Journal of Interdisc+linary History, xxv1:3 (Winter, 1996), 367-392.
Daniel Scott Smith and 1.David Hacker
Cultural Demography: New England Deaths and
the Puritan Perception of Risk Perception of risk has a
connection to actuality but does not mirror it. For example, recent
news reports would not lead travelers to think that they would
have to take a round-trip on a United States commuter airline
everyday for 5,000 years before dying in a fatal crash. This investigation examines how New Englanders of the seventeenth and
eighteenth centuries perceived the risks of death. The rich and
suggestive language of printed funeral sennons provides the main
body of source material. W e situate the ministerial discourse on
mortality risk in these documents in two theoretical contexts:
First, the discipline of demography provides a useful metric for
the notion of actuality, principally in its construct of the life
table-a
device that generates averages from data on death rates
by age and other characteristics. Second, a multidisciplinary literature on risk provides suggestive interpretive insights. In response to public policies and controversies about environmental
hazards, psychologists and other social scientists have explored the
relationship between actual and perceived risks.'
In recent years, cultural historians have made mortality a
major topic for study. Scholars of death in early New England
have featured, in particular, the role of Puritanism in enhancing
Daniel Scott Smith is Professor of History, University of Illinois, Chicago. H e is author of
"Female Householding in Late Eighteenth-Century America and the Problem of Poverty,"
Journal ofsocial History, XXVIII (1994), 83-107; "American Family and Demographic Patterns
and the Northwest European Model," Continuity and Charge, VIII (1993), 6 7 9 1 .
J. David Hacker is a graduate student in the Department of History at the University of
Minnesota. H e is the author (with Steven Kuggles and Matthew Sobek) of "Order out of
Chaos: General Design of the Integrated Public Use Microdata Series," Historical ~Vkthods,
XXvIII (19951, 33-39.
This article was first presented at the annual meeting of the Social Science Hlstory
Association, in Baltimore, Maryland, on November 4, 1993. The Institute for the Humanities
of the University of Illinois, Chicago, provided financial support for the research. In addition
to the participants in early American history seminars at the University of Minnesota and the
Newberry Library, the authors are indebted to Stephen Foster, Janles Lehning, Katherine
Lynch, Gigi Santow, and Susan Watkins for comments.
O 1995 by The Massachusetts Institute of Technology and the editors of T h e Journal qf
Interdisciplinary History.
I
Arnold Barnett supplied the figure for mortality risk o n commuter flights
368
1
DANIEL S C O T T S M I T H A N D J . DAVID H A C K E R
the terrors of death by promoting uncertainty of salvation. This
article seeks to bridge the gap between these cultural studies and
studies of mortality patterns by historical demographers. Central
to the origmal formulation of the theory of demographic transition
was the tenet that fertility decline resulted from a prior or concomitant decrease in mortality, as well as from economic modernization. The logic underlying the former relationship is simple:
If mortality decreases, fewer births are needed to attain a desired
number of surviving ~ h i l d r e n . ~
Decades of empirical research have left the fact of demographic transition in place; in many societies, low birth and death
rates now prevail where once they were high. However, the theory
of demographic transition has not fared as well, and the American
case provides a striking instance of apparent perversity in the
relationship between mortality and fertility indices. Although
child-to-woman ratios in the United States declined from the early
nineteenth century onward, mortality does not appear to have
decreased substantially until after 1880, and it may even have risen
during the first half of the nineteenth ~ e n t u r y . ~
One reason for the absence of a stronger relationship between
mortality and the decline of fertility is that the impression of
typical mortality risk, rather than its numerical incidence, informs
the behavior of people. In response to this possible cultural filter,
one demographer has formulated propositions relating perceptions of average mortality risk to fertility behavior. Put in another,
more historic context, we are interested in the relationship between perceptions of mortality and the larger mental framework
within which an emphasis on such concepts as "average level"
and "typical risk" of mortality attain meaning. Cognitive nurneracy
in mortality might be defined as a stress on the order provided
2
Lawrence Stone, "Death," in idem (ed.), The Past and the Present (Boston, 1981),242-259;
David E. Stannard, The Puritan Way qf Death: A Study in Religion, Culture, and Social Charge
( N e w Y o r k , 1977); David Hackett Fischer, Albion's Seed: Four British Folkways in America
( N e w Y o r k , 1989), I 11-1 16; Sanluel H . Preston, "Introduction," In idem (ed.), The Effects c f
Infant and Child Mortality on Fertility ( N e w Y o r k , 1978)~1-18.
3 Emphasizing regularities, Jean-Claude Chesnais provides a comprehensive review o f
national data o n the fertility/mortality relationship in the demographic transit~on-The
Demographic Transition: Stages, Puttems and Economic Implications. A Lor~xitudinalStudy c f SixtySeven Countries Coverir~gthe Period 1720-1984 (Oxford, 1992), I ~ O - - I S I , 346-354. Michael
Haines, "The Use o f Model Life Tables to Estimate Mortality for the United States in the
Late Nineteenth Century," Demography, XVI (1979),289--312.
CULTURAL DEMOGRAPHY
/
369
by the quantified magnitudes of death rates. The life table is a
perfect illustration of such an ideal/typical model of numeracy in
mortality.'
Problems ofMortality perception
In the traditional regime of
high (and highly variable) mortality, it is not all that surprising
that perception of the risk of death diverges from measured
realities. By showing the separate impact of fertility and mortality
levels on both period and cohort age patterns of death, model life
tables and stable populations provide a succinct overview of the
range of situations relevant to perceptions of death risk in populations. Figure I outlines, and Table I elaborates, the sources of
the difficulty in perception. They display indicators of the central
tendency and variation in mortality at three divergent expectations
of life at birth in model life tables: 25, 50, and 75 years, respectively. These values are shown for three rates of natural increaseo, 10, and 2 5 per 1,000 per annum-in stable populations.
When mortality is high, the impact of death is strongly
bimodal by age-very high in infancy, high in early childhood
(ages one through four), declining to a minimum at the beginning
of the teenage years, and then increasing gradually before finally
accelerating rapidly after middle age. Hence, few people die at
ages in the vicinity of the average age at death. With life expectancy at birth (eo) of twenty-five or fifty years, under one-sixth
die between ages fifteen and thirty-four and forty and fifty-nine,
respectively. However, with the low mortality shown in the third
column of Table I (eo = 75), more than half of the deaths occur
to those dying within ten years of the expectation of life at birth.
In a population closed to migration, the level of fertility
overwhelmingly determines the degree of youthfulness in the age
distribution. Since higher fertility makes the population younger,
the average age at death is lower in a rapidly growing population
than in a stationary one. The high rate of natural increase in early
American populations would have made inference especially
tricky. Of the demographic regmes portrayed in Table I, that of
colonial New England most closely approximates that shown in
4 David M . Heer, "Infant and C h ~ l dMortality and the Demand for Chldren," in Kodolfo
A. Bulatao and Ronald D. Lee (eds.), Deternrinants of Fertility in Developing Countrie.~( N e w
Y o r k , 1983), I , 369-387. See the analogous discussion o f "numeracy in children" in Etienne
van de Walle, "Fertility Transition, Conscious Choice and Nurneracy," Demo~raphy,XXIX
(19921, 487-502.
Fig.
Median Age at Death in Stable Populations by Life Expectancy at Birth
1
(P)
and Growth Rate
Panel C, e,
Panel B, e, = 50
Panel A, e, = 25
= 75
u
>
-z
h t-
80
5
i
1
([I
m
([I
401
c
30
m
0
-
:
ci
3:
9
50t
z
0
%
--Growth Rate
SOURCE
4
4
a,
cl
60
50 C
p
0
c
u
9
C
u
30
I
z
9 i
Growth Rate
0
0
Growth Rate
MODEI.WEST figures from Ansley J. Coale and Paul Demeny, Kegional Life Tables and Stable Populations (New York, 1983; zd ed.).
