Counting People of Color: Worcester County, Massachusetts 1790-1860
A Preliminary Report
Jack Larkin
Old Sturbridge Village




“…a wide field of research, with much ground to traverse and little to gather, and that little so scattered and hidden in the mass of other matter, everywhere surrounding . Yet the returns were necessary, and must be had …”
John Milton Earle, 1859i


Counting people who have been made nearly invisible is a perilous process for any student of population. The history of people of color in New England is full of omissions and oversights, and certainly the history of population and social structure for African American and Native American people in 19th Massachusetts century is an area about which we know little. This is a first report on an effort in progress to advance that “wide field of research” by narrowing it a little, to focus on what we can learn systematically about the people of color of central Massachusetts – the fifty-odd towns of Worcester County – in first half of the nineteenth century. Throughout the nineteenth century, Worcester County’s people of color were a group of modest size – about .6% of the population on average, amounting in total to about the size of one of the county’s smallest towns. Their numbers were not large, but their significance - in terms of cultural endurance and persistence of Indian and African identities within “a sea of white faces” - is great.ii


In his A Statistical View of the Population of Massachusetts of 1846, pioneer demographer Jesse Chickering turned in his last chapter to look at the “colored population.” He maintained that the real Indians of Massachusetts were virtually extinct, that those of “pure African race” were very rare, and that the state’s people of color were a biologically and socially weak “mixture of races, of Africans, Indians and whites, in various degrees of purity … a circumstance regarded by physiologists as unfavorable to the increase of a healthy and hardy progeny.” For Chickering, unable to escape the conceptual structure his society employed for thinking about race, and operating with a quasi-biological concept of purity of blood, this was the defining characteristic of Massachusetts’s people of color. They were a mixed people, and it was thus quite possible that they would eventually disappear completely, leaving both their lineages extinct or nearly so.iii.


We know, of course, that he was wrong, not only in his presumption that races were real biological entities, but in his predictions about population. In Massachusetts today, the descendants of those he was discussing, people of Native and African American descent, are very much present. They have endured. A rich literature of historical investigation, most of it quite recent, has uncovered the important, and to some extent hidden histories of New England’s people of color in the eighteenth and nineteenth centuries. This paper returns to the population sources, although not to Chickering’s interpretation, and tries to look anew at the question of counting people of color in early and mid-nineteenth-century Massachusetts. As is usually the case in the historical study of people who have been marginalized, the sources are often both fragmentary and refractory, created and maintained by the dominant group, and subject to bias and perhaps extra carelessness.


The title of this piece, and its focus on people of color as a definable population group, do not at all reflect an acceptance of the notion that they - as seen by some white observers of the time - were only a jumble of fragmented identities. Clearly they knew who they were. John Milton Earle’s listings of Indian families and individuals in his Report of 1859 makes it clear that a structure of Indian group identity, naming, and self-recognition was very much in evidence, even if he, as a white outsider, did not fully understand it. Others who were counted in the census had a sense of their lineages from West Africa, as well as of generations spent in New England households, farms and shops. Still others combined Indian, African and Anglo-American traditions, identities and self-recognition in various ways. All who were listed as “free colored” struggled with the bleak and pervasive racism of nineteenth-century America, and the census counted them together.




This is a preliminary report based on the analysis of census data from 1790 to 1850. Starting with the initially opaque but increasingly informative censuses of 1790 through 1840, it looks at patterns of household size, geographical mobility, age structure, and the proportion of people of color living in white-headed households. Working with the far more detailed census of 1850, it also explores patterns of household composition, literacy and school attendance, occupational structure and property ownership, and fertility. In all these analyses, Worcester County’s total “free colored” population will be compared with a reasonably representative white population in the region – the people of the town of Shrewsbury.iv Because this is an initial report on work in progress this presentation has had to be more schematic than narrative. Eventually, a more complete and perhaps compelling version of this study will weave the stories of individuals and families around an expanded statistical framework.




Populations. It is widely understood, and needs no long rehearsal here, that the Indian peoples of southern New England underwent not only invasion and expropriation of their lands, but a demographic catastrophe in the late 16th, 17th and 18th centuries. They endured, but their numbers – from epidemic disease, warfare, voluntary and forced migration from the region – and the attendant ills of physical and cultural dislocation – less frequent marriage and lower fertility within marriage – diminished shockingly. In their homelands, substantial populations, once numbered in the thousands or tens of thousands, shrank over the long decades to be counted in the hundreds.v


The population of African descent in New England initially had a very different trajectory. African people came to the region as enslaved domestic workers and laborers for merchants, magistrates, and ministers. Rarely amounting to more than one or two in a household, they became widely scattered throughout New England communities, as the “servants” of a few elite families. At first greatly hindered by highly unbalanced sex ratios, they gradually established patterns of marriage, procreation and family life, although often in situations that kept the husband in one household and wife and children in another. For these reasons, marriage was less common and generally later, and the number of children considerably smaller, for enslaved New Englanders than for their white masters.vi


New England’s colonizing English settlers, by contrast, had demographic success virtually without parallel in early modern Europe or America. In the 17th and 18th centuries their population growth was explosive, doubling every 25 years, as they experienced a demographic regime with higher fertility and lower mortality than virtually anywhere else. Marriage rates and fecundity within marriage were higher than in England, France or the American South, and the life expectancies of white New Englanders, for infants, children and adults, were although low by modern standards, probably the best in the world at the time. In mid-18th century Sturbridge, Massachusetts, for example, married couples had an average of eight children born, and could expect to see six of these children survive to adulthood. Indeed, this burgeoning population growth was the force that almost from the beginning of New England settlement was pressing hard against the ecological boundaries of the New England landscape and the productive possibilities of New England farming – creating the continuing “hiving off” of New England communities and the settlers unending, relentless search for new land.


