Settlement growth and military conflict in early colonial New England 1620–1700

Abstract

Property rights are secure, and violence over land can be attenuated when the treatment and delineation of the property are consistent, stable, and interpreted similarly by each party. In the absence of a mutual understanding of property rights, land-use stability becomes strained as the area of contested land between two rival parties expands—when one party (or group) is perceived as asymmetrically and rapidly accumulating land at another’s expense. While relations between Algonquian tribes and English settlers were generally peaceful in the first half of the 17th century, subsequent colonial growth accelerated and lead to violent conflict. The latter half of the 17th century experienced some of the most devastating conflicts during early colonial American development—beginning with Pequot’s War, peaking during King Philip’s War, and ending with a European proxy war in North America during King William’s War. Using probate data for 72 settlements in New England to measure the growth of farmers as a proxy for colonial territorial growth, I find a general pattern that English settlements with higher rates of population and territorial growth experienced more violent conflict during King Philip’s War. The same relationship between territorial growth and violent conflict was not as strong for wars that preceded and succeeded King Philip’s War.

1 Introduction

Social norms and formal rules aid in the definition of property rights and the resolution of conflicts over land. Historically, these institutions have either been relegated to formal states (Acemoglu & Robinson, 2019; Johnson & Koyama, April 2017; Dincecco & Katz, 2016) or deferred to stateless channels of resolution (Candela & Vincent, 2020). The violence-reducing characteristics of these institutions include: (1) defining property rights over space and time, (2) adjudicating disputes over property, and (3) enforcing those resolutions after adjudication. Areas with competing concepts of property rights institutions, such as the early colonial period in New England, lead to many points of contention and confusion across all three characteristics between English settlers and Algonquians.

During the period of rapid colonial expansion in the mid-17th century, the mutual understanding of land use between English settlers and Algonquian tribes fluctuated. Colonial population growth and expansion after the arrival of the Mayflower in 1620 made the process of defining and redefining property rights more costly through peaceful means. As settlements grew, English settlers sought to acquire more land for development, access to pastures, and rights to fishing, hunting, and timber. The marked terrain and land-use arrangements made with Algonquian tribes changed frequently and often became contested.¹ The sharp change in relative growth between English settlers and Algonquians expanded the area of contested land. The area of contested land grew in proportion to the expansion of English settlements.

The types of institutions that addressed and resolved potential land disputes impacted the likelihood and severity of violence in colonial North America. In colonial French Acadia, French colonists saw little to no violent conflict with Acadians and Mikmaqs. The rules of collective decision-making for settling land disputes favored consensus and greatly reduced the returns to conflict (Candela & Vincent, 2020). All parties had to come to a collective agreement, eschewing the formation of special interests who could potentially benefit from fighting and thus spill the external costs of collateral damage onto the rest of the population.

Conflicts of interest over natural resources (like beaver) constrained the types of land-use practices that emerged in Canadian North America. English trading posts around Hudson Bay responded to those constraints with different beaver-population management practices. In response to competition from French trappers, Hudson Bay Company managers, in remote Forts Albany and York, raised the price of furs paid to Native Americans which resulted in a larger harvest and a rapid depletion of the beaver population (Carlos & Lewis, 1993, 1999, 2001). Fort Churchill did not face the same level of competition in the supply of beaver pelt. Prices for furs in that region were more stable, and the beaver population was not depleted as quickly (ibid). Both colonial episodes in French Acadia and the Hudson Bay Company demonstrate how initial settlements, endowments, and incentives impacted the evolution of social agreements that helped define property rights and resolve property conflicts between Native Americans and European colonists.

Seventeenth-century colonial New England did not have the same types of consensus-building institutions. Rather, New England colonists had a circumscribed method for marking and claiming land tenure through royal prerogative, town charter, and protection from the Crown. Over the course of early colonial development, however, the Crown became distracted with civil war and unrest back home. Settlers applied similar rules and tools of land management they had used in England. They also settled and expanded with relative autonomy after receiving a royal grant.

Following a “mixed system” of land use, English settlers surveyed and demarcated land prior to settlement, exerted greater control over land development, but lacked uniform application, exacerbated by rough terrain (Libecap et al., 2011). While the template for scouting and demarcating land for settlement was somewhat standard, each township operated independently and was not coordinated by a colonial government or representative of the Crown (ibid). English colonial development imposed an exclusionary delineation of territorial boundaries or natural-resource rights which were often at odds with Algonquian tribes who used the land and natural resources for shared, multiple purposes (Pluymers, 2021).