CULTURAL D E M O G R A P H Y
Tabk
I
/ 371
Indicators o f Central Tendency and Distribution o f Deaths for
Females at Different Expectations o f Life and Rates o f Natural
Increase
MEAN EXPECTATION OF LIFE
AT BIRTH
GROWTH RATE AhrD INDICATOR
(MEANAGE AT DEATH
BIRTH COHORT)
FOR
25
50
75
YEARS
YEARS
YEARS
(A)v = o (life-table stationary
population)
Birth rate per 1,000
Death rate per 1,000
Mean age at death in
population
Median acre
" at death
Percentage of deaths w i t h n
+/- 10 years of mean
Percentage of deaths in
modal 5-years
(Modal 5-years)
(B)v = .OI (?table population)
Birth rate per 1,000
Death rate per 1,000
Mean age at death in
population
Median age at death
Percentage of deaths within
+/- 10 vears of mean
Percentage of deaths in
modal 5-years
(Modal 5-years)
(C)Y= ,025 (stable population)
Birth rate per 1,000
Death rate per 1,000
Mean age at death in
population
Median age at death
Percentage of deaths w i t h n
+/- 10 years of mean
Percentage of deaths in
modal 5-years
(Modal 5-years)
SOURCE
Model WEST figures from Ansley J. Coale and Paul Demeny, Regio~aiModel
L$e Tables and Stable Populations (New York, 1983; 2d ed.).
372
/
DANIEL S C O T T S M I T H A N D J . DAVID H A C K E R
the middle colunln of the bottom panel (eo = 50, F = .o25). In
that stable population, the mean age at death for any gwen time
period is 26.5 years, and half of those dying are under age 12 (the
median being I I .2 years), even though the expectation of life at
birth (the average age at death) was far higher.5
There is, unfortunately, no reliable body of empirical evidence from seventeenth- and eighteenth-century New England
about the age patterns of death as experienced from year to year.
Historical demographers have not resolved the extent of underregistration of deaths, either overall or by age. The median age
of death in the first American life table-based on late eighteenthcentury bills of mortality from long-settled communities in which
fertility was lower and age structure older than average-was only
18.1 years. For most of New England throughout most of the
colonial period, the median of I I .z years is probably closer to the
mark.6
The findings of psychologists concerning the framing of risk
have important implications for the history of the perception of
mortality. For example, the assessment of risks, even by experts
today, lacks logical consistency. In one study, both lay persons and
physicians changed their preferred type of treatment for hypothetical cases of lung cancer when the outcomes were framed in
terms of substantively identical proportions surviving rather than
fractions dying.'
Psychometric studies indicate that the criteria underlying the
evaluation of risk may be grouped into two main classes. The
more important factor is the extent to which risk is feared because
it is uncontrollable, catastrophic, involuntary, etc. The second is
the element of concern that depends on the degree to which risk
is unknown because it is not directly observable, is novel, or has
delayed effect^.^
5 Ansley J. Coale, "How a Population Ages or Grows Younger," in Ronald Freedman (ed.), Population: T h e Vital Revolution (New York, 1964), 47-58. 6 Edward Wiggleswrorth, "A Table Shewing the Probability of the Duration, the Decre- ment, and the Expectation o f L ~ f ein the States of Massachusetts and New,-Hampshire, fornled from sixty-two Bills of Mortality on the files of the American Acadenly of Arts and Sciences in 1789," American Acadenly of Arts and Sciences, 12.lrmoirs (Boston, 1793), 11, part I, 133. 7 Barbara J. McNeil et al., "On the Elicitation of Preferences for Alternative Therapies," New Ergland Journal of :\ledicine, CCCVI (1982), 1259-1262. 8 Paul Slovic, "Perception of Risk: Reflections on the Psychometric Paradigm," in Sheldon Krinlsky and Dominic Golding (eds.), Social Theories cf Risk (Westport, 1992), 117-152. CULTURAL DEMOGRAPHY
/ 373
Both psychometric and cultural analyses of risk strongly indicate that there is no "attitude toward death" in the historical
record that can be isolated atomistically. Rather, we can expect
to find variable perceptions of mortality risk that depend on how
the hazard is embedded in situations and in the cultural norms
and expectations that people bring to it. In law, the definition of
homicide ignores the age of the victim; yet this neglect does not
contradict sentiments that the death of a ten-year-old is more
regrettable than that of an eighty-year-old. Motive and circumstances shape the definition of murder rather than the number of
potential years of life foreshortened. The approach of cultural
demography is to attempt to explore how a group that dealt with
death-in
this case, New England ministers of the seventeenth
and eighteenth century-defined its subject matter.'
T h e Fzrrze~ulSermorz us Sotlrce a~zdG e n ~ e Like other sources
for that much-studied group in that much-studied society, the
published funeral sermons of early New England ministers have
attracted the attention of historians and literal-): scholars. These
sermons followed the standard tripartite fonnat of text, explication
or development, and application. The final section often included
a discussion of the meaning of death from the ministers' perspective on the life and death of the saint. Expounding on a particular
death, a minister could admonish his parishioners to "get ye ready
also." Since the purpose of the sermons was didactic, the historian
learns mainly what ministers wanted their listeners to think about
death. However formulaic these passages may be, they were the
precepts of an established religious culture. Moreover, careful
readers can find, between the lines, hints of what ministers' listeners and readers believed, or wanted to believe, beyond what
was theologically correct."
9 For a relativistic position on cultural construction of risk, see Mary Douglas and Aaron
Wildavsky, Risk and Culture (Berkeley, 1982). Less extreme is Charles E. Kosenberg, "Introduction, Frarning Disease: Illness, Society, and History," in Rosenberg and Janet Golden
(eds.), Framing Diseuse: Studies in <7ultural History (New Brunswick, N.J., 19921, xii-mi.
10 Lonna Myers Maltnsheitner, "New England Funeral Sertnons and Changing Attitudes
Toward Woman, 1672-1792," unpub. Ph.D. diss. (Univ. of Minnesota, 1973); Harry S. Stout,
Tlze Neiv England Soul: Preuclziwg and Religious Culture in Colonial Neb, England (New York,
1986), 123, 161-162. Luke 1x40. In 1722, sixteen-year-old Bensatnin Franklin satirized the
elegy, calling it for "the greatest part, wretchedly Dull and Ridiculous" ("Silence Do-Good,
No. 7," in Leonard W. Labaree and Whitfield J. Bell, Jr. [eds.], T h e Papers qf Benjilrniu 1:railkliti
[New Haven, 19591, I, 23-26).
374
1
DANIEL S C O T T SMITH AND J . DAVID HACKER
New England funeral sern~ons-of which fewer than fifteen
were printed in the seventeenth century-numbered
c. 600 by
the year 1800. O f these, we sampled fifty in each of three time
periods delineated by the bibliographical volumes compiled by
Evans. Since some ministers were more prolific than others, our
study includes the works of I I O separate authors.''
Funeral sermons were not intended to provide reflections on
a random sample of deaths or lives in seventeenth- and eighteenth-century New England. Despite their high death rates, no
infants or very young children were memorialized in sermons.
Nearly 70 percent of the sermons were for men, and the average
age of death of those so honored was about 55. Among the men,
ministers (43 percent) and political leaders or merchants (43 percent) were the principal subjects. Some fourteen of the forty-seven
women were ministers' wives."
The social or pious status of 12 of the 150 subjects is not clear
from the sermons themselves. Doubt about status applies especially
to those too young at death to have attained significant social
rank-cases of interest because of the demographic questions. For
example, the status of Elisha Lyon-a Woodstock, Connecticut,
youth killed by a musket accident in 1768-and Michael Griswold, Jr., of North Illingsworth, Connecticut-"suddenly kill'd,
by a stick of Timber which fell on him"-is
not clear. Possibly,
the unusual circumstances of an obscure person's death led to the
In his bibliography of 39,161 surviving items published in all of mainland British America
II
and the United States before 1801, Charles Evans (American Biblio~raphy[Chicago, 1903-19 j j],
14v.) lists 789 funeral sermons, half of which appeared after 1785. O f the 19,448 items
published before 1786, 501 were funeral sermons. Using the sequence numbers provided by
Evans as lndlcators of the dates of publication, the first jo relevant cases were taken from
vol. I, which covers 1639-1729, vol. IV (1765-17731, and vol. XI11 (1799-1800). However,
since the final volume lacked a suff~cientnumber of sermons, we also drew 18 cases from
the end of vol. XI1 (1798-1799). W e excluded sermons with places of publication outside of
New England and those that were originally preached elsewhere but reprinted in that region.