Pridefully riding this constantly rising tide, it was all too easy for white New Englanders to see people of color as a numerically inconsiderable minority, a remnant people, a people on their way to extinction. Over the years they frequently construed their fecundity, in sharp contrast to the numerical decline of an oppressed people, as signs of divine favor or biological superiority.vii


By the time that Chickering wrote, intermarriage between Indians, African Americans, and whites had long been a reality in Massachusetts. It may be something of an understatement to say that this is, and was, a complicated issue. An overbalance of males among slaves (due to patterns of forced migration) was in many places matched by a corresponding disproportion of females (due to warfare and seafaring that pulled men out of the community) in Native communities. Despite social and legal proscription, some white men and women, almost always people in the same straitened economic circumstances, continued to find marriage partners among people of color. Both African Americans and Indians were marginalized on account of skin color, cultural difference, and poverty, but in Massachusetts there were for a time some significant legal distinctions between them. Marriage between whites and “Negroes or Molattoes” seems always to have been proscribed, but the prohibition appears to have first been formalized in the Act of 1705 “For the better preventing of a spurious and mixt Issue.” viii Intermarriage between whites and Indians, however, was not prohibited, although the issue was debated at that time in the councils of government. In its provisions for punishment, in fact, the statute clearly assumes the enslaved status of “negroes or molattoes.” Indians then, however oppressed, were legally free to marry with either the dominant whites or the enslaved Africans. And to some extent they did so, at least sometimes purchasing the freedom of enslaved marriage partners.


But in 1786, the legal landscape for people of color changed strikingly. That year signaled the beginning of judicial emancipation for Massachusetts slaves, as the Supreme Judicial Court interpreted the Declaration of Rights of the state constitution of 1780. Yet while African Americans were entering into freedom, Natives were finding their legal status newly circumscribed. During the 1780s, a host of statutes were enacted to codify the law of the commonwealth under the new constitution. The new statute regulating marriage repealed the act of 1705, and replaced it with a blanket prohibition of marriages of whites “with any Negro, Indian or Mulatto.”ix Thus in the same year that the status of slave disappeared in Massachusetts, a broader racial category was created that included Indians together with recently freed slaves, others who had been manumitted earlier, offspring of any unions between them or between them and whites. All were defined together as “free colored,” and set off distinctly and dichotomously from whites. They were defined as the population that the census schedules sought separately to record. Chickering’s words of 1846 were ungentle, but he described what social definition and law had sought to create: “most of them are a mixed breed of whites with Indians and Negroes, and have been so, to a great degree, for the last fifty years and more.” x


Censuses. Thus in studying the “free colored” population on the census schedules, we are looking at people (and peoples) intertwined, who were counted and defined by others. Although historians are glad enough to have them, the earliest censuses of the Untied States provide frustratingly sparse information. Yet the census procedures and categories themselves are artifacts of racial definition, constructed on assumptions of difference and inferiority. Before 1850, they listed only the names of the heads of each household, while categorizing the (white) population in terms of age groups and sex. Yet they were even more opaque about people of color. For the years 1790, 1800, and 1810, the census schedules provided five age categories for white inhabitants of each sex but no information at all about the age and sex of “free colored” people, simply noting the total number in each household. In 1820, 1830 and 1840, age categories were introduced for the “free colored” but they were systematically different from those for whites. In 1820 the schedules had six age categories for the white population and four for the “free colored”; in 1830 and 1840, there were 13 age categories for whites and 6 for people of color. For every census, the classifications were considerably less precise, as if detailed information about them was less valuable; as well, the “break points” for specifying age groups were different, as if whites and people of color had different rates of maturation or availability to the labor force.xi .


Census taking in Massachusetts, conducted in communities that were geographically compact by American standards and by men who usually knew their localities well, was about as good as it got in the early nineteenth century Untied States. However, there is little doubt that people of color were undercounted to some extent. An uncertain number of them lived semi-migratory lives, making them quite difficult to find. A few Indian families were consistently not counted: some members of the Sprague and Pegan Nipmuc families of Dudley/Webster, living on the surviving reserve, seem ordinarily to have been excluded.. And then, perhaps most important, they were poor and often living on the geographical fringes of their towns, leading to the “overlooking” of marginal people that is a feature of virtually every process of enumeration. John Milton Earle noted of his own efforts to locate Indian families and individuals in Massachusetts that “The difficulty of tracing them is much increased by their humble position and obscure station in life, known only to a few directly abut them, and those mostly persons whose position in the community is similar to their own.”xii.