Recent economic history literature has not examined the relationship between land-use disputes and violent conflict in the colonial New England setting. There is literature that discusses the relationship between land conflict and contract choice in a modern context. Alston and Bernardo (2010) analyze how land conflict and initial property endowments impact subsequent contract choice and the type of tenancy arrangement chosen in Brazil. Similarly, Conning and Robinson (2007) examine how property insecurity impacts the type of agricultural organization selected among competing claimants. They use a model of potential land reform to demonstrate how an agent’s expectations of property insecurity, instigated by the likelihood of land reform, are likely to modify their current choice of contract. Both papers measure the impact of property insecurity on subsequent contract choice. This paper examines the relationship in reverse. How does property insecurity combined with an increase in the area of contested land lead to conflict?

2 Historical background

After Jamestown was founded in 1607, English emigrants settled in the colonies of Plymouth and Massachusetts Bay. Pilgrims and Puritan settlers left Holland and England to escape religious-conformity restrictions imposed by the state religion of their home country. Disheartened by religious persecution, they wished to preserve their English identity and brought with them their own form of self-governance and legal framework (Winslow, 1646). As described by English jurist and politician, Sir William Blackstone, colonial America was a ‘conquered land’ in which ‘all the land in the kingdom is supposed to be holden, mediately or immediately, of the king’ (Byer, 2023). The absolute and exclusionary application of private property and land tenure extended from England to the colonies.

The area of land spanning the coastline of Maine to Long Island Sound included many Algonquian peoples from the Massachusett, settled around Massachusetts Bay, to the Wampanoag, settled around Cape Cod Bay, and the Narragansett in the west of modern day Rhode Island (Washburn, 1989). Over the course of the 17th century, English settlers established various diplomatic, commercial, and religious connections with surrounding tribes, sachems (male leaders), and sunksquaws (female leaders). European settlers and indigenous peoples conversed, intermarried, and formed treaties (Schultz & Tougias, 1999; Warren, 2018).

They also exchanged goods and land. As the children and grand-children of England’s first “Great Migration” came of age, by 1670 the English population burgeoned to over 60,000 people, almost double the Native New England population (Silverman, 2019). Early interactions between Algonquians and the English (1620–1670) were pacific. What sparked tensions and violent conflict was the rapid rise of English pastures, farms, and lumberyards—a dramatic increase in the demand for land.

At first, English settlers attempted to purchase land through various exchanges including manufactured goods or wampum, a string of beads used by Algonquian tribes to signal social ties and which European colonists repurposed as a form of currency (Brooks, 2018; Schultz & Tougias, 1999). These trades often took a strategic tone within the context of internecine power relations among European powers and rival Algonquian tribes (ibid). The source of conflict resulted from fractured and often competing concepts of property rights as well as the perceived encroachment or illegitimate purchase of land between English settlers and the Wampanoag, Narragansett, and Nipmuck. As the English settler population grew, more land was contested.

From the perspective of English settlers, an open expanse of land, a significantly depopulated Algonquian territory devastated by disease prior to the arrival of the English (Steckel et al., 2002), and initial comity with the Wampanoag made the subject of contract choice relatively simple. Under consent and charter from the Crown, most English requests for land were made through purchase ( Pagden, 2015; Roback, 1992). Challenges over occupancy with nearby Algonquian tribes were generally met through treaty, gift, or exchange². However, ambiguous land-use arrangements, illicit encroachment for land and timber, and rapid colonial settlement growth through land-intensive animal husbandry and demand for wood set the stage for one of the most devastating periods of conflict between the English and Algonquians in the Americas (Pluymers, 2021).

The Wampanoag, Narragansett, and Nipmuck settled inland, away from the coastlines, and along well-protected rivers and marshes (Washburn, 1989). Algonquian tribes had a deep knowledge of the terrain and exposed points of attack along coastlines where English colonists predominately settled. Generally, Algonquian forces were mobile and used their knowledge of the land to their advantage while the English typically fought in fixed points of defense near their settlements (Schultz & Tougias, 1999).

The introduction of livestock, expansion of land-intensive agriculture, demand for timber and rapid settlement growth characterized early colonial development in New England. Four years after the arrival of the Mayflower, Edward Winslow, one of the early Pilgrim Fathers, brought from England “three heifers and a bull, the first of any cattle of that kind in the land” (Anderson, 1994, p. 602). The task of improving the land was met as a general measurement of their prosperity. By 1627, Plymouth Colony had accumulated—either through husbandry or import—up to “fifteen animals, whose muscle power increased agricultural productivity” Anderson (1979 and PARP). In order to accommodate this land-intensive form of agriculture, Plymouth began expanding its borders beyond the boundaries of the compact village established at the time of the first Thanksgiving. They traded manufactures and agricultural products with Massasoit, sachem of the Wampanoag and father of Metacomet, in exchange for more land.