W e also rejected execution sermons, broadsides announcing the death of an eminent person,
and the numerous short eulogies given after the death of George Washington in 1799
Wondrously productive, Cotton Mather published 67 of the sernlons on those who
perished from a total of 76 for New England between 1721 and 1730; he authored nearly
half (23) of the sermons in our first period, which concludes in 1716 (Gordon E. Geddes,
Welcomejoy: Death in Puritan hreu~E n ~ l a n d[Ann Arbor, 19811, 161). Mather's most frequent
dream, w h ~ c hoccurred from early in his life, was of his own death (Robert Middlekauif,
T h e :'vlathers: Three Generations cf Puritan Intellectuals, 1596-1725 [New York, 19711, 320).
12
In her more comprehensive survey ("New England Funeral Sermons," 32) Malnlsheimer
estimates that 75% of the sermons were on men, 25% on women, and 5% on chddren (those
not financially independent of parents). Perhaps as many as 65% of those on Inen were on
ministers, and half of those on women were on the wives and daughters of ministers.
CULTURAL DEMOGRAPHY
/
375
publication of a funeral sermon. Supporting this inference is the
existence of the related, and considerably more popular, genre of
the execution sermon. Just as the lives and deaths of saints could
be advanced as role models, those of notorious sinners served even
more compellingly as negative examples, especially as the minister
traced the horrendous sin meriting capital punishment back to
minor transgressions committed earlier in life.13
Puritan ministerial thought
about the risks of mortality may be summarized briefly in two
statements: (I) Death is certain. ( 2 ) Its timing is uncertain. These
basic principles encompassed or generated other themes: that life
was bounded by the upper limit of a fixed life span, and was, in
any case, short; that the capricious timing of death made preparation for it necessary at all stages of life, even though ministers
and their audiences discerned simultaneously that the average risk
of death varied by age; and that variance of all sorts in the
phenomenon of death was more impressive than what might be
considered average or typical.
From their reading of the Garden of Eden story in the Old
Testament book of Genesis, orthodox Puritans believed that both
disease and death were punishments for the orignal sin of Adam.
The first clause of I Corinthians I 5 : 2 2 made the point succinctly:
"Allmankind is involved in Adams7 sin, and therefore all must
dy." With the corpse as proof, the funeral sermon was the ideal
occasion to stress the theme of death7scertainty. Though supportive of hierarchy on this earth, ministers used the death of the
high and mighty to portray death as the great leveler. "All,high
and low, rich and poor," one minister proclaimed at the funeral
of a Connecticut governor, "are alike subject, to the common
laws of mortality. It is appointed for all once to diesn1'
THE THEMES OF MORTALITY RISK
13 Abel Stiles, Death God's monitor to the Living (Providence, 1768), 2; William Seward, Due Consideratiow of; awd Preparation for our latter End, or Death, Recotntnended and C'vged: In T w o Semons (New Haven: 1772); Malnlsheimer, "New England Funeral Sermons," 33-34; Karen Halttunen, "Early American Murder Narratives: The Birth of Horror," in Richard Wightman Fox and T. J. Jackson Lean (eds.), T h e Potver of Culture: Critical Essays in Americaw History (Chicago, 1993). 67-101; Daniel A. Cohen, Pillars of Salt, ~\/lonuments of Grace: New England Critne Literature and the Ovtgins of Atnerican Popular Culture, 1764-1860 (New York, 1993). 14 Eliphalet Williams, T h r Ruler4 Duty and Honor, In Serving His Generation (Hartford, 1770)~ 21; see also Cotton Mather, T h e Glorious Throne (Boston, 1714), 30-31; John Mellen, T h e ~\/leritof doing good (Boston, 1765). 21. 376
1
DANIEL SCOTT SMITH AND J. DAVID HACKER
As with the principle of death's certainty, the Bible was also
the authority for the corollary that life was short. Ministers repeatedly invoked passages depicting the brevity of life. References
to life as a "vapour," a "shadow," as "altogether vanity" and as
"withering grass and flowers" were common throughout the period. Compared with biblical instances of extreme longevitysuch as that of the antediluvian Methuselah, who was said to have
lived 969 years-as well as to present experience, life was short.
Although young people perceived a life of many years as long in
prospect, the old knew just how rapidly the years flew by. "Now
ask any man that may be Seventy years old; my Father, How long
does the Time of your past L f e now seem to you? He will say,
'Indeed, when I look'd forward, it seem'd long; but now I look
backward, Oh! how swiftly is it passed; how quickly ended!"'15
The concepts of Ife span
T h e L f e Span and Stages of L f e
and stages of l f e provided the framework that nearly forced a
recognition that mortality risks varied with age. The biblical
"three-score and ten years" was the normal life span if a man was
"to live out all his Days." For someone of unusually uncommon
strength, as the Psalmist first claimed, the life span could be
extended to eighty years of age. Fixing the maximum length of
life in this range implicitly demonstrates an awareness that death
rates at these older ages were much higher than at younger adult
ages. In seventeenth-century England, old age was regarded as the
final stage of the life span-a phase certainly to be terminated by
death. Ministers in New England concurred with this assessment:
"Elder People are so near their E n d , so soon to dye, and to give up
their Account, that surely they above any shou'd be very careful
about their Behaviour. "I6
15 The biblical texts are, respectively, James 4:14, I Chronicles 29:15, Psalms 39:4-5, Job
I ~ : I - 2 , and Psalms 9o:j-6; Cotton Mather, L$e S~tu$ly Passing awd Quickly Endiwg (Boston
17161, 7.
16 Psalm y o : ~ o ;M ather, L$e S~tu$ly Passing, 7; Charles Chauncy, A Discourse Occasioned by
the Death o f t h e Reverend Thotnas Foxcroft (Boston, 1796), 19-20; Mather, L$e Stviftly Passing,
7; Jatnes Dana, Illustratiows ow Hurnaw Frailty awd Vawity (New Haven, 1765)~5; Cha~mcy,
Discourse Occasioned, 18. For a cohort, the expectation of life at a given age is identical to the
reciprocal of the death rate after that age. Steven R. Smith, "Death, Dying, and the Elderly
in Seventeenth-Centuty England," in Stuart F. Spicker, Kathleen M . Woodward, and David
D. Van Tassel (eds.), Aging and the Elderly: Hutnanistic Perspectives in Gerontology (Atlantic
Highlands, N.J., 1y78), 21 I; Benjamin Coln~an,T h e D u t y and Honour o f A g e d W o m e n (Boston,
171 I), 3 I; Chauncy, Discourse Occasioned, 18
CULTURAL DEMOGRAPHY
1
377
Drawing on both folk and Enlightenment culture, ministerial
writings on old age underlined the point. According to a common
saying, "the young may, the old must die." By the end of the
eighteenth century, the idea was being couched in statistical terms:
"There is a probability that they who are in the prime or morning
of their days may continue many years; but there is no probability
that the aged should." Hence, death in old age was not as regrettable as death in youth.17
In their sermons, ministers clearly showed that they understood the differing risks of mortality at each stage of life. Using
the death of Elizabeth Wainwright-the youngest decedent in our
study-to survey the successive stages of life, Reverend Benjamin
Colman referred to "brittle and tender" babes, "blooming" childhood, "verdant, budding7'youth, and "the prime of strength" of
manhood. Elsewhere, youth is described as "hale, strong, and
vigorous, " and as a stage of "vigour, " "blooming vigour, " or "vigor
and strength." Nevertheless, ministers also argued that vigor and
health were no security against early or sudden death, for as
Wadsworth insisted in 1715 , they "are often almost as Changible
as the Weather."18
Ministers could still emphasize the cumulative impact of "high"
stage-of-life-specific death rates. As a well-read scientist, Cotton
Mather sometimes made his points quantitatively. In two separate
sermons, Mather drew upon the work of Graunt, the English
political arithmetician, to assert that "computation [has] found that
more than half of the children of men Dy before they come to
be Seventeen Years of Age." In his sermon on the schoolmaster
Ezehel Cheever, who died at age ninety-four, Mather warned,
17 Ebenezer Gay, T h e Old Man's Calendar. A Discourse on Joshua X I K i o Delivered in the First
Parish ofHingham on the Lord's Day, August 26, 1781 (Boston, 1781; repr. Hmgham, 1846), 22;
Job Orten, Discourses to the Aged (New York, 1980; repr. Salem, 1801; orig pub. 1771), 29-30;
Jonathan Parsons, A Funeral Sermon Occasiowed by the Death ofMr. Ebenezer Little (Salem, 1768),
14.