But if there is no doubt that the census takers missed some Indian families, there is also no doubt that they included many. In every census for Worcester County between 1800 and 1850, we can link some enumerated heads of households to vital records, town records, and genealogical records that identify individuals as Indian. For 1840 and 1850, we can find many significant linkages between names on the census and names listed in John Milton Earle’s Report of 1859. For the 1840 census, we can link 32% of the people of color heading households on the census to Indian identities; for the 1850 census, we can link 42%. For the Hasssanamisco and Dudley Nipmuc communities, 80% and 65% of the household heads listed in 1859 in Earle’s Report respectively, can be found nine years earlier in the 1850 census. Thus census “capture” of the Central Massachusetts Indians although imperfect, is substantial.


[Charts 1-5]


Families and Households of Color. Poverty, a precarious living on the margins of rural communities, and the pain of racist scorn and contempt were the common experience of people of color in Massachusetts . Despite their partial legal equality (the vote for taxpaying males, but bans on intermarriage and militia service), as a white observer noted, “wherever they go a sneer is passed upon them, as if this sportive inhumanity were an act of merit.” At the heart of a number of the accounts we have of growing up as a person of color in rural New England in the nineteenth century is a broken family narrative. In his virtually unique autobiography, A Son of the Forest, William Apess tells the painful story of his early separation from his parents, his grandparents’ alcoholism, inability to care for him and physical abuse, and his subsequent life in a string of white households as a bound “Indian boy.”xiii The story of Angela Sprague’s childhood in central Massachusetts has similar resonances – the early death of both her parents on the Nipmuc Pegan reservation in Webster, the subsequent death of her grandmother who had given her care and protection, and her coming to live in the household of the white Bemis family of Sturbridge – as “adopted daughter (a white woman’s later reminiscence) or “servant” (from the census schedules for the Bemis household) xiv There is as well the comparably tangled family narrative in Harriet Wilson’s Our Nig, or Sketches from the Life of a Free Black The story of the “beautiful mulatto” Frado, whose African American father dies of consumption, whose white mother descends into poverty and despair and abandons her, and who becomes a more or less involuntary, and often abused, servant in the household of the cruel Mrs. Bellmont. xv These sources present us with wrenching portraits of families struggling and often breaking under the strain of poverty, illness, and oppression. Imperfect as it is, the census data may provide us with a way to set this testimony in perspective, as we look for measures that will tell us something about the circumstances of life for a broader and more representative group of individuals and families. This closer look becomes possible with the 1850 census, the first fully nominative one for the United States.


Mobility: “we expect their localities to be more changeful.” Tracking household heads from one census to another allows us to compare patterns of geographic mobility. Americans in general were known to European observers as a traveling and migratory people, but central Massachusetts people of color were exceptionally so. Consistently over the decades, household heads of color were about twice as likely as white household heads from a sample of rural towns in Worcester County (Shrewsbury, Sturbridge, Barre) to leave their communities from one census to the next. Ten-year persistence rates for household heads of color ranged from 25% to 30%, while those for white heads of household averaged just over 60%. For some people of color this high degree of mobility/transience may well have been linked to a continuing preferences for a partially migratory way of life. Chickering, who was in some ways an insightful sociological observer, suggested an explanation in terms of relative social isolation. (Although he was unable to measure population turnover, he seems to have assumed that there would be a significant difference.) He thought that “on account of their scattered position” and “their want of …easy intercourse with each other, by which they are deprived of social enjoyments” people of color would be more transient and expected “their localities to be more changeful, and their increase less regular from year to year.”xvi But a variety of studies of geographic mobility for largely white New England communities suggest an additional interpretation – a lack of economic, as opposed to experiential and spiritual, connection to the land. In terms of propensity to move, household heads of color of all ages were about as transient as propertyless white men in their twenties – exhibiting comparable mobility patterns to farm laborers in rural New England.xvii


[Chart 6]


Wealth and Work: “they continue poor …“they are excluded from the more honorable and profitable employments.” In 1850 the census began to collect information about the value of real estate owned by individuals, as well as about male occupations. Unsurprisingly, comparative analysis of the 1850 data on wealth holding shows a stark contrast between people of color and ordinary rural whites. Fewer than 20% of all household heads of color owned real estate of any value. Fewer than 6% owned property worth $1000 or more. Only 3% owned $1500 or more – equivalent to a small central Massachusetts farm. In Shrewsbury, 50% of all men over 20 owned some real property, 30% owned property worth at least $1000, and 22% owned the equivalent of a small farm. The contrast is particularly striking when looked at over the life cycle. It seems that only rarely did men of color in Worcester County have ownership of land at any time in their lives. When wealth holding is broken down by age, the data strongly suggest (as do other studies of the rural American North in the 19th century) that Shrewsbury’s white men accumulated property as they advanced in years. Although relatively few men in their 20s owned land, each successive age cohort showed a higher proportion of landowners. The evidence is very strong that this rarely happened for men of color in Massachusetts; men in their forties and fifties were little more likely to own property than men just starting out in life.