The most lethal and politically potent of these manufactured goods were matchlock and flintlock firearms. The proliferation of guns was initially spurred by early colonizers (Silverman, 2016). Colonial governments were late to enforce the arms trade; and even when they began imposing restrictions on the movement of guns, black markets rose. The arms build-up—predominantly spurred by Dutch, French, and English arms traders—revolutionized warfare for the Wampanoag. Metacomet,³ who the English called King Philip, saw the colonies as attempting to drive them off their land. Exacerbating tribal agitation, colonial governments banned the arms trade entirely and began confiscating arms from the Wampanoag (Lepore 1998 and Silver 2016). The confiscation of arms and the perception of conquest came at a time of increased tensions between English colonists and the Wampanoag, Nipmuck, and Narragansett.⁴

By the 1640s, colonial authorities in Plymouth and Connecticut’s General Court issued orders prohibiting English settlers from purchasing land or timber rights. Metacomet notified and submitted grievance to English colonial authorities that settlers were illicitly felling woods in a swamp in Wampanoag territory (Pluymers, 2021). Encroached timber rights, unsanctioned timber trade with Algonquians, and colonist incursions into Algonquian territory were major contributors to increased tensions. “The steady stream of trespasses and encroachments onto Wampanoag land escalated to the point where Metacom launched a war to defend their territory and sovereignty” (Virginia DeJohn Anderson, 1994 and Pluymers, 2021).

John Sassamon, a Massachusett raised by an English family, was a mediator with significant influence with both the English and Wampanoag. He served as an advisor to Metacomet and was in a favorable position to ease the rising tensions. It was not until Sassamon was found dead in Assawompset Swamp in the winter of 1676 when conflict erupted. While the nature of Sassamon’s death has never been confirmed, three of Metacomet’s closest advisors were arrested, tried, and executed by Plymouth colonists for Sassamon’s murder later that summer (ibid). Metacomet understood the colonists’ summary judgement and execution as a threat. It meant the Wampanoag were a subject people and beholden to a foreign form of justice (Silverman, 2016). King Philip’s War had begun.

3 Theoretical model of land tenure and conflict

Disagreements over land use can be resolved by jointly defining property rights in which both parties come to a peaceful resolution or lead to violent conflict. In a setting with competing conceptions of property rights and no mutually-agreed authority (or set of rules) to resolve disagreements over land use, the likelihood of violent conflict increases as the area of contestation—the overlapping geographic area where both parties lay claim to the same plot of land—spreads. Absent reliable institutions to resolve land-use disagreements, increases in the area of contestation between Algonquian tribal settlements and colonial English settlements are more likely to be targets of violent conflict.

3.1 “Raid or trade”

In Anderson and McChesney’s “raid or trade” model, a “zone of controversy” between disputed parties arises when a harm has been committed, like the trespass of land, and a claim has been pressed on the disputed property (1994). Along a section of land with uniform value, the “zone of controversy” arises when the marginal benefit of exerting force to acquire more land exceeds the marginal cost of fighting. The assumption of uniform land value refers to the permanent topographic features related to the land such as access to the coast or waterway, or the presence of a forest or pasture.

One historical example of this “zone of controversy” was in 1661 when English colonists negotiated a land acquisition from Pequots “in exchange for some convenient portions for their subsistence located elsewhere.” This afforded Pequots points of “ingress and egress” through the land, but they were specifically prohibited from using timber (Pluymers, 2021). There were different responsibilities and confounding expectations attached to the English and Pequots over how the land and its resources were used.

In a state of nature, absent a mutual authority to appeal to, each side chooses (1) how much to invest in the use of force to secure the disputed property⁵ or (2) whether to parley through peaceful negotiation and exchange. The decision to resolve disputes through violent conflict also comes with significant deadweight loss for both parties.

The surplus from negotiation (S) is a function of the costs of fighting (CF) and negotiation (CN):

$$\begin{aligned} S=CF-CN > 0 \end{aligned}$$

(1)

If the costs of fighting exceed the transaction costs of negotiation, the conditions for violence are not met (ibid).