18 Colman, A Devout Cowtemplation on the Meaning ofDiviwe Providewce iw the Early Death of
Pious and Lovely Children (Boston, 1714)~5 ; Benjamin Wadsworth, Death is Certain, the Time
When Uwcertain (Boston, I ~ I O )9;, Ebenezer Baldwin, A Funeral Oration iw Memory of Mr.
Jowathaw Lyman (New Haven, 1767), 14; Thomas Wells Bray, The Knowledge of Our Ewd, with
the Advantage awd Importance of it (New Haven, 1740)~9; William Mason, A Sermon Delivered
at Castine (Castine, Me., 1799), 10; Wadswoah, Man's Present State Compared to Withering
Grass awd Flowers (Boston, 1715), 10; Baldwin, Funeral Oration, 14; John Foot, A Discourse
Delivered Jaldwuary 8th, 1769 (New Haven, 1769), 27; William Wells, A Sermon, Preached at
Brattleborough (Brattleborough, Vt., 1798), 17; Mason, Sertnon Delivered at Castine, 10.
"Children, 'tis your Dawning time. It may be your Dying time.
. . . Go unto the Burying-place; There you will see many a Grave
shorter than yourselves. "19
The high cumulative incidence of death since birth depended
on the "multitudes" who died in infancy and early childhooddistinct stages of life. Children hearing Mather's words would have
already survived through these most dangerous ages. If they or
young adults chose to recognize the better chances of current
survival, they might well have ignored admonitions to "Dy daily"
in preparation for death that in the long run was inevitable.
Alternatively, as the example of a nineteenth-century diarist,
Samuel Rodman, suggests, they may have felt guilty or fearful
that they survived longer than others did.20
Despite the normal life span and variations in risk with age,
ministers relentlessly contended that God determined the timing
of death: "The lives of all men are in God's hands: He lengthens
them out, or cuts them shoa, as he pleases. . . ." In his poetic
and popular rendering of Calvinist theology, The Day of Doom,
which was reprinted ten times before 1800, Wigglesworth directly
addressed the sense of false security from death in the present and
near future:
How many thousands have this strong delusion
Already brought to ruin and confusion
................................ They thought of many years, as thou dost now,
But were deceived quite, and so may'st thou."
To refute this naive, arrogant presumption, some ministers
came close to contending that all, regardless of age, faced similar,
I 9 Cotton Mather, Corderins Americanus (Boston, 1708), 17; Mather, L$e Stvi$ly Passiwg, 7.
John Graunt, Natural and Political Obsewatiows Mentiowed in a follotviwg Iwdex, awd made upow
the Bills of Mortality (New York, 197s; orig. pub. 1662), 69. Since the bills of nlortality did
not record age, Graunt infen-ed mortality before age 6 from information on cause of death.
Ian Hacking, T h e Etneugence of Probability: A Plzilosophical Study of Early Ideas about Probabilit):
Induction and Statistical Inference (Cambridge, 1984), 107-110; Mather, Corderius Atnericanus,
'/.
Maris A. Vinovslus, "Angels' Heads and Weeping Willows: Death in Early America," in
idem (ed.), Studies in American Historical Demography (New York, 1979), 196-198.
21 Joseph Pe~ry,T h e Character of Moses Illrrstrated and Itt~proved(Hartford, 1767). 16; Michael
Wigglesmorth (ed. Kenneth B. Murdock), T h e D a y of Doom: O n a Poetical Descriptiow o f t h e
Great and Last Judgment tvith other poems (New York, 1966; orig. pub. 16621, 81.
20
CULTURAL DEMOGRAPHY
1
379
high death rates in the short run. This sentiment was convenient
computationally for Graunt, who assumed a constant death rate
per decade of life (about three-eighths) between the ages of six
and seventy-six. Citing what he called an old saying, Mather
remarked that "lfan old Man has Death before his Face, a Young Man
has Death behind his Back; the Deadly Blow may be as near to one,
as it is t70ther."In a Boston sermon of 1713, an Anglican assistant
minister of King's Chapel noted that those "in the midst of their
Youthful follies . . . may be soon snatched away as the Hoary
Head, which bends to the Grave. Yet, the conditional "may be"
in these two passages suggests what was possible rather than what
was likely. At least one sermon, from I 771, contains a recognition
that people, foolish as they were, felt secure from imminent death.
Referring to experience, Bisset found it strange and incredible
"that mankind, who in the midst of life are in death, and who
are daily surrounded by a thousand diseases and accidents, should
stand in need of so many exhortations, in order to awaken them
from the sleep of sin into seriousness and a t t e n t i ~ n . " ~ ~
The more common response eschewed the inchoate framework within which notions of average or relative risks flickered
and occasionally grew stronger. The traditional view of each life
as an individual pilgrimage terminated by death was a powerful
one within the Christian-especially the Protestant-tradition.
Instead of perceiving the individual as one among many with a
common mortality risk, ministers focused their funeral sermons
on the individual journeying alone through life. For example, a
twenty-year-old man reasoning on that basis would not conclude
that he had 97 chances out of IOO to survive to age twenty-five
but that after five years, his "chances" of death had been,
retrospectively and definitively, either zero or one.23
Variance in death was important for reasons beyond its divinely arbitrary timing. Indeed, the hand of God was revealed
more in the unusual than in the ordnary, particularly in the
earliest period of intense Puritanism. Accidents and other unusual
22
Hacking, The Etnevgence of Probability, 108-109; Mather, L$e Sw@ly Passing, 16; Henry Harris, A Sertnow Preach'd at the Queen's Chapel (Boston, 1713). 16; George Bisset, A Sermon Preached in Trinity Church (Newport. R.I., 1771). 5 . 23 Thonlas R. Cole, The Journey o f l f e : A Cultural History ofAginy in America (New York, ~ y y z )10-47
,
Some 969 out of 1,000 20-year-olds would survive the next five years, according
to Model WEST, Level 13 mortality. See Table I .
causes of death were subsumed under the category of "Remarkable Providences." "That God is pleased to call off some by a
sudden and unexpected death," argued Bisset on the death of
Abigail Wanton, age thirty-six, serves "to awaken us from this
fatal sleep of security, and to teach us the awful importance of
improving our time." Rev. William Steward of New Haven even
claimed that sudden death was common, for "it is the portion of
great numbers of mankind to be snatch'd away suddenly by
death. "24
What made the uncertain timing of death especially important religiously was the need for individuals to be prepared to face
judgment. Reflecting on the death of Elisha Lyon, approximately
twenty years of age, who was killed by the accidental discharge
of a musket while on a military exercise in 1768, the Rev. Stiles
of Woodstock, Connecticut, stated the point and its implication
precisely: "It is not for us to fore know the time, the occasion,
the manner of our Death; but surely it is for us to know the
necessity of being prepared. "25
Sudden deaths to younger persons provided doubly ideal
opportunities to emphasize the need for preparation now. Altogether, there were twenty-two sermons in which the decedent
was under thirty years of age. In all of these but one or two, a
main theme in the discussion was the uncertainty of life and, as
a consequence, the need to prepare for death immediately. Although this theme also appeared in sermons for those dying at
ages above thirty, its ubiquity in the texts for those who died
young indicates that ministers implicitly recognized the biblical
life span of seventy years and the surprising feature of early death.26
n/laternal Mortality, Actual and Perceived
During the last
three centuries, political arithmeticians, demographers, and historians have pointed out how the mortality risk of childbearing has
been exaggerated. In the index of contents to his Obsewations
24 Franklin caustically advised, "For the Subject of your Elegy. Take one of your Neighbours who has lately departed this Life; it is no great matter at what Age the party dy'd, but
it will be best if went away suddenly, being Kill'd, drown'd, or Froze to Death" ("Silence
Do-Good, No. 7," 26); Bisset, Serrnon Preached, 8; Seward, D u e Conrideration, 14.