[Chart 7]


If few men of color owned real property, none had professional or mercantile occupations. The largest number of men of color in Worcester County, nearly 50%, were laborers and landless farmers. Next came a significant group (15%) of barbers and workers in hotels, tavern and stables in the county’s largest communities. Following that were men (12%) who had “no occupation” listed after their names – a designation that sometimes surely reflected age or incapacity, but might also have reflected a combination of activities – trapping, fishing, the making of brooms, baskets and wooden wares – that could not be easily explained to the census takers. 11% were shoemakers and bootmakers. “Sale” shoemaking for the mass market as organized in central Massachusetts was a semi-skilled trade that was accessible to some men of color in New England. 6 % practiced more highly skilled “mechanic” trades –blacksmith, carpenter, cooper, upholsterer, trades that were generally, as Frederick Douglass noted of Massachusetts, difficult of access for men of color because of the unwillingness of white mechanics to train them or work alongside them. A few men of color were listed as basket makers, a traditional Native craft, and as herbal physicians – a role played by healers in both the African and Indian traditions.


Occupation 1850 People of Color Shrewsbury (whites)
Laborers and Landless Farmers
47%
10%

Barbers and other Service Trades
15%
0%

No Occupation
12%
5%

Shoemakers
11%
20%

Artisans
6%
19%

Farmers
5%
40%

Basket makers
2%
0%

Herbal Physicians
2%
0%

Merchants/Professionals
0%
6%



Household Structure: An Emerging Pattern of Independence. Through the early 19th century, poor and marginal New England families of all races frequently sent young family members out to live and work in the households of the more prosperous. This was a particularly pervasive experience for people of color. The condition of household slaves, of course, was in one sense an extreme, involuntary form of this pattern. William Apess’ account of his childhood as a virtually orphaned and bound-out “Indian boy” in white households suggests that it was common for Indian children as well; indeed, John Milton Earle noted frequent Indian employment and residence in white households and praised it as a beneficial mechanism of assimilation.xviii


The census schedules bear this out. In the four censuses between 1790 and 1830 (as late as some 45 years after the abolition of slavery) a very high proportion of Worcester County’s total population of color was enumerated as living in white-headed households – ranging from 30% to 35%. About one-third of all people of color were thus living as dependents – household helps, laborers – in white households. A more precise look at age structure (possible for the censuses of 1820, 1830 and later) reveals that great majority of those living in white-headed households were young people between 10 and 29. What this means is that through 1830 well over half of all older girls and boys, young men and women of color, were living as dependents in households not their own, under the economic and social control of whites. Over those same decades the mean size of households headed by people of color was consistently substantially smaller than that of white headed households. This of course is consistent with a pattern of “exporting,” or sending large numbers of young people to live in white households at early ages.


But the census schedules for central Massachusetts track a remarkable change in household structure starting in 1840. The proportion of people of color living in white-headed households dropped below 20% in 1840 and declined to 12% by 1850. A comparison of the 1830 and 1840 data, shows that most of the decrease was concentrated in the younger age groups. Along with this went a related change in household size. The mean size of households headed by people of color, previously consistently lower than for whites, began to rise. In 1840 it was almost equal to mean household size for whites; by 1850 it was higher. The trend seems clear: families of color were increasingly keeping their children at home rather than sending them out to live and work elsewhere. In this it appears that they were following, a bit belatedly, the same pattern as white New England households had evidenced from the end of the eighteenth century – the decline of binding out and living out for young children. These trends attest to a convergence of structure between white households and households of color. For families of color, it may have been gradually improving economic status and household stability that made this possible. A broader overall explanation has to do with a cultural phenomenon – the emergence of a view of childhood as a uniquely malleable time and an expanding emphasis on nurturing, supervising motherhood within the family. xix


[Charts 8,9,10]]


Household Composition in 1850. The census of 1850 offers a more detailed comparative portrait of households of color, and here we see broad evidence of stable family life. Households of color were considerably more likely to be headed by women at all stages of the family cycle; but it is important to note that this was true only of a minority of households. 70% of all households of color had both spouses/parents living together; although this was somewhat lower than the figure for white Shrewsbury households (88%), it represented a substantial majority. Households of color and Shrewsbury households had the same percentage – about 5% - of families where children were clearly being raised by grandparents alone. In 1850, children of color under age 15 were just as likely – 93% - as white Shrewsbury children to be living in their families of origin – another strong indication that the sending out or binding out of young children of color, particularly those 10 to 14, had significantly declined. For older children, there were significant differences in residential and household status. Youths of color ages 15 to 19 were substantially more likely to be living and working outside their parental household than their white counterparts. However, this proportion, although higher than that for whites, was significantly reduced from the extremely high rates of youth of color who were “outliving” and “outworking” before 1840.


[Charts 11-14]




Literacy and Schooling in 1850. The 1850 census also reveals that families of color were also able to provide basic schooling and literacy for their children. 75% of children of color ages 5-14 were listed as having attended school at some point during the census year; 85% of those ages 7-9 were in school. These figures were lower than those for Shrewsbury whites, about 95% of whose children ages 5-14, and 98% ages 7-9, were in school, but would have been impressive in most other places in rural America. The differences can probably be accounted for by a combination of family need, greater residential mobility, and informal barriers of discrimination that kept some children of color out of school. Yet taken as a whole, these figures indicate that the great majority of children of color not only had access to schooling but were schooled.