Anderson and McChesney’s starting point is after the encroachment of European settlers on Indigenous land. The gains from fighting are determined by each party’s subjective value of the trespassed land. In this context, ex post English trespass, the surplus to negotiation is also a function of Algonquians’ willingness to retake the land. The surplus of negotiation for Algonquians is,

$$\begin{aligned} S_{A}=CF_{A}-GF_{A}-CN_{A} > 0 \end{aligned}$$

(2)

and the surplus of negotiation for English settlers to hold onto the land which could be lost from fighting $LF_{E}$ is,

$$\begin{aligned} S_{E}=CF_{E}+LF_{E}-CN_{E} > 0 \end{aligned}$$

(3)

If the combined surplus of negotiation is greater than zero then conflict is avoided.

$$\begin{aligned} S=S_{E}+S_{A}> 0 \end{aligned}$$

(4)

If person A has been trespassed and values the disputed territory more than person E then the surplus for negotiating is lower for person A compared to person E. Person A is more willing to resort to violence to recover the disputed property. For simplicity, assume each party values the disputed land equally.

$$\begin{aligned} T=GF_{A}=LF_{E} \end{aligned}$$

(5)

The losses and gains from fighting (excluding deadweight loss for both parties) net out to zero, returning to the inequality in Eq. 1 (from the perspective of both parties).

Under a fog of war, the prospects for winning or losing a violent conflict over territory T comes with uncertainty, potentially affording one party asymmetric information about the outcome of violent conflict. Introducing uncertainty into the equation, the probability Algonquians assign to gaining (retaking) the disputed land is $P_{A}$ and the probability the English assign to keeping (defending) the land is $P_{E}$. The probabilities are subjective and therefore may not sum to 100%. Each party may over or underestimate the outcome of violent conflict. The surplus for negotiation under uncertainty then becomes:

$$\begin{aligned} S=CF-CN -(P_{A}T) + (P_{E}T) > 0 \end{aligned}$$

(6)

The surplus of negotiating disappears and fighting occurs if the difference in assigned probabilities of success through violent conflict are greater than the difference in the costs of fighting and negotiation relative to the value of the disputed territory.

$$\begin{aligned} (P_{A}-P_{E}) < (CF-CN)/T \end{aligned}$$

(7)

If the costs of fighting (such as building battlements, loss of life, destruction of property) exceed the costs of negotiation (such as customs, language, treaty and exchange), then the difference in assigned probabilities must be positive for fighting to occur.

3.2 Extensions and implications of “raid or trade”

The empirical analysis in Sect. 4 focuses on historical examples of premises drawn from the “raid or trade” model: (1) after English settlers have trespassed, (2) after Algonquians have acknowledged this harm, and (3) a decision is made by both parties to either negotiate or resort to violence. Perhaps counterintuitively, ex post non-violent trespass, trespassers value negotiation more than the trespassed even if both parties value the disputed land equally, $GF_{A}=LF_{E}$ (See: Eqs. 2 and 3). This does not mean trespassers will stop trespassing. The model reflects a point in time when harm has been acknowledged and both parties decide how to proceed. Trespassing would also begin to slow in the face of violent pushback.

Section 4 covers three periods of conflict in and around English settlements in colonial New England. During later periods of sustained conflict, English fighters are sometimes on the “offensive” in and around their own settlements. For simplicity, all raids and battles, regardless of who is the attacker or the defender, are recorded as a conflict event.

Determinants of overlapping land claims and the outbreak of violence over Anglo-Algonquian land disputes include (1) the collective presence of colonies and county size (at the time of conflict), (2) the distance between a colonial settlement and a rival, tribal village, and (3) colonial settlement growth. While all of these factors are likely to increase the “zone of controversy” and lead to disputes over land, each factor has a different implication for the decision to resort to violence.

Prior to assessing the marginal trespass of disputed land within the “zone of controversy”, the initial collective colonial presence of a settled region would have two potential impacts on the inequality in Eq. 7. A larger colonial presence would narrow the divergence in expectations of violent conflict—underestimating the prospects for Algonquian success $\downarrow P_{A}$ and overestimating the prospects for English success $\uparrow P_{E}$ in the event of a conflict—reducing the left-hand side of the inequality. The size of nearby colonial presence would also increase the costs of fighting $\uparrow CF$ requiring more investment in offensive and defensive capabilities.

The costs of fighting scale with the number of people, the size of the territory being contested, and military capabilities (Baker, 2003). Imbalanced military capabilities do not directly imply a preference for the use of violence but indirectly alter the terms of negotiation (Umbeck, 1981). In general, the initial size of colonial presence would narrow divergent expectations on the likelihood of success through violence, greatly increase the costs of fighting for both sides, and reduce the likelihood of conflict near or an attack on a particular town.