Stiles, Death God's Monitor, 27.
25
26 Patricia Cline Cohen argues that the Puritans did "not share our notion that a young's
person death is untimely or premature," hut explains in a footnote that they dld not view
such early deaths as unnecessary o r preventable ( A Calculating People: T h e Spread ofhrumeracy
in Early America [Chicago, 19821, 93, 241).
CULTURAL DEMOGRAPHY
/
381
( I 662), Graunt implicitly emphasized its rarity, demonstrating
from bills of mortality that "not one woman in a hundred dies in
Child-bed." In the text, taking note of competing causes of death,
Graunt reduced his estimate of the risk of dying a month after
birth to approxinlately one in two hundred. Relying on modem
family reconstitution data, Schofield calculated estimates for English parishes that are roughly similar to those of Graunt. Echoing
his predecessor's skepticism about the threat of deaths from causes
related to childbirth, he concludes that "Childbearing in 'the
world we have lost' tums out to have been a rather less mortal
occasion than we may have been inclined to believe." In the
practice of Martha Ballard, a Maine midwife, in the period between 1785 and 1812, only 0.6 percent of mothers died per
confinement. Ballard, however, routinely referred to delivery as
a "sickness" and thanked God for its successful outcon~e."
Only two of the forty-eight women whose funeral sermons
appear in our sample died of maternal causes-a ratio, coincidentally, of the same order of magnitude as the cumulative figure
based on the empirical observations of maternal mortality. Ministers connected their deaths in childbirth to "the Cuvse pronounc'd
upon ouv$vst iWothev Eve" from Genesis 2:16. They also referred
to the risks of maternal mortality in other sermons. The extended
duration of pregnancy before the perilous event of travail served
more effectively than shorter episodes of sickness ("which do more
surprise G ternt fov the time, but are usually soon over and theiv Impressions vanish") to bring women to a seriously pious state. Increase Mather quoted from the private papers of Maria-Cotton's
mother-"I went thvo' many Deaths in the bearing of~hildven."~~
27 Graunt, Natural and Political Observations, 10, 42-43; Roger Schofield, "Did the Mothers Really Die? Three Centuries of Maternal Mortality in 'The World W e Have Lost,"' in Lloyd Bonfield, Richard Smith, and Keith Wrightson (eds.), T h e World W e Have Gained: Histories of Populatiort and Social Structure (Oxford, 1986), 260; Laurel Thatcher Ulrich, "'The Living Mother of a Living Child': Midwifery and Mortality in Post-Revolutionary New England," William and Mary Qtrarterly, XLVI (1989), 27-48. The percentage of women who eventually died &om maternal causes must be multiplied by the total number of chddren born to them-including
those women who died before the end of childbearing years. 28 Cotton Mather, Eureka the Vertuons W o m a n Found (Boston, 1704); idem, Memorials of Early Piety (Boston, 1711); Colman, T h e D n t y and Honorrr o f A g e d W o m e n (Boston, 171I), ii; Cotton Mather, A n Essay to Describe the Good Works of A Vertrrorrs W o m a n (Boston, 1713); Colman, D n t y and Honour of Aged W o m e n ; Increase Mather, A Sermon Concerning Obedience and Resignation T o the W i l l of God (Boston, 1714), iv. 382
/
DANIEL S C O T T S M I T H A N D J. DAVID H A C K E R
Can the discrepancy between the facts of maternal mortality
and perception of the risk of death in childbirth be reconciled?
Historians a r p e that women in early modern Anglo-America
sided with the ministers rather than with the political arithmeticians and later demographers. The apparent emphasis placed on
the dangers of maternal mortality seems to contradict a central
finding of the risk-perception literature. Unlike accidents, maternal deaths were not sudden-being anticipated during the long
months of pregnancy; the proximate cause of mortality risk was
not unknown, novel, or unobservable. Moreover, unlike in epidemics, deaths in or after childbirth occurred singularly rather than
in large numbers within a short period.29
Psychologists emphasize that strong prior belief makes it
difficult for subsequent information to alter the perception of risk.
The apparent disjunction between the incidence and perception
of maternal mortality is, at present, the most substantial evidence
that religion purveyed, or at least reinforced, among ordinary
people an interpretive framework that overestimated the risks of
death. Given the authority provided by the first book in the Bible,
the difficulties of pregnancy and childbirth could be ascribed
clearly to acts of God.30
As in the case of the simultaneous recognition and denial that
mortality varied with age, we should contemplate the possibility
that women both believed and doubted these messages of danger
concerning childbirth. The paradigm of sickness and providential
assistance that midwife Ballard repeated after each delivery was
indeed formulaic. Its use does not necessarily indicate that she or
the mother were greatly in fear of death. Similarly, the closeness
of death was not the only theme of childbirth available to ministers. Mather incorporated an earlier sermon about pregnancy and
labor into his unpublished treatise on medicine. Although
Mather's chapter-the
second longest of some sixty-six sections
in the manuscript-focuses mainly on religious matters, it also
provides nearly three pages of remedies at the end to ease various
29 Linda A. Pollock, "Embarking on Rough Passage: The Experience of Pregnancy in Early-Modern Society," in Valerie Fildes (ed.), Women as Mothers in Pve-Indrrstvial England: Essays in Memory of Dorothy McLaren (London, ~ggo),47-49; Catherine M. Scholten, Child-
bearing in American Society: 1650-1850 (New York, 1985), 19-22. 30 Slovic, Baruch Fischoff, and Sarah Lichtenstein, "Behavioral Decision Theory Perspec- tives o n Risk and Safety," Acta Psychologica, LVI (1984, 183-203. CULTURAL DEMOGRAPHY
1
383
problems associated with labor. The activist/fatalist and instmmental/fatalist labels that Fischer uses to summarize Puritan theological practice with respect to death also are relevant to Puritan
social p r a ~ t i c e . ~ '
The Tangle of Folk, Religious, and Scientijic Views
Religon
was not the only framework for dealing with disease and death.
Alternative traditional perspectives-both
folk and medicalcompeted with the outlook offered by religon, but, in practice,
alternative views coexisted with, complemented, and intertwined
with the Puritan perspective. New England ministers served as
physicians or collaborated with them in particular cases. Deatheither of the individual interpreted in funeral sermons, or of many,
the result of epidemics and other mass calamities-was amenable
to the religous idea of providential retaliation against individual
or collective sin. Fatalism was a plausible response to death after
it occurred, and ministers made the most of it. Before death,
behavior was not necessarily unrealistic or fatalistic. Patients
tended to avoid treatment except for serious illness, and their
acceptance of such heroic medical interventions as bloodletting
suggests a will to survive, even though God supposedly had
already determined the time of death.3z
Nor did a conflict emerge between religion and the worldview provided by the Newtonian science of the Enlightenment.
Ministers could adapt the new science to their traditional religious
assertions without apparent contradiction, or, when necessary,
elide it metaphorically. Ever eclectic, Mather, in his medical
treatise, invoked a notion of religious quality time: "Three Sevens
of years" of an awakened good life is longer than "Nine hundred
and sixty nine" years for a "drowsy and thoughtless wretch."
Living in the constant expectation of death permits one to "Live
Long in a Little Time." Nonetheless, Mather argued for a tem-
3 I Gordon W . Jones (ed.), T h e Argel of Bethesda (Barre, Mass., 1972), 235-248; Fischer,
Albion's Fatal Seed, 813.