Essentially the same story is told by the census data for literacy. 92% of people of color above the age of 14 were recorded as able “to read and write.” Though somewhat lower than the 98% recorded for Shrewsbury whites, it is also an impressive figure. Even with the heavy freight of racial discrimination, central Massachusetts seems to have become “the land of schools’ as Henry Ward Dana of Shrewsbury put it, for children of color as well as for whites. When analyzed by age, the literacy figures suggest that access to schooling may have modestly increased over the first half of the 19th century. Illiteracy for the oldest age cohort of people of color – those over 60, born in the previous century – was 12%; for those between 40 and 60, it was 9%; for those under 40 it was 5%.
[Charts 15-16]


Overall, the census data paints a picture, hardly of wealth, but of largely intact families, children at home with their mothers, and schooling. In important ways it is at odds with much testimony that was inclined to see the lives of all people of color as marginal and problematic. Thus it is more than interesting to note John Milton Earle’s observation about the domestic life of Massachusetts Indians in 1859. Its condescension is jarring, but its content actually resonates with the patterns we have just described. “The people are nearly all,” he wrote, “comparatively poor,” but he found significant signs of domestic order: “some of their dwellings will compare, not unfavorably, with those of other people of equal means, in their furniture and order in which they are kept.” He continued with praise for the tenor of their household arrangements: “I have repeatedly partaken of their hospitality, and their tables, their cookery, their lodgings, and whatever else came under my observation, in their humble dwellings, exhibited a neatness not excelled in the mansions of the more affluent whites.”xx Whatever Earle’s biases and deficiencies, it seems clear that he was seeing respectable if modest houses and functioning households, not disorganized families in hovels.


Fertility and Mortality: “not likely to increase much in Massachusetts.” Jesse Chickering’s pessimistic forecast about Massachusetts’ people of color clearly assumed that they would continue to experience relative and perhaps absolute demographic decline, with relatively few marriages and low marital fertility. Their chances for “a healthy and hardy progeny,” he asserted, were low, and believed that the incomplete vital statistics he had seen “strongly indicate the great inferiority of the proportion of the births of colored children to that of the whites.”xxi For Massachusetts in the eighteenth century, there is some demographic evidence; a recent study of the Indian town of Natick suggests that the number of children per family size was significantly smaller for Natick than for the neighboring white community of Dedham.


Our best evidence for understanding of the demographic patterns of people of color would come from a substantial time series of family reconstitutions that would allow for a direct comparison with the abundant data for rural New England’s white population. Unfortunately, we do not have such evidence; the records are far too fragmentary. But another comparative look at the 1850 census offers us some useful measures.


Indeed a look at the 1850 evidence, gathered just four years after Chickering wrote, is quite surprising. When the age structure of Worcester County’s people of color for 1850 is compared with the age structure of the county as a whole, or that of the single community of Shrewsbury, the results are striking. In both cases the population of color had somewhat higher proportions of children – the very opposite of the statistical “signature” of a group with a declining birth rate. This observation is greatly reinforced by a more detailed comparison of indirect measures of fertility between the 1850 families of color and the families of the rural community of Shrewsbury. When we look at these two indices:


q
Proportion of households with children under 10
q
Number of children under 10 per women 20-49


We again find that they are somewhat higher for people of color. Over the previous couple of decades, it strongly appears that Worcester County’s women of color were actually bearing more children than the women of Shrewsbury. These findings runs directly counter to Chickering’s assumptions about the continuation of low fertility for people of color.


Certainly an increase in the birth rate from what were clearly quite low levels in the eighteenth century resonates with other trends that indicate the achievement of greater family stability by mid-century. This evidence needs considerable further exploration, particularly analysis of comparable data for later census years – but it is tantalizing, and suggests a somewhat speculative interpretation of demographic change: The birth rate for people of color in Massachusetts had begun to increase during the early nineteenth century, as families were able to reverse some of the damage done by centuries of slavery, disease, war and dislocation. People of color were establishing more stable patterns of family life, patterns that allowed the maintenance of intact households over the long term. Greater stability kept couples together longer, encouraged marriage and led to the birth of more children. Certainly the population of color in Worcester County, and in Massachusetts as a whole, after very slow growth and stagnation in the 1790-1830 years, began to register a steady decennial increase from 1840 on. .


Our assumption is that the fertility patterns of people of color and whites converged over the course of the first half of the nineteenth century, just as their household patterns did. Fertility for families of color increased; fertility for white families continued a pattern of decrease that had begun in the late eighteenth century (and would continue, unbroken except for the baby boom, until the present.). Our next assumption is that convergence continued later in the nineteenth century as the birth rate for people of color almost certainly began to decrease, paralleling the continued decline in the white birth rate. xxii It is likely then, that after increasing the size of their families, parents of color later in the nineteenth century began to limit the number of their children, probably for the same reasons (reducing the economic burden of establishing children in life, women’s interest in controlling pregnancy) as their white neighbors. Confirming (or refuting) this interpretation will require a good deal of additional research – but it offers one way of understanding the population dynamics of people of color in the nineteenth century.