The proximity of an English township to an Algonquian tribal settlement also alter the expectations and costs of conflict. Closer targets would increase the expectations of Algonquian success $\uparrow P_{A}$ and English failure $\downarrow P_{E}$ in the event of conflict. A shorter distance travelled also reduces the cost of attack $\downarrow CF$. Given the pattern of colonial and tribal settlement, townships settled on the colonial frontier were more susceptible to attack and often near Algonquian tribal settlements.

Finally, the growth of colonial settlements would also instigate conflict, causing more encroachment on surrounding land $\uparrow T$. The surplus of negotiation shrinks, and the viability of conflict under a fog of war becomes more likely, as the area of contested land increases. If the costs of fighting are greater than the costs of negotiation and the expectations of an Algonquian success are greater than the expectations of the English, settlement growth would instigate conflict. Taken together, English settlements in sparsely populated regions (counties), in close proximity to tribal settlements, and with high growth rates were more likely to instigate conflict. The next section explores the relationship of these factors using historical probates for New England townships and records of battles near or attacks on English settlements during three periods of Anglo-Algonquian conflict in the 17th century.

4 Cross-township analysis

The model and theoretical narrative above describe how the geographic distribution and growth of English townships⁶ increase the area of contested land, break down the possibility of peaceful resolution, and spur violent conflict. In order to measure how these determinants of land-use conflict interact, I examine the relationship between colonial presence, tribal settlement proximity, and the likelihood of conflict at the township level. I first quantify the magnitude of early colonial expansion across all settlements in colonial New England and identify whether a township was attacked (damaged or destroyed) by rival tribes or where a battle occurred during three periods of Anglo-Algonquian conflict in the 17th century: (1) the Pequot War of 1636–1638, (2) King Philip’s War of 1675–1676, and (3) the New England theater of the early part of King William’s War from 1688–1690.⁷

A first-best measurement of colonial expansion is the change in land area for each township over the preceding decades. Growth in colonial territory expanded the area of potentially contested land and spurred more land-use disagreements between the English and Algonquians. Detailing differences in land demarcation institutions across various European colonies worldwide, Gary Libecap et al. note that New England colonial settlements were centralized and occupied a “compact state of freeholders” marked within square townships of 6–10 miles and divided fields (2011). Surveying and the demarcation of land was typically administered within a year of a town grant and independent squatters away from the town center was limited. This uniform pattern of land demarcation for New England townships was weaker for settlements along the frontier or in rugged terrain (ibid).

The centralized “compact” New England township is the unit of analysis for comparing land-use conflict during this period. Unfortunately, data on cadastral⁸ size and development for each township is not readily available at the cross-settlement level (beyond just a handful of early colonial settlements). I use the growth rate in the population of farmers since the decade of initial founding as a proxy for land growth over the same time period.

4.1 Data

4.1.1 New England probates

The data for farmer population comes from a sample of colonial New England probate records from 1620 to 1700 (Main et al., 2013). The universe covers all deceased individuals—including landowners and landless tenants—in southern New England over this time period. The probate data also includes categories for each individual’s occupation, value of real property, wealth,⁹ debt, age, and sample weight by age group. The sample weight equals the inverse of the probability of selection for an individual colonist in a certain age group to die and be recorded in the probate sample (ibid). For example, a deceased individual is more likely to be older than younger. The probability of a probate record mentioning an older person is higher than a younger person. The sample weights for younger people are larger than the weights for older people. I construct a “farmer population” measure as the sum of sample weights for deceased individuals with a farmer occupation.¹⁰

The area of a township cadaster is a stock measurement of land; the number of deceased individuals reported in the probate is a flow approximating the colonial population. To account for this discrepancy, I aggregated the (deceased) farmer population¹¹ by decade. Annual growth rates had many gaps, were too variable, and did not reasonably represent changes in colonial territory.¹² The model assigns one decennial growth rate to each township. The decennial growth rates were computed following a straight-line approach.¹³

This measurement of farmer population is also more economically representative of land growth relative to real estate value. Seventeenth century frontier farming in the American Colonies can be characterized as a factor minimizing production function between land (capital) and labor: $F(K,N)=min[K,N]$. This type of production function treats land (K) and labor (N) as complements. It assumes that a given acre of land did not yield more output (or be anymore productive) after substituting production towards more labor. Given that assumption, any increase in agricultural production would need to be met with a one-to-one increase in both factors.

