Full text of DOI:10.1111/1468-0289.00175

Economic History Review, LIII, 4 (2000), pp. 706–741
Two textile townships,
c. 1660-1820: a comparative
demographic analysis1
By PAT HUDSON and STEVE KING
ince Wrigley and Schofield’s classic work of 1981, it has been generally
accepted that the sustained population acceleration in England and
S
Wales in the eighteenth century resulted from earlier and more universal
marriage. This was in turn taken to reflect a response to rising income
levels coupled with social changes, such as the decline of live-in farm
service and apprenticeships which had delayed the setting up of new
households.2 Our subsequent understanding has been enriched by his-
torians who have placed stress upon European-wide stimuli to earlier
courtship and marriage, and hence to increases in fertility, especially the
greater mobility and economic and sexual freedoms of young people
resulting from the processes of proletarianization and proto-industrializ-
ation.3 Such causal analysis and model building at national and supra-
national levels have been invaluable but they pose fundamental questions
about the dynamics of population change in varied regional and local
environments.4 Understanding the diversity of experience behind aggre-
gate indices and averages of vital variables calls for complementary
research looking at regional and local patterns, at the distributions (as well
as the means) of demographic variables, and at individual experiences. By
digging beneath the surface of aggregate indicators, and by making more
direct and immediate connections between the processes of economic,
social, cultural, and demographic change, it is possible to uncover worlds
of cause and effect very different from those which satisfy the aggregated
variables and which dominate the large-scale causal analyses.
In western Europe as a whole, geographical variations in demographic
1 Earlier versions of this article have benefited from exposure at the Eighth International Economic
History Conference, Madrid, 1998, and at seminars at the Cambridge Group for the History of
Population and Social Structure and All Souls College, Oxford. We are grateful to the seminar
participants for their helpful comments. The research upon which the article is based forms part of
a larger project on economic, social, and cultural change in industrializing West Yorkshire, elements
of which have been financed during the past decade by the ESRC, the British Academy, the Nuffield
Foundation, and the Leverhulme Trust.
2 Wrigley and Schofield, Population history. This analysis owed much to ideas about the prevalence
of the nuclear family since preindustrial times, originally aired by Hajnal, ‘European marriage
patterns’, Laslett, World we have lost, and idem, Household and family, ‘Introduction’. For a rebuttal
which places much more emphasis on declining mortality, see Razzell, ‘Conundrum’.
3 Seccombe, Millennium of family change; Levine, Family formation; idem, Reproducing families.
4 Wrigley and Schofield highlighted this need as part of their future agenda in 1981: Population
history, pp. 9-10. It was also stressed early by Drake, Population history, and endorsed by Hudson,
Regions and industries, pp. 11-13 and by Wrightson and Levine, Poverty and piety.
Economic History Society 2000. Published by Blackwell Publishers, 108 Cowley Road, Oxford OX4 1JF, UK and 350 Main Street, Malden,
MA 02148, USA.

