wendell, making them a much more pervasive part of this network.

I also think the sparsity of this network is for a different reason than the Munsee network: the Munsee near Ulster were relatively fixed, many of them indentured to white landowners or dependent on cash wages from working as daylaborers. They didn’t have many economic or political options. Wendell’s Haudenosaunee customers, on the other hand, often just left if they were unhappy. There were no white settlements in the Iroquois territories of what’s now New York state until the 1790s, meaning that Wendell and other traders dealing with Haudenosaunee customers really had no options for collecting debts or observing social connections. Where the Ulster trader probably saw his Munsee customers on a weekly or monthly basis because they lived nearby, Wendell’s Haudenosaunee customers lived anywhere from fifty to three hundred miles from him, and usually only came to trade once per year.

So with two networks built from similar documents, in similar locations, in the same time period, we have some major factors shaping differences in network morphology:

1. Geographic distance between indigenous and settler communities has an impact on how well indigenous networks were observed.

2. Colonial power dynamics influenced the visibility of women’s participation and influence.

[Slide 5, Network 3: Fort Hunter]

This next network tests those two ideas. This is a representation of connections between individuals who stood as godparents to children in the Anglican church at Fort Hunter, NY, about forty miles west of Albany. The Native people in this network are all Haudenosaunee, and unlike the previous networks we looked at, this network includes European settlers as well as indigenous people. This network represents a community of people who lived within about twenty miles of each other, who attended church regularly, and yet whose subcommunities within this network overlapped very little and have different morphology. I don’t think that the changes we see here are due to change over time: I have a couple of economic networks like the ones we saw previously which cover this and later time periods, and they look very similar to the first two networks we looked at. I’m showing this network instead of more economic networks because it illustrates the difference that document genre makes in network morphology, as well as the importance of colonialist surveillance in shaping network morphology.

The settler portion of the network is much denser than the Haudenosaunee side of the network. The prominence of women in the two sides of the network is also striking: Haudenosaunee women are much less prominent here than either settler women or Haudenosaunee women in the previous network. That indicates that the Anglican minister who kept the baptismal network, either 1. preferentially recorded settler women’s ties or 2. simply saw less of the Haudenosaunee side of the congregation because women didn’t participate as much as men, although other archival sources suggest that Native women converted more often than men. European networks in this period tend to be very dense, like this little knot of Scottish Presbyterians here, with outliers of marginalized groups like these recently immigrated Palatine Germans up here. The fact that Barclay recorded such density in the settler portion of his congregation and not in the Haudenosaunee portion suggests that his view of the Haudenosaunee community at Fort Hunter was limited in some way.

[Slide 6, Network 5: War of 1812 Pensions]

Before I finish, I really briefly want to show you a network from a later period to talk a bit more about document genre and archival colonialism. After the War of 1812, American and Native veterans filed pension claims from about 1815 to 1855, and named the executor of their estate—especially important because the US government didn’t get around to paying these pensions until almost the Civil War, and many of the veterans themselves were dead by that time. This network is an archival representation of the social structure the US government was trying very hard to enforce on all reservations: one central chief with his trusted captains, their warriors, and the warriors’ dependents. This is what most of the 19th century indigenous networks built from government sources look like, especially after 1848 when the federal government started forcing reservation governments to adopt models of government styled after the Constitution which barred women’s participation. This is what colonialism looks like. And yet, Haudenosaunee women here act as structural connectors—their connections as executors of estates tied together this network despite active colonial pressure against it.

[Slide 7, questions???]

This project is in its very, very early stages, so I’m going to finish with some questions rather than answers. What I’m hoping to do is to build a website hosting several dozen networks like these in order to compare the effect of document genre, ethnic and national differences, change over time, and other factors on the visibility of women in early American networks. I’m tentatively planning to include an interface where others can upload their own networks to compare in order to extend the project and for use as a teaching tool. Some of the big questions I’m interested in with this project:

  • Is network density a product of actual connections or of recorded connections?
  • Is recording of connections a product of colonialist surveillance or the strength of those connections in colonialist interactions?
  • What can network morphology say about ethnic community creation and boundaries?
  • And finally, how does visualizing the limits of the archive through networks help us better understand the agnotology of women’s archival presence? I argue that using network analysis, the colonial process of archival erasure can be made visible, giving us a better sense of how and why indigenous and other women influenced their communities.