[Charts 17-23]


For understanding the fertility of people of color there are some clues; mortality presents a greater puzzle. Taken as a whole, the life expectancies of the people of rural Massachusetts were about as high as could be found anywhere in the world in the nineteenth century; still, they were painfully low by modern standards, with nearly 30% of all children failing to reach adulthood. This old regime of mortality was somewhat worse for those who lived in Boston and a few other large cities, and it did not really begin to change until quite late in the nineteenth century. Jesse Chickering believed that the mortality rates for people of color in Massachusetts were higher than those for whites. “It is remarked,” he wrote, “that a larger proportion of colored than of white persons are, in early life, found to be subjects of fatal disease.” xxiii There is what demographers would call “anecdotal evidence” to buttress this claim. Certainly there are accounts of the lives of people of color at this time that describe early deaths, ascribing them to the toll taken by poverty, drink, and consumption. Timothy Dwight around 1800 pontificated about “the smallness of the number of children who survive their childhood” among the Indians of southern New England.xxiv More concretely, for the eighteenth century, Daniel Mandell’s study of Natick suggests that Indian child mortality rates were higher than for whites, and David Hackett Fischer has found evidence to indicate that mortality for New England slaves was higher as well. But given the evidence currently available for the nineteenth century the truth is that it is impossible to know. Stories about early death are both wrenching and individually plausible, but are difficult to evaluate, given the small size of the population of color, against the background level of overall mortality for nineteenth century Massachusetts. xxv It is at least possible that in keeping with the other trends suggested by this study the mortality of people of color in the nineteenth century was actually improving. (Only detailed work combining the censuses with the post-1846 Massachusetts birth and death registers would offer some hope of resolving this question – a daunting task.)


Identity. One additional complication of studying the population history of people made nearly invisible is that their numbers may well change by processes of social and cultural recognition rather than demography. Over time, it likely that many individuals from families of color made transitions out of a socially and economically difficult status. Although racism and social and economic discrimination against people of color were severe, social boundaries among poor and marginal folk in Massachusetts were porous enough, that a gradual transition out of “colored” status was possible for many over the space of a generation or two. In Massachusetts, a decision of the Supreme Judicial Court in 1810 seems to have recognized and confirmed this process, creating a sort of legal transition zone between “colored” and “white. It declared, in effect, that the children of two “mulatto” parents, or of a white parent and a “mulatto” parent, were not themselves mulattos in the eyes of the law. Since Massachusetts law made no recognition of more precise categories of color, such children would then be white by default.xxvi Three decades later, in 1843, the state legislature removed all racial restrictions on marriage by repealing the statute of 1786.xxvii


The continuing reality of intermarriage between whites and people of color throughout the first half of the nineteenth century is clearly evidenced in the racial designations of the 1850 census. 38% of Worcester County’s people of color were enumerated as “mulattos; the proportion was 30% for those over 50 years old and over, but was substantially higher – 45% - for those under 30. Out of 99 married couples identified in the census that included persons of color, six were intermarried – in five cases, a white woman was married to a man of color; in the sixth, a “mulatto” woman was married to a white man. Under Massachusetts law their children would have been recognized as white; the social recognition of skin color was different, and the census takers listed the children as mulattos. Yet over another generation or two some of their descendants would no longer be easily recognizable – either by the law or the census - as people of color. Chickering clearly believed that this process would join with the demographic ones of fertility and mortality to diminish the people of color, arguing that “a further mixture with whites will, from time to time, cause a portion of them to be undistinguishable in the community from the whites themselves.”xxviii But again we must say that he was wrong; amidst all these processes of change, they, or at least a great portion of them, did not become “undistinguishable.” In various ways, they maintained their memories, traditions and identities, sometimes holding them close, or, as now, vigorously reasserting them.. Ultimately, then, we are back in territory where the census cannot reach. Identity and survival in a social and cultural sense, cannot ultimately be categorized or enumerated.


This study’s preliminary findings might best be summarized this way: Although intense and bitter racism continued to shadow their lives, the analysis of data from the census schedule suggests that by mid-century the people of color in Central Massachusetts were living in a world that was in some significant ways better than it had been at the end of the eighteenth century.. Slavery had disappeared, and long-standing patterns of outliving in white households were waning. Poverty, lack of access to the land, and confinement in a narrow range of occupations persisted. But literacy was substantial and school attendance widespread. The statistics of household composition reflected the achievement of significant family stability. Measures of fertility suggest that birth rates were rising well beyond the level of earlier times. At the beginning of the period of this study the people of color were more or less legally defined as all those forbidden to marry with whites; by its end, all legal barriers to intermarriage had been removed.