two textile townships, c. 1660-1820
707
indicators were always greater than temporal variations between the
sixteenth and eighteenth centuries, suggesting that it is vital to consider
spatial differences and their causes.5 Levine pointed a way forward by
stressing the potential importance of occupational determinants and by
suggesting that the English demographic regime contained several demo-
graphic response patterns working alongside one another in differently
weighted combinations over time and space: the peasant, the proletarian,
and the proto-industrial.6 Seccombe refined this type of model in a
European context but questioned any inevitable link between proletarianiz-
ation and changing demographic behaviour. He argued that there was a
difference between demographically damaging or insignificant ‘negative
proletarianization’, caused by displacement from the land, and ‘positive
proletarianization’ which stimulated the take-off in population growth
rates after 1750 because it was accompanied by a commensurate rise in
waged work opportunities.7
The recent collective analysis of 26 reconstitution studies by the Cam-
bridge Group has also been partly concerned to highlight variations in
experience. Certain typologies of demographic change are suggested,
associated with ‘agricultural’, ‘retail trade and handicrafts’, and ‘manufac-
turing’ communities.8 Although only a small sample (8, 5, and 3 respect-
ively of each economic group), these parish types appear to have differed
markedly from each other in nuptiality levels and in other demographic
experiences between the seventeenth century and the nineteenth. Adair
has analysed variations in bastardy in the period 1538-1754, suggesting
that here, too, there were distinct and enduring regional differences in
the nature of courtship and in attitudes both to premarital sex and to
marriage.9 Szreter’s study of the late nineteenth-century demographic
transition further illustrates the enduring nature of regional and social
variations in demographic behaviour in England: far from being a ‘unitary
and unifying’ event, the fertility decline in Britain was one of ‘fundamental
cultural and socio-demographic diversity’.10
Explaining varied experiences is no easy task. Continental scholars
have successfully used ‘micro-history’ to question macro-level connections
between demographic and economic change, and to try to identify more
precisely the stimuli and motivations affecting behaviour in different
5 A point made by Seccombe, ‘Marxism and demography’.
6 Levine, ‘Population history’. Occupational variations had already been demonstrated empirically
in Flinn, European demographic system. Levine, Reproducing families; idem, ‘For their own reasons’;
idem, ‘Asymmetrical, non-linear population dynamics’; idem, ‘Proletarian family’.
7 Seccombe, Millennium of family change.
8 Wrigley et al., English population history. These categories are defined according to the concen-
tration of employment by sector at the time of the 1831 census. But this census enumerated only
the employments of men over 20 years under seven specific headings. It is thus a rather blunt
instrument with which to measure dominant occupations in the early nineteenth century, let alone
during the previous century and earlier.
9 Adair, Courtship, illegitimacy and marriage. These sorts of regional and local differences in
illegitimacy have been identified by others including Blaikie, Illegitimacy, sex and society; Laslett et
al., Bastardy; Sabean, Power in the blood.
10 Szreter, Fertility, class and gender, pp. 533, 539-40.
Economic History Society 2000

708
pat hudson and steve king
communities.11 But relatively little is still known about when and how
inter-generational relationships, relationships between the sexes, the court-
ship process, and marriage ages were affected by real wage shifts, prolet-
arianization, manufacturing employments, or changes in poor relief poli-
cies, especially in Britain.12 And, although the question has been raised
many times, gender-specific demographic reactions to broader cultural
and economic changes have rarely been researched, despite sufficient
evidence to suggest that female motivations in courtship and marriage
were often very different from those of men.13
The seemingly close relationship between proto-industrialization, prolet-
arianization, and the breakdown of ‘traditional’ ways of life, including
demographic behaviour, as argued so forcefully by Levine, Mendels, and
others, is particularly ripe for some local-level rethinking in the British
context.14 European studies over the past two decades reveal that the
economic, social, and demographic consequences of rural manufacturing
and of proletarianization varied widely. Increased manufacturing and
expanding trade were often compatible, rather than in conflict, with
normative social relations and ways of life at local level. Vardi, for
example, has shown that the linen industry of the Cambresis arose from
a response of substantial peasant weavers to additional income-earning
opportunities which fitted in with their existing agrarian culture.15 Tilly,
11 The literature on European communities is considerable. Schlumbohm gives a survey and a
contribution in ‘Micro-history and the macro-models’. See also NEHA, Economic and social history,
and J. Family Hist., 16 (1991), special issue on European marriage patterns. The term micro-history
has started to be used to describe such studies, although originally micro-history was much more
closely associated with detailed biographical or ethnographical work by Italian scholars in particular.
What the two very different types of micro-history have in common is the use of small-scale study
to ask big questions. For wider discussion of the role of micro-history, see Levi, ‘On micro-history’.
For broader views of the value of an analytical local history, see Phythian-Adams, ‘Local history
and societal history’ and Marshall, Tyranny of the discrete. For a recent example of a demographically
oriented micro-study, see King, ‘Chance encounters’, and for discussion of the implications of
extending this approach in studies of industrialization, see Hudson, ‘Challenge of micro-history’.
For further examples and a debate about the relationship between micro and macro accounts, see
Schlumbohm, ed., Mikrogeschichte Makrogeschichte.
12 The best overview of the British research is found in Schofield, ‘British population change’,
where he considers changes in labour demands, urbanization, and the operation of the poor law as
well as ‘future earnings prospects’ in underpinning demographic change. He indicates that the poor
law in particular may have injected some stability into the demographic system in the eighteenth
century. Paradoxically, the recent collective analysis of 26 reconstitution studies by the Cambridge
Group, while adding much to the macro-framework, does little to elaborate the economic and social
experience of individual communities or reconstruct the micro-histories which would allow a perspec-
tive on detailed demographic motivations: Wrigley et al., English population history. On wage and
poor relief factors, see Wrigley and Schofield, Population history; Boyer, Economic history of English
poor law. The most notable British community-level studies are Wrightson and Levine, Poverty and
piety and the same authors’ Making of industrial society.