Appendix 3 attempts to reconcile the data limitations for using changes in farmer population as a proxy for changes in territorial footprint. Appendix Figs. 7 and 8 compare the farmer population growth rates observed in the probate data to archaeological data on pasture size for Plymouth township over the same time period. The decadal growth rates in farmer population for Plymouth show a similar pattern as the growth in pasture size. The two time series have a Pearson correlation coefficient of 0.825.

4.1.2 Historical accounts of conflict

Historical accounts for the three periods of conflict come from various sources. The source I use for the Pequot War comes from Alfred Cave (2003) who cites English colonists who recounted (e.g. Increase Mather) or participated (e.g. Captain John Mason) in the conflict. Nearly all of the English settlements and forts attacked during the Pequot War were along the coast and river of Connecticut. Two of the battles occurred at outposts such as Fort Mystic and Saybrook. Only one township that was attacked during the Pequot War appears in the probate data—Fort Mystic (Groton, MA).

There is more historical documentation for King Philip’s War—more than two generations after the arrival of the Mayflower. The secondary sources for colonial conflict come from Washburn (1989), Figure 2, and Starkey (1998), Figure 3. Detailed accounts of every battle fought in King Philip’s War are found in Schultz and Tougias (1999). They mostly rely on contemporary narratives, tombstones, and markers to indicate the locations of battles. The primary source that features prominently in secondary historical research is John Foster’s engraved “map of New England”—the first map printed of the English Colonies, surveyed by William Reed and illustrated in William Hubbard’s “Narrative of the Troubles with the Indians in New England” (1677).

Figure 1 highlights in red or yellow each of the townships that were damaged or destroyed during King Philip’s War. The source for township damage comes from a collection of accounts from the war mentioned in the primary and secondary sources cited above. The Washburn map, Figure 2, includes other areas of conflict in colonial New England 30 years before and after King Philip’s War. The Washburn (1989) map and Drake (1897) historical narrative are cited accounts of conflict for the New England theater during the early stages of King William’s War in 1688–1690.

No probate data is available for the Providence Plantations. The Main, Main & Lindert data only report probate records for the United Colonies of New England¹⁴: Connecticut, Massachusetts (including Maine), and New Hampshire. English settlers in Rhode Island, mostly Quakers, as well as Narragansett leaders who resided in the Providence Plantations either refused to join or remained neutral at the start of the war (Mandell, 2010). Narragansetts entered the war after the English attacked a Narragansett fortified camp, Canonchet’s Fort, in the Great Swamp Fight at the end of 1675 (Schultz & Tougias, 1999).

I then assigned a binary variable of 1 to any nearby battle or settlement attacked in the 17th century. Although Figure 1 distinguishes between destroyed and damaged towns, the relative magnitude of destruction is unknown. I consider any sign of war-related property damage or nearby battle as sufficient for indicating conflict. As mentioned in Sect. 3, while this predominately focuses on Algonquian attacks of English settlements, there are also battles fought near settlements that reverse the Algonquian-attacker and English-defender battle formation.

Figure 4 is a magnified map showing tribal settlements in yellow and “Indian praying towns” (Christian missions) in black (Lepore, 2018). Both the Starkey map Figure 3 and Lepore map Figure 4 were referenced to identify Algonquian tribal settlements.

5 Findings

Estimates of farmer populations were aggregated and growth rates of decade-over-decade changes in farmer population were constructed for every township over the period 1620–1700. Farmer population estimates, population growth rates, and the number of townships (reported in probates for the United Colonies) were then aggregated by colony and county (or region). The number of towns with nearby conflict—either a nearby battle or a raid on an English settlement—were also aggregated by region. Tables 1–4 summarize those findings during each period of conflict in 17th century New England: Pequot’s War, King Philip’s War, and King William’s War.

Table 1 17th century Colonial New England conflicts

Table 1 compares the probate aggregate statistics and conflict data across all three periods of conflict. Column 3 reports the number of towns that appear in the probate data for each time period. Column 4 shows the share of New England townships within 20 miles of a tribal settlement. Column 5 reports the average town growth¹⁵ in farmer population since the prior period of conflict. Over time, English presence increased on both extensive (number of towns) and intensive (population growth) margins. The biggest jump in growth, across both margins, was the period leading up to King Philip’s War in the 1670s. Among the three wars examined in the New England theater, King Philip’s War was the most violent conflict, both in terms of the share of towns attacked and the share of (farmer) population attacked. Tables 2–4 summarize the probate growth rates and conflict data for each conflict separately by colony and county.