Clearly, the people of color of Massachusetts survived and adapted under difficult circumstances, defying predictions that they would decline or disappear. This essay has tried to sketch out some of the paths they took on that nineteenth-century journey of endurance




Notes
Endnotes
i John Milton Earle, Report … Concerning the Indians of the Commonwealth, Under the Act of April 6, 1859 (Boston: William White, Printer to the State, 1861), 6.


ii For a powerfully argued study of Native American peoples in Central Massachusetts that refutes the mythology of disappearance and establishes their enduring significance and the signifance of their endurance, see Thomas L. Doughton, “Unseen Neighbors: Native Americans of Central Massachusetts: A People Who Had ‘Vanished’,” in Colin G. Calloway, ed., After King Philip’s War: Presence and Persistence in Indian New England (Hanover, N.H. and London: Dartmouth College/University Press of New England, 1997), 207-230.

iii Jesse Chickering, A Statistical View of the Population of Massachusetts, From 1765 to 1840 (Boston: Charles C. Little and James Brown, 1846) 157.


iv As part of the long-term research effort of Old Sturbridge Village to understand and interpret the diversity of life in early nineteenth-century rural New England, the museum is currently undertaking a comparative study of household and family structure of the people of color in Worcester County, Massachusetts 1790-1860. The primary evidence for this study consists of two population databases, both based on the Federal Censuses of Population for Worcester County, Massachusetts, 1790-1850. The first is a database for the town of Shrewsbury. The second is a database for the “Free Colored” population of Worcester County, drawing from schedules for all the towns in the county. Both were created from the printed transcript of the census schedules for 1790 and the manuscript census schedules (on microfilm) for 1800, 1810, 1820, 1830, 1840, and 1850. The printed schedules and census microfilms were consulted at the Old Sturbridge Village Research Library. Transcription and data entry were undertaken as part of the Tradition and Transformation Project at Old Sturbridge Village, funded in part by a grant from the National Endowment for the Humanities. Special thanks also go to Mary Hill Stone for additional assistance. Analysis of the data was undertaken with SPSS. I would like to thank my colleague Ed Hood, Research Historian at Old Sturbridge Village, for numerous illuminating discussions of the issues dealt with by this paper over the last several years. The study seeks to put this data in comparative perspective with what is know about the Massachusetts and central Massachusetts population as a whole from the extensive scholarly literature on these topics for Massachusetts, from aggregate state and federal census data, and from detailed Old Sturbridge Village studies of the towns of Sturbridge and Shrewsbury.


v See Bruce G. Trigger and William R. Swagerty, “Entertaining Strangers: North America in the Sixteenth Century,” in Cambridge History of the Native Peoples of the Americas, Vol. 1, North America ( Cambridge: Cambridge University Press, 1996) ed. Bruce G. Trigger and Wilcomb E. Washburn, part 1, 361-369, 389-391; Neil Salisbury, “Native People and European Settlers in Eastern North America, 1600-1783,” in idem, 455-456. Colin G. Calloway, New Worlds for All: Indians, Europeans and the Remaking of Early America,(Baltimore: The Johns Hopkins University Press, 1997) ch. 2, “Healing and Disease,” 24-41. See also Bert Salwen, “Indians of Southern New England and Long Island: Early Period,” Handbook of North American Indians, gen. ed., William C. Sturtevant, Vol. 15 Northeast ed. Bruce G. Trigger (Washington: Smithsonian Institution Press, 1977) 160-176; Laura E. Conkey, Ethel Boissevain, and Ives Goddard, “Indians of Southern New England and Long Island: Early Period,” 177-189; William S. Simmons, “Narragansett,” 190-197.-176.


vi For a discussion of African Americn population and family structure, see William D. Piersen, Black Yankees: The Development of an Afro-American Subculture in Eighteenth-Century New England (Amherst, University of Massachusetts Press, 1988) ch. 1,2.. There is a substantial amount of population data for New England’s people of color developed in Roger Parks, “Early New England and the Negro,” unpublished research paper, Old Sturbridge Village, 1969 – a good work of scholarship that in its assumptions about the identity of people of color reflects the historiographical framework of thirty years ago.


vii For an overview, see Robert V. Wells, Uncle Sam’s Family: Issues and Perspectives in American Demographic History (Albany: State University of New York Press, 1986); Jack Larkin, The Reshaping of Everyday Life 1790-1840 (New York: HarperCollins, 1988) 62-104; for New England specifically, see Daniel Scott Smith, “The Demographic History of Colonial New England,” Journal of Economic History 32, no. 1 (March 1972): 165-83 Maris A. Vinovskis, Fertility in Massachusetts: From the Revolution to the Civil War (New York: Academic Press, 1981; Nancy Osterud and John Fulton, “Family Limitation and Age at Marriage: Fertility Decline in Sturbridge, Massachusetts 1730-1850,” Population Studies 40, No. 3 (1977):481-494.


viii Acts and Laws of His Majesty’s Province of the Massachusetts-Bay in New England (Boston, 1759) Anne IV, Chapter VI (1705).


ix “An Act for the Orderly Solemnization of Marriages,”June 22, 1786, Section 7 The Laws of the Commonwealth of Massachusetts from November 28, 1780 to February 28, 1807 (Boston, 1807) I 321-323.
x Chickering, A Statistical View, 112. Daniel R. Mandell, Behind the Frontier: Indians in Eighteenth Century Massachusetts (Lincoln and London: University of Nebraska Press, 1996), ch. 6, “Indians and their mixt posterity,”164-202, has an extensive discussion of intermarriage patterns for Massachusetts Indians.