13 Drake was one of the first historians to emphasize gender differences strongly with respect to
marriage behaviour: Drake, Population and society. See also Hill, ‘Marriage age of women’; Tilly,
‘Women’s history and family history’; Sundt, On marriage; O’Day, Family and family relationships;
Mackinnon, ‘Were women present?’; Gullickson, ‘Proto-industrialization’; idem, ‘Love and power’;
Sharpe, ‘Literally spinsters’; Gandy, ‘Illegitimacy in a handloom weaving community’; Maynes,
Taking the hard road.
14 Levine, Family formation. This study followed a continental tradition initiated by Mendels,
‘Agriculture and peasant industry’, idem, ‘Proto-industrialization’, and Kriedte et al., Industrialization
before industrialization.
15 Vardi, Land and the loom.
Economic History Society 2000

two textile townships, c. 1660-1820
709
Liu, Gullickson, Leboutte, and other researchers suggest a similarly com-
plex relationship between proto-industrial development and existing social
relationships with regard to labour, consumption, material support, and
demographic experience.16 Hendrickx’s study of the Twente weaving
region of the Netherlands shows that neither proto-industry, nor later
industrialization, nor deindustrialization, had any profound influence on
demographic or socio-sexual behaviour.17 Schlumbohm’s study of linen
production in Belm found that the connection between marriage, econ-
omic resources, and the formation of new, self-supporting households
was not apparent, while Spagnoli’s work on Lille also confirms the
absence of a link between industrialization and earlier marriage.18 Lehn-
ing’s research on proto-industry among the peasants of Marlhes near St
Etienne showed no marked shift in the age of marriage: the diversification
of the economy was not accompanied by the fracturing of community or
of existing patterns of reproduction.19 Vardi has demonstrated that the
increased wealth of linen weavers in the village of Montigny did not
cause them to marry earlier or produce more children. They stayed in
the village (when they might, in the absence of proto-industry, have been
forced to leave) and experienced increased prosperity. It was this stability
and improvement in living standards which lay behind population expan-
sion in the Cambresis, largely through reduced mortality.20 Even for
Flanders (the region for which Mendels originally developed his theory
of the close links between rural manufacturing prosperity and marriage
behaviour) this stability and mortality improvement is now the favoured
explanation of the connection between proto-industry and demographic
increase.21
In England, the analysis of industrializing communities has not pro-
ceeded as far. The three proto-industrial areas used by the Cambridge
Group (Shepshed, Gedling, and Birstall) appear to have had a remarkable
uniformity of trend, if not level, in demographic indicators. All three
areas experienced a sustained eighteenth-century fall in marriage ages,
buoyant fertility, and a move from the bottom quartile of a ranking of
the 26 family reconstitutions in infant, child, and adult mortality terms
during the period 1675-1749 to near the top of the rankings by the early
16 Liu, Weaver’s knot; Gullickson, Spinners and weavers; idem, ‘Proto-industrialization’; Leboutte,
‘Adaptation, reconversion, mutation’; essays in Medick and Sabean, eds., Interest and emotion; Medick,
Weben; Schlumbohm, Lebenslaufe.
17 Hendrickx, ‘From weavers to workers’; idem, ‘In order not to fall into poverty’.
18 Schlumbohm, ‘Micro-history and the macro-models’; Mendels, ‘Proto-industrialization’; Spag-
noli, ‘Industrialisation, proletarianisation and marriage’.
19 Lehning, Peasants of Marlhes, p. 43. Mean age of first marriage in 1841-70 was 25.6 for women
and 30.56 for men, compared with 25.05 and 29.78 respectively during the following three decades.
For an excellent survey of research on France see Lewis, ‘Proto-industrialization in France’. He
argues that historians of France have been much more willing than those of Britain to adapt the
original conceptual approach of proto-industry theory rather than just to use case studies to find it
wanting: p. 162.
20 Vardi, Land and loom. For similar arguments see Spagnoli, ‘Industrialisation, proletarianisation
and marriage’; Lehning, Peasants of Marlhes.
21 Vandenbroeke, ‘Proto-industry in Flanders’, p. 107; idem, ‘Regional economy’; Mendels,
‘Proto-industrialization’.