32 See the balanced survey by Andrew Wear, "Puritan Perceptions of Illness in Seventeenth-Century England," in R o y Porter (ed.), Patients and Practitioners: Lay Perceptions of
Medicine in Pre-industrial Society (Cambridge, 1985), 55-100; Patricia A. Watson, T h e Angelical
CorGur~ction:T h e Preacher-Physicians of Colonial New England (Knoxville, 1991), 7-35; David
D. Hall, Worlds of Wonder; Days of Judgment: Popular Religious Belief in Early New England
(Cambridge, 1990)~196-210; Edward Shorter, Doctors and Their Patients: A Social History (New
Brunswick, NJ., 1991), 55-74.
384
I
DANIEL S C O T T SMITH AND J . DAVID HACKER
perate approach in diet and for moderate exercise as strategies to
extend longevity measured in years rather than metaphors.33
Mather's use of Graunt's figures to illustrate the shortness of
life was ironically appropriate, for Graunt based his figures on a
fixed mortality ratio during the seven decades of life after age six.
So great was the emphasis on the orderly character of mortality
among eighteenth-century statisticians that they often assumedradically in conflict with what is now well known-fixed geographical death rates and nearly constant death rates over the life
span. These quantitative assumptions, which ironically concur
with the central precept in the funeral sermons, namely, that death
is equally possible at each age, had little impact at the time, even
in their direct application to the insurance business.34
There was an imVariance within Families and over Time
portant difference between eighteenth-century religon and empirically oriented science with respect to regularities, especially the
problem of an orderly relationship between age and mortality risk.
For the mechanistic mind of the Enlightenment, order was found
in the regularity in nature, but for the religious nlind, order was
to be revealed in the arbitrary workings of providence. Empirical
investigations could serve to sustain the belief that order lay hidden
in disorder. One minister gathered facts on an epidemic for just
this purpose. In assessing the significance of the high childhood
mortality of the throat distemper epidemic in New Hampshire in
1735, Fitch concluded that the "Grave is a Land of Darkness
without any Order" with respect to age; "so many Younger ones
have gone to the long Home before the Elder." Divine order still
persisted as each person died "in his own Order, in respect to God's
appointment who has deterrnin'd the Time of every ones Death."
At the very end of the eighteenth century, another minister
reiterated this exact point: "We shall fall before him [Death]
indiscriminately, and go down to the grave without any apparent
order. "35
33 Jones, Angel of Bethesda, 14-19, O n Cotton Mather's use of Newtonian science, see Middlekauff, T h e Mathers, 27cp304. 34 Lorraine Daston, Classical Probability in the Enlightenment (Princeton, 1988), 126-138. 35 Jabez Fitch, Atz Account of the Numbers that have died ofthe Dirtemper in the Throat, Within the Province of New Hampshire, With rome Rejlections thereon, July 26, 1736 (Boston, 1736), 7; reprinted in Rosenberg (ed.), Disease and Society in Provincial Marsachuretts: Collected Accountr, 1736-1939 (New York, 1972). See also the general discussion in Cohen, Calculating People, 81-108; Eliphalet Porter, A Sermon Delivered to the First Religious Society (Boston, 1799), 9. CULTURAL DEMOGRAPHY
1
385
The emphasis on the uncertain timing of death-one of the
central themes in the sermons-appears in a variety of contexts
about death and the risks thereof throughout
the seventeenth and
eighteenth centuries. For example, mortality varied within families. In his analysis of the New Hampshire epidemic, Fitch not
only tabulated deaths by age but also reported, for each town, the
number of children lost per family and the number who had lost
all of their children. In a sermon after the death of Rev. Samuel
Checkley in 1769, Bowen noted that Checkley's children were
almost all "cut down in the flower and prime of life. . . . So that
out of twelve, there is but one survives him." Other sermons
focused on clusterings of deaths within families.36
Variation in death rates due to disease provided the most
obvious environmental support for the emphasis seen on the
uncertainty of life in funeral sermons. Although historians have
studied outbreaks of major epidemics-primarily smallpox-they
have barely investigated such vital fluctuations in early New
England. In Boston-the residence of many sermonizers-estimated crude death rates, which averaged 37.2 per 1,000 between
1701 and 1774, exceeded 50 per 1,000 only in 1702, 1721, 1730,
and 1752. During the smallpox epidemic of 1721, the death rate
reached a peak of 103 per 1,000. Although epidemics visited the
city in the nineteenth century, never did the crude death rate after
I Soo exceed 40 per I ,000.~'
36 Fitch, Account of the Numbers, 1-7. Another ministerial study reported the differential
impact within families. John Brown, T h e Number of Deaths i n Haverhill, and Also Some
Cornfortable Instances Thereof Among the Children, under the Late Distemper in the Throat, with a n
Address to the Bereaved (Boston, 1738; zd ed.), cited in Cohen, Calculating People, 9 0 3 1 . What
Fitch and Brown failed to report is the distribution by number of children in the families
that lost no children in the epidemic; i.e., they offer no demographic concept of population
at risk; Penuel Bowen, A Discourse Occasioned by the Death oJ the Reverend Santuel Checkley
(Boston, 1770), 3 I. See William Lockwood, A Sernlon Delivered at the Funeral o f Mrs. Jerusha
Woodbridge (Middletown, 1799), 18; Isaiah Parker, A Ftlneral Discourse Occasioned by the Death
qf 1VIr. Josiah Bowles (Boston, I ~ O O )17;
, Edward Don-, A Discourse Occasioned by the 11fuciz
Lanlented Death, o f the Honorable Daniel Edu)ards, Esq (Hartford, 1765), 1 9 2 0 ; Parsons, T o
Live is Christ, T o D i e is G a i n (Portsn~outh,1770), 38; Cotton Mather, h'epenthes Evangelicunt
(Boston, 1713), 42; Increase Mather, A Sernlon Concevnirg Obedience (Boston, 1714), iv; Samuel
Moody, T h e Children o f t h e Covenant (Boston, 1716), 52.
37 John Du&, Epidemics i n Colonial Anlen'ca (Baton Rouge, 1953). Deficiencies in sources
or organization are rrlainly responsible for this gap in the literature. Nearly all of the published
volurrles of New England vital records are organized by surname rather than year; furthermore, the corrlpleteness of the records declines between the seventeenth and midnineteenth
centuries. John B. Blake, Public Health in the ? b ~ v n of'Boston, 1630-1822 (Cambridge, Mass.,
1959), 247-249.
386
1
D A N I E L S C O T T S M I T H A N D J . DAVID H A C K E R
Compared with elsewhere in the Anglo-American world,
eighteenth-century rural New England Gas distinctly characterized by a low level of mortality punctuated with episodes of
severe epidemics that spread across a wide area. Not surprisingly,
ministerial discourses on epidemics (which we examined separately from the I 50 funeral sermons) took advantage of the surge
in deaths to emphasize the uncertainty of life.3"
From the begnning, collective events or phenomena-epidemics, droughts, Indian attacks, faltering piety, and the annual
spring planting season-rather
than the death of an individual,
stimulated authorities to declare days of fast, prayer, and humiliation. Calamities and natural disasters allowed ministers to emphasize general themes of divine punishment for sin and the need
for religious revival. The greater complexity of an epidemic compared with an indvidual death also elicited more diverse and
elaborate responses from the ministers. Both the smallpox epidemic of I 72I and the diphtheria epidemic of I 73 5-1 740 led them
to statistical investigations. In addition to tabulating deaths by age,
John Brown, following the model of James Janeway's popular
T o k e n for Children (London, 1671)~presented thirty-four short
biographical sketches of youthful victims, including deathbed
scenes and evidences of salvation. Although epidemics were unusual, their striking devastation merely exemplified, in magnified
form, the familiar uncertainty of life. In a sermon for four young
men who died of smallpox, Grafton-a Baptist preacher in Newtown-argued that these cases were not uncommon, for many
youth also perished from consumption and from accidents.39
FROM PURITAN TO MODERN LIFE TABLES
Metaphorically and
statistically, the dominant ministerial tradition in early New England focused more on the variance than on the average death rate.