xi The changing structure of the schedules of population can be examined in: Returns of the Whole Number of Persons in the United States … 1790 (Washington, 1802); Returns of the Whole Number of Persons in the United States … 1800 (Washington, 1802); Aggregate Amount of Persons within the United States in the year 1810 (Washington, 1811); Census of 1820, Fourth Census Book One (Washington, 1821); Fifth Cenus, Enumerating the Inhabitants of the United States 1830 (Washington, 1831); Compendium of the Sixth Census .. of the Untied States 1840 (Washington, 1841; The Seventh Census of the United States 1850 (Washington, 1853).


xii Earle, Report … Concernng the Indians of the Commonwealth, 7.


xiii William Apes(s) A Son of the Forest: The Experience of William Apes, a Native of the Forest, Written by Himself 2nd ed., revised and corrected (New York: published by the author, 1831) 7-15. For a penetrating analysis of Apes’s autobiography, and the conditions of Native American life in early nineteenth-century New England, see Barry O’Connell, “Introduction,” in O’Connell, ed., On Our Own Ground: The Complete Writings of William Apess, A Pequot (Amherst: University of Massachusetts Press, 1992) xiii-lxxvii; and O’Connell, “William Apess and the Survival of the Pequot People,” in Algonkians of New England: Past and Present, Annual Proceedings of the Dublin Seminar for New England Folklife 1991 (Boston: Boston University Press, 1994) 89-101.


xivFor the Sprague family story, see the reminiscence of Octavia M. Sweetser, 1949, in a letter, Octavia Sweetser to Ruth M. Wells, August 30, 1949, Old Sturbridge Village Archives. And [Helen Holley] “Angela Sprague Leach: Her Story” Brimfield (Massachusetts)
xv Harriet E. Wilson, Our Nig, or, Sketches from the life of a free Black : in a two-story white house, north, showing that slavery's shadows fall even there (Boston: Rand and Avery, 1859).
xvi Chickering, A Statistical View, 137.


xvii Robert Doherty, Society and Power: Five New England Towns 1800-1860 (Amherst: University of Massachusetts Press, 1977) 30-45 and 46-55; Jack Larkin, "'Labor is the Great Thing in Farming': the Farm Laborers of the Ward Family of Shrewsbury, Massachusetts 1787-1860, " in Richard D. Brown, ed., Farm Labor in Southern New England During the Agricultural-Industrial Transition (Worcester: American Antiquarian Society, 1990) 189-226.


xviii Earle, Report … Concerning the Indians of the Commonwealth (Boston: William white, Printer to the State, 1861) 11.


xix See Larkin, Reshaping and Children Everywhere: Dimensions of Childhood in Early New England (Stubridge: Old Sturbridge Village, 1988). Asa Sheldon, who was a bound child in the first decade of the nineenthe century, criticizes the practice from the vantage point of midcentury; see The Life of Asa Sheldon, Wilmington Famer, (Boston, 1863).


xx Earle, Report … Concernng the Indians of the Commonwealth, 11.


xxi Chickering, A Statistical View, 137

xxii Fertility measures for Shrewsbury whites, in contrast, reflected pervasive change in the opposite direction. White birth rates in New England, although still high by modern standards, were declining from 1790 on. New England whites were undergoing their own long-term revolution in childbearing, marked by the adoption of conscious family limitation and a steady gradual decline in family size See again Wells, Uncle Sam’s Family, Vinovskis, Fertility in Massachusetts, and Osterud and Fulton, “Family Limitation and Age at Marriage.”


xxiii Chickering, A Statistical View, 157.


xxiv Timothy Dwight, Travels in New-England and New-York ed. Barbara Miller Solomon 4 vols. (Cambridge, Mass.: Harvard University Press, 1969) III, 22.


xxv Well, Uncle Sam’s Family, ; Larkin, Reshaping of Everyday Life, 62-85, and “No Force Can Death Resist”: Some Perspectives on Infant and Child Mortality in American History,” unpublished paper, Old Sturbridge Village, 2000; Maris A. Vinovskis, “The Jacobson Life Table of 1850: A Critical Reexamination from a Massachusetts Perspective,” Journal of Interdisciplinary History 8 (Spring, 1978) 4: 703-24; Vinovskis, “Angels Heads and Weeping Willows: Death in Early America,” Proceedings of the American Antiquarian Society 86 (1978) 2: 275-302; Michael R. Haines, “Estimated Life Tables for the United States, 1850-1910” Historical Methods 31 (Fall, 1998) 4: 149-169All that can be said is that a comparison of the census data on age structure for 1850 does not give any indication of very large differences in mortality between whites and people of color; but this is a very uncertain measure which could miss quite sizeable differences.

xxvi Medway v Natick, 1810, in Dudley Atkins Tyng, Reports of Cases argued and determined in the Supreme Judicial Court of the Commonwealth of Massachusetts … 1810-1811 (Newburyport, 1812) 88-89/


xxvii “An Act relating to Marriages between individuals of certain races,” Chapter 5, Acts and Resolves of the Legislature of Massachusetts in the Year 1843 (Boston, 1843).


xxviii Chickering, A Statistical View, 159.