Probate data for the 1630s is the most sparse and did not have enough frequent records to produce growth rates. The Pequot War occurred less than 20 years after the first Thanksgiving. Given those limitations, farmer populations and settlement size serves as a decent baseline for comparing two of the three factors that partially describe land-use conflict. By the 1630s, the largest colonial presence was in Connecticut, Plymouth, Rhode Island¹⁶. The theater of the Pequot War occurred in Connecticut, along the coastline of Long Island Sound and the Connecticut River. Setting aside single towns with reported probates in Boston and Rural Suffolk, Table 2 shows the relatively sparsely populated Hartford county had one township with a reported battle an attack on Wethersfield on the frontier outside of Hartford township.

The top panel of Table 3 summarizes the farmer population estimates and growth rates for each colony and county region during King Philip’s War. I present growth rates both on an intensive margin, with the median growth rate observed at the township level within a region, and on an extensive margin, with the weighted average growth of an entire region. The latter measure applies the same decade-over-decade growth rate calculation for the entire county. The bottom panel shows the number of towns within 20 miles of a tribal settlements and the number of towns attacked during King Philip’s War. Three counties report episodes of conflict: Hampshire and Rural Suffolk, Massachusetts—frontier counties outside of Essex and Boston—and Plymouth—the strategic and diplomatic center of activity leading up to the war.

Essex and Plymouth sustained the highest growth rates up until the war. Those counties also had the largest estimates for the population of farmers when conflict erupted. English settlements in Essex were closer to Boston harbor and far from Algonquian settlements. On an extensive margin, Essex had more towns than Plymouth by the 1670s and could sustain a larger population of farmers and a higher growth rate without encountering conflict. Plymouth, also large in terms of the number of towns and farmers, had settlements with much closer proximity to Algonquian tribes. Both Hampshire and Plymouth had half of their settlements near an Algonquian settlement with Hampshire in the western frontier of Massachusetts Colony experiencing conflict up and down the Connecticut River. The probate estimates also indicate a higher concentration of farmers per township in the areas where conflict occurred. Hampshire, Rural Suffolk, and Plymouth all had upwards of 100 farmers per township. Whereas, Connecticut had roughly 50 farmers per township, and New Hampshire even fewer at 17 per township. Essex, away from the main theaters of King Philip’s War, was able to develop and grow without suffering conflict.

The findings for the early period of King William’s War serves as a counterexample to King Philip’s War for the same New England region. The source of raids were more strategic, spurred by French colonial belligerents in coalition with Wabanaki Confederacy (an Algonquian people), and less motivated by English land encroachment (Drake, 1897). As observed in Table 4 and Figure 6, Wabanakis lead by a French military officer raided northeast New England settlements as a preemptive strike on perceived English encroachment into French Acadia (ibid). Given the strategic concerns of European powers, it is worth noting that Essex, which escaped conflict during King Philip’s War, continued to grow at a high rate but did experience conflict at Haverhill. Most of the French-directed Wabanaki raids occurred in New Hampshire and in what is today Maine. The only factor from the theoretical model in Sect. 3 that could partly explain conflict in New Hampshire is the proximity to a Wabanaki settlement. The strategic direction lead by French interests during King William’s War demonstrates a departure from the Algonquian-lead strategy of military targeting during King Philip’s War.

Table 2 Early colonial population and conflict during Pequot’s war 1630s

Table 3 Township growth and conflict during King Philip’s war 1670s

Table 4 Township growth and conflict during King William’s war 1680–1690s

6 Conclusion

Many factors contributed to Anglo-Algonquian conflict from the Pequot War in the 1630s, to the onset of King Philip’s War in the 1670s, to the early period of King William’s War in the 1680s–1690s: colonial expansion, disparate understanding of land use, the proliferation of guns, inter-colonial and inter-tribal rivalries, and the failure of diplomacy. Threats, such as weapons confiscation and perceived loss of land, drove the Wampanoag and other Algonquian tribes to violence. Peaceful resolutions to land disputes became more strained—in the absence of shared property-enforcement rules—as the area of contested land between Algonquian tribes and English settlers expanded.

This relationship was strongest during King Philip’s War, compared to the periods of conflict before and after. In the same New England theater, the Pequot and King William’s wars included other strategic concerns and co-belligerents such as competing Indigenous North American and European rivals. The relationship between contested land use and violent conflict was relatively dampened during the wars that bookended King Philip’s War.