3 8 Mary J. Dobson, "Mortality Gradients and Disease Exchanges: Comparisons &om Old
England and Colonial America," Social History ofMedicine, I1 (1989), 281-284. See also Ernest
Caulfield, "The Pursuit of a Pestilence," Proceedings o f the American Antiquarian Society, LX
(1950), 21-51.
39 Ronald A. Bosco, "Introduction," in idem (ed.), Tize Puritan Sermon in America: Illumiliation and Execution Sermons (Delmar, N.Y., 1978), I, mi-xxxxvii. According to Charles
Shively, drought was the most common reason for the proclanlation of a fast (Shively, "A
Histor). of the Conception of Death in America, 1650-1860," unpub. Ph.D. diss. [Harvard
Univ., 19681, 5). Dennis D. Melchert, "Experimenting on the Neighbors: Inoculation of
Smallpox in Boston in the Context of Eighteenth Century Medicine," unpub. Ph.D. diss.,
CULTURAL DEMOGRAPHY
1
387
The successful propagation and general acceptance of Enlightenment and deistic notions in the eighteenth century could well
have depended on the reduction of s;dden and unexpected deaths
due to epidemics. In Boston, as in Western Europe from the late
seventeenth century onward, the diminution of mortality peaks
associated with epibemics was central to the early phase i f the
mortality transition, though sensitivity to fearful upsurges in epidemic diseases seems to have persisted even after the frequency
and magnitude of such crise; lessened in the eighteenih and
nineteenth cent~ries.~'
Unlike other students of the cultural history of death in New
England, we find relatively little change over time in the perception of mortality risks. N o doubt, the unchangng circumstances
of death that prompted a funeral sermon account substantially for
the absence of any dramatic change. In addition, the emergence
of the funeral sermon in late seventeenth-century New England
was itself a mark of religious declension. Early Puritan divines
eschewed the preaching of the funeral sermon, fearing its deterioration into false elegy and consolation; no funeral sermons were
published before 1672. Changes were more subtle than dramatic
between the first and last samples of our study. The increasing
attention paid to aspects of the-saints' lives and.to the manner of
their deaths suggests a waning emphasis on warnings to the living
death became more
based on death. Discussion of the good
extensive in the funeral sermons during our second and final
(Univ. of Iowa, 1973); Caulfield, "A True History of the Ternble Epidemic Vulgarly Called
the Throat Dlsterrlper which Occurred in his Majesty's New England Colonies between the
Years 1735 and 1740,'' in Rosenberg (ed.), Disease and Society in Provincial Massachrrsetts:
Collected Accorrnts (New York, 1972), 5-1 13; Brown, T h e Number $Deaths in Haverhill; Joseph
Grafton, A Sernlon, Occasioned by the Death of Samuel Bixhy, W z o Died Sept. zg, aet. 17; Jonathan
Shepard, W h o Died Sepr. 28, net. 29;James Ward, W h o Died Sepr. 29, aet. 25; Michael BtQht,
Jun., who died October 1 0 , aet z o (All of'the small pox) (Boston, 1793), 8.
40 For the range of lsease and death phenomena, see the study of Rev. Ebenezer Parkman's
diary by Rose Lockwood, "Blrth, Illness and Death in 18th-Century New England," Journal
of Social History, XI1 (1978), I I 1-128. William H . McNeil, Plagues and Peoples (New York,
1976), 227-228. In America, attachment to Protestant Christianity dampened and channeled
the rationalistic thrust of the Enlightenment. Henry F. May, T h e Erllkhterlnterlt in Anlerica
(New York, 1976). Schofield and David Reher, "The Decline of Mortality in Europe," in
Schofield, Reher, and Alain Bideau (eds.), T h e Decline of ~lfortalityin Europe (Oxford, 1991),
1-3. N o pessimist, Adam Smith, for example, in 1776 casually referred to the "very uncertain
duration of hurrlan life" (An Inquiry into the Natrrre and Causes of the Wealth ofh'ations [New
York, 1937; orig. pub. 17761, 101).
388
/
DANIEL S C O T T S M I T H A N D J . DAVILI H A C K E R
periods. Paradoxically, the spare seventeenth-century Puritan approach to death seems to mesh easily into a quantifjring mentality
that does not pay attention to the particularity of death. A death
is a death is a death.41
The certainty of death and the uncertainty of its timing
remained prominent themes in funeral sermons at the end of the
eighteenth century. The orientation of the late eighteenth and
early nineteenth centuries combines a continuation of the religous
view of death with a greater emphasis on the distinctiveness of
the individual, alive or dead.4'
These glacial changes cannot be said to constitute a transformation of conceptions of mortality risk from a Puritan life table
to the one familiar to modern demographers. Some ministers after
the middle of the eighteenth century started to keep mortality
journals of a more modem character. They omitted deaths from
other locations that previously entered into the eclectic category,
"Remarkable Providences," as well as deaths of other ministers in
New England, concentrating instead on their own parishes and
adding details about age at death and sometimes its cause.43
In retrospect, the perspective of the demographers' life table
seems radical, even bizarre, compared with that of tradition. In
his ironic depiction of a midnineteenth-century British official in
India, Farrell, the Australian novelist, captures effectively the vast
cultural distance between today's skepticism about quantitative
social science and the Victorian passion for statistics: "At the
thought of statistics, the Collector . . . felt his heart quicken with
joy. . . . For what were statistics but the ordering of a chaotic
universe?" The key to modern conceptualization and computation
was a perception of the aggregate. Unlike the moral sciences of
the Enlightenment, nineteenth-century social science made society its subject. Within an aggregate framework, analysts could see
beyond the traditional scheme that placed death in relation to an
41 See especially Stannard, Puritan W a y oJDeafh. 42 The themes persisted in funeral sermons preached during the first half of the nineteenth century in Schenectady, New York. Ilobert V. Wells, '"Frorn Remembering Death to Remembering Life': Changing Styles of Funeral Discourse in Nineteenth-Century Schenec-
tady, New York," paper presented at the Social History Association meetings, 1993. Shively, "History of the Conception of Death," 3-4. 43 Janles H . Cassedy, "Church Record-Keeplng and Public Health in Early New England," in Philip Cash, Eric H. Christianson and J. Worth Estes (eds.), Medicine in Colonial ~2.lassachusetts 1620-1820 (Charlottesville, I~XO),259-263. CULTURAL DEMOGRAPHY
1
389
individual aging through the stages of life. Nineteenth-century
demographers employed a tripartite framework of deaths, population at risk, and time, with numerators being drawn from a death
registration system and denominators from a census. In the new
worldview, "while nothing- is more uncertain than the duration
of a single life, nothing is more certain than the average duration
of a thousand lives."44
Indicating the extent of departure from traditional religous
beliefs is the fact that most of the leaders of the burgeoning science
of demography were liberal in religous belief and radical in
politics. From Joseph Priestly and Richard Price in eighteenthcentury England to William Farr and Lemuel Shattuck in the
nineteenth century, Unitarianism-shockingly liberal by mainstream Protestant standards-was the conservative pole of those
in the statistical movement. Unitarians promoted, both for themselves and for society, activist and rationalistic approaches to
health. Although holding evangelical sentiments in his youth,
Elizur Wright-the leading American actuary of the nineteenth
century-became a prominent atheist. The modem life table, in
sum, was a device well removed from the conventional religious
context .45
44
The quotation &on1 Farrell continues as follows:
Statistics were the leg-irons to be clapped on the thrrfs of ignorance and superstition
which strangled Truth in lonely byways. Nothing was able to resist statistics, not even
Death itself, for the Collector, armed with statistics, could pick up Death, sniff it,
dissect it, pour acid on it, or see if it was soluble. The Collector knew, for exarnple,
that in London during the second quarter of 1855 . . . that out of 10,157 tailors 108
had passed to a better world; that 139 shoerrlakers had gone to their reward out of
26,639 . . . and that was still only a fraction of what the Collector could have told
you about Death. If mankind was ever to clirrlb out of its present uncertainties,
disputations and self-doubtings, it would only be on such a ladder of objective facts.
(Jarnes Gordon Farrell, T h e Siefe of Krirhnapur [New York, 19731, 186).
Daston, Classical Probability, 298-301; John M . Eyler, Victorian Social Medicine: T h e Ideas
and .Methods of William F a r (Baltimore, 1979), 69; Lawrence B. Goodheart, ADolifionist,
,
See the revealing
Actuary, Atheist: Elirur Wright and the Reform Inlprrlse (Kent, Ohio, ~ g g o )145.
reflections o n the contrast between medeval artistic renderings and rrlodern notions of
mortality risk by the statistician Karl Pearson, "The Chances of Death," in idem (ed.), T h e
Chances of Death and other Studies i n Evolution (London, 1897), 1-41,
45 Spencer Lavan and George Huntston Williams, "The Unitarian and Universalist Traditions," in Ronald L. Numbers and Dan-el W. Arrlundsen (eds.), Carirg and Crrriq: Health
arld.Medicine in the Western Religious Traditioru (New York, 1986), 354-375; May, Enliglztenntent
in America, 158-159; Eyler, Victorian Social medicine, 14; Goodheart, ADolitionist, Actrrary,
Atheist; see also Thomas A. Kselman, Death and the Afterlfe iin .Modern France (Princeton, 1993),
28-34. For an argument linking the decline of Calvinism to the flourishing of the life-lnsur-
390
1
DANIEL S C O T T S M I T H A N D J . DAVID H A C K E R
What this shift from an emphasis on the variability and
uncertainty of death to the average life span might mean for the
demographic behavior of individuals can be illustrated by stylized
strategies of fertility response to the death of children. Under a
simple replacement strategy, couples attempt only to replace children who die. Under an insurance or hoarding strategy, however,
couples continue to have children as a hedge against future losses,
even if none has died. In order to adopt contraception, they have
to be reasonably sure that their target family size will be achieved.
The Puritan emphasis on the uncertainty of life militated against
such security and, hence, against contraception. T o the extent that
it was reasonable to be uncertain, couples would overinsure
against what objectively was an unlikely c a t a ~ t r o p h e . ~ ~
For women who married in Hingham, Massachusetts, before
1800, there is no consistent evidence that the deaths of children
before the mothers reached age thirty affected the number borne
from that age until the end of the mothers' childbearing years. In
marriage cohorts of the first forty years of the nineteenth century,
however, such replacement occurred, even though overall mortality of children increased substantially for women married between 1821 and 1840. This contrast clearly cannot be accounted
for by actuality, but it is consistent with a mentality focused on
the predictability of averages rather than the uncertainty of particular cases."
Although one could quip that the Puritans conceived of a
death table rather than a life table, there were several reasons why
the modem life table was not established as a means of both
calculation and perception of mortality until the nineteenth century. New England ministers saw life as a passage through stages
over time. But ministers saw the individual, rather than an aggregate, to be inserted as the denominator of a "table7-an age group
whose members were subject to the risks of death. For them,
ance industry after the 183os, see Viviana A. Rotman Zelizer, .Morals and Markets: T h e
De~/elopmertto j L$e Insurance in the United States N e w York, 1979). For an emphasis on the
importance of the shift in focus &om gambling to benefit the gambler to insurance that
benefits a family, see Daston, "The Domestication of Risk: Mathematical Probability and
Insurance, 1650-1830,'' in Lorenz Ktiiger, Lorraine J. Daston, and Michael Heidelberger
(eds.), T h e Probabilistic Revolution. I, Ideas in History (Cambridge, 1987), 237-260.
46 Preston, Efirts, 10-1 I.
47 Smith, "Population, Family, and Society in Hingham, Massachusetts," unpublished Ph.D.
diss., (Univ. of California, 1973), 248-253.
CULTURAL DEMOGRAPHY
1
391
deaths, too, were particularized, with the more sudden or dramatic
being more remarkable than the routine. T o be counted in a
numerator required that their similarities be appreciated more than
their differences. It would have helped the statistically minded
among the clergy if colonial or local authorities had enumerated
populations by age and reported the results. By the late eighteenth
century, Wigglesworth, among others, saw the need to relate
deaths by age to the age structure. But in their funeral sermons,
ministers continued to advance their favored themes-the lives of
the saints, the individual's process of entry into death, and the
religious implications of death itself for the living. The clustering
of deaths, as in an epidemic or within a family, had collective
implications through themes of divine punishment or generalized
dread.
Puritan ministers failed to understand the life table because
its grasp required comprehension of several distinct elements. An
emphasis on the variance could even be considered more sophisticated statistically than the life table's stress on averages; the mean
is the first moment of a distribution and the variance the second.
Nineteenth-century statisticians, such as Adolphe Quetelet, became fixated on the mean and such notions as the "average man."
Departures from the mean were called "errors," as if a short person
made a mistake in not being taller. Comprehending both mean
and variance and seeing populations as aggregations of unique
individuals is now the sophisticated position. However, there was
enough variance in actual mortality experience to understand why
the ministers exaggerated it. The professional need to promote
religious concems impelled them to deny what they and some of
their hearers clearly recognized-namely, that the young and the
healthy were less likely to die in the near future than the old and
The triumph of the modem life table over the Puritan life
table represents only a limited victory of science over religion in
Western cultural history. The secular decline in mortality to its
current low level has masked the partial nature of the achievement
of rationalism. Today the chance of survival until old age is highly
certain for nearly all groups of American society (see the last
column in Table I), and the modern demographic outlook is
Ernst Mayr, T h e Growth ojBiologica1 Thought: Diversity, Evolution, and Inheritance (Cambridge, Mass., 1982), 45-47.
48
392
1
D A N I E L S C O T T S M I T H A N D J DAVIl3 H A C K E R
virtually unchallenged within the domains of insurance, govemment, and academic science. Yet, to the irritation of some experts,
public opinion hesitates to follow modern demography in certain
areas. In response to the nuclear accident at Three Mile Island in
Pennsylvania and a host of environmental hazards brought under
the scope of public policy in the past two decades, scholars have
contrasted the perception of risk by experts and by the public.
Experts tend to assess risk in terms of the expected number of
fatalities, whereas psychometric studies show that strong a priori
perceptions about a given hazard and an absence of information
about it affect the public's assessment of risk. There is agreement
between the two sides, however, in the evaluation of the elements
underlying risk, such as dread and o b s e r ~ a b i l i t y . ~ ~
The public's current difficulty in assessing risk, when measured against the standards of modern science, recalls the assunlptions underlying the Puritan life table. Labels used to describe
"dread risk'-the
type most feared by the public-include
"uncontrollable, " "dreadful, " "globally catastrophic, " "not equitable, "
an'd "of high risk to future generations." In addition, the impact
of a catastrophic event relates to what that event portends. Similarly, the response to the plight of an identified individual (for
example, a child trapped in a well) is much greater than the
response to a larger number of statistical lives estimated to be saved
by the same expenditure of effort in public health.50
This modem perception of risk is reminiscent of the days
when fasting and humiliation were the appropriate responses to
epidemics and extended discussions of the life and death of saints
were the grist of funeral sermons. It is not so much a return to
the past as a public resurfacing of a perception that has always
been present in the American psyche.
49 The interpretation of secularization as the declining scope of religious authority fits the
American case during the nineteenth century. See Mark Chaves, "Secularization as Declining
Religious Authority," Socinl Fourex, LXXII (1994), 749-774; Douglas and Wildavskv, Risk and
Culture; Slovic, "Beyond Number;: A Broader Perspective on Risk Perception and Risk
Communication," in Deborah G. Mayo and Rachelle D. Hollander (eds.), Arreptable Evidenre:
Srience and Vnlues i n Risk .Unnagernent (New York, 1991), 48-65.
50 Slovic, "Perceptions of Risk," Science, CCXXXVI (19873, 280-285. With the exceptions
of nuclear power and x-rays, and to a lesser extent, police work and nonnuclear electric
power, the similarity of expeas' ranking of risks with those of educated groups in the
population is impressive. W. Kip Viscusi, Fnfnl Tradeoj?: Public nnd Pvivafe Responsibilities for
Risk (New York, 199z), 21-22.