Colonial growth in New England was rapid and geographically heterogenous in the early to mid 17th century. From 1620 to 1680, over 50 English townships were founded, and townships grew from a doubling in size to as much as a factor of 45. By 1680, the median township growth since founding was just over 200%. An English settlement’s proximity to a nearby Algonquian settlement also influenced the perceived expectations and realized costs of violent conflict. After accounting for proximity to the nearest Algonquian tribal settlement, English townships that grew more rapidly during the early stages of colonization were more likely to experience violent conflict.

Future research may consider expanding the list of potential sites of conflict to include English attacks on Algonquian settlements, exploring other factors that contributed to Anglo-Algonquian violence during the colonial period, and identifying points of contention over land rights, such as the Plymouth colonists’ felling of timber in woods thought to be held in commons with nearby Wampanoags. The combination of expanding colonial presence, proximity to potential rivals over land use, and the perception of land trespass in settings with overlapping and incongruent perceptions of property-rights institutions, like that of 17th century colonial New England, contributed to violent conflict.

Appendices

Appendix 1: Historical Accounts

1.1 English and algonquian interactions prior to King Philip’s war

See Table 5.

Table 5 A summary of secondary sources categorizing the types of interactions among Algonquian tribes and English settlers

Appendix 2: Figures

See Figs. 1, 2, 3, 4, 5 and 6.

Fig. 1

Towns Damaged or Destroyed. Source: American History Online

Fig. 2

Washburn’s Areas of Conflict. Source: Washburn, Vol. 4

Fig. 3

Starkey’s Areas of Conflict

Fig. 4

Hassanamesit and mission communities, with Missitekw, Cambridge, Boston. Source: Brooks, Lisa. Our Beloved Kin. 2018

Fig. 5

Probate Growth and King Philip’s Conflict Map

Fig. 6

Areas of Conflict During King William’s War.Sources: Wikimedia Commons; Sites of raids also mentioned in Cave (1996)

Appendix 3: Plymouth settlement pasture growth tabulations

The cross-township analysis suffers from an absence of data on the growth of land in English settlements as measured in surface area. In order to corroborate the magnitude of growth rates observed in the cross-township analysis—based on estimates of the population of farmers who entered probate—I also looked at archaeological data on livestock and pasture size over the same time period for Plymouth township to compare how growth in livestock-intensive land reconciles with the probate proxy variables used in the cross-township analysis. Using United States Department of Agriculture (USDA) guidance on pasture size capacity, I imputed the total number of pasture acres required to accommodate the number of livestock reported in the archaeological site.

1.1 Plymouth archaeological rediscovery project

Craig Chartier and other members of the Plymouth Archaeological Rediscovery Project (PARP) used faunal analysis to catalogue and substantiate the number of cattle remains recorded across three sites within Plymouth township. The faunal results reported in Fig. 7 show the total stock of cattle for each decade. Ratios of other types of livestock relative to each cow were also confirmed in the archaeological analysis. I applied those ratios to the total stock of cattle to derive the total number of livestock for each decade. The total number of livestock and decennial growth rate in the number of livestock are reported in the last two columns of the lower table in Fig. 7.

1.2 Pasture size

Using the USDA formula, I then imputed the pasture size required to accommodate the livestock from Fig. 7. The USDA formula is adapted for small-scale farm use and mixed operations including foraging and crop rotation (NRCS, USDA). These assumptions are more relevant for 17th century agriculture and colonial English farming methods. I scaled down the average size of each animal¹⁷ as well as reduced the utilization rate and average yield per acre to reflect a level of farm productivity that is closer to subsistence. Settlement growth, in this analysis, is entirely driven by the growth in cattle reported from archaeological evidence and the ratios of livestock weight.

The total number of acres required to house each type of livestock are reported by decade in Fig. 8. The assumptions I made for average livestock weight, utilization rate, grazing days per year, and average yield per acre are listed on the top left. The pasture size imputations reported in Fig. 8 are constructed using the USDA formula and total livestock figures reported in Figure C7. The decennial growth rate in pasture size for Plymouth township is reported in column 7 of Fig. 8. The highlighted records in Figure C8 compare the decennial growth rates of the faunal analysis to the growth in farmer population reported in the probate data for Plymouth which was used in the cross-township analysis. The growth rates show a similar pattern: a rapid expansion in 1630–1640, moderate growth in 1650–1660, followed by a decline in the aftermath of King Philip’s War in the 1670s. The Pearson correlation coefficient in decadal growth rates for Plymouth’s pasture size and farmer population is 0.825.

Fig. 7

Faunal Statistics in Plymouth Colony

Fig. 8

Pasture Growth Rates in Plymouth Colony