Economic History Society 2000

710
pat hudson and steve king
nineteenth century. Over the long eighteenth century, the population of
these proto-industrial areas grew by 240 per cent, somewhat behind
urban districts, but way ahead of small towns and rural areas.22 There
is much in these observations to support a link between the disruptive
effect of rural industry and demographic change which would contrast
with the recent continental literature. But such a conclusion may be
premature. Birstall was a parish of 14 townships whose individual demo-
graphic experiences were conflicting rather than complementary.23 And
in other areas of West Yorkshire protoindustry failed to have the impact
on demographic behaviour which was initially suggested by historians.24
In the midlands and elsewhere, a lowering of the age at marriage most
often appears to have preceded industrial expansion and was more likely
to have resulted from prior agrarian changes than from proto-industry.25
Both continental and some British studies thus show that localized demo-
graphic patterns appear to have had a marked variability and a durability.
Industrial development did not merely, or always, disrupt these localized
patterns but, on the contrary, often adapted significantly to them.
This article examines the demographic experiences of two Yorkshire
textile manufacturing townships, Sowerby and Calverley, in the eighteenth
and early nineteenth centuries.26 It seeks to highlight the ways in which
micro-history can deepen our understanding of the character and chron-
ology of change which has been suggested by research at macro-level.
The approach taken here is comparative in two respects. First, because
the two townships in question, though both experiencing rapid proto-
industrial expansion, had different agrarian structures and institutional
histories and produced different sorts of cloths for different markets. This
involved very different labour demand conditions, different production
processes and technological changes, contrasting patterns of organization
of trade, and varying fluctuations in prosperity. They also had different
patterns of local governance and of poor relief administration. The simi-
larities and the contrasts between these industrial communities enable
both general and specific connections between social, cultural, economic,
and demographic change to be identified more clearly. Second, the
approach is comparative in that it creates the opportunity to compare
our family reconstitution results for Yorkshire with those produced for
other proto-industrial communities and for rural and urban localities in
other parts of Britain and western Europe. This makes it possible to say
22 Wrigley et al., English population history; Wrigley, ‘Brake or accelerator?’.
23 A copy of the reconstitution documentation can be found in the library of the Yorkshire
Archaeological Society, Leeds. This suggests that infant mortality in Wyke was substantially greater
than in Tong, a feature which would be intensified by the addition of the numerous Moravian
infant and child deaths relating to Wyke from the Moravian registers of Calverley.
24 Hudson and King, ‘A sense of place’; King, ‘Nature and causes’.
25 Carpenter, ‘Peasants and stockingers’; Wall, ‘Real property, marriage and children’.
26 This is the first stage of broader research considering the links between economic, socio-cultural,
and demographic change as well as the integrated nature of these variables and their impact upon
everyday life in the townships. For other results see Hudson, ‘Landholding and the organisation of
textile manufacture’; Hudson and King, ‘Rural industrialising townships’; Hudson and King, ‘A
sense of place’; King, ‘Nature and causes’; idem, ‘Reconstructing lives’; idem, ‘Migrants on the
margin?’; idem, ‘Dying with style’; Hudson and King, Industrialization.
Economic History Society 2000

two textile townships, c. 1660-1820
711
something about the varieties of demographic profile and behaviour which
lie behind macro-level indices and about the similarities and differences
between communities across Europe, rather than concentrating upon
national patterns.
I
The townships of Calverley-cum-Farsley and Sowerby were located within
the West Yorkshire textile district which came to dominate much of the
English woollen cloth trade by the later eighteenth century (figure 1).27
From an early date, Sowerby was more than double the size of Calverley
in population terms. In the hearth tax of 1664, Sowerby had 312
households while Calverley had 127. By the mid-1750s the population
of Calverley had risen to about 1,400, while that of Sowerby in 1764
was 3,004. The 1811 census records Sowerby with 5,177 inhabitants
compared to Calverley with 2,390.28 Late eighteenth-century population
growth in both places involved substantial net in-migration as well as
sustained natural increase.
R. Wharfe
IIkley
Worsted area
200
Arthington
Otley
200
Woollen area
Guiseley
Keighley
Oakworth
Fancy woollen
area
Rawden
Bingley
Shipley
Horsforth
Heavy woollen
area
Haworth
Oxenhope
Bramley
Armley
Hunslet
Denholme
200m contour
LEEDS
CALVERLEY
BRADFORD
Pudsey
SOWERBY
Heptonstall
A
.
R
i
200
Ovendon
Morley
r
e
Gomersal
Batley
Dewsbury
Hipperholme
HALIFAX
R. Calder
Cleckheaton
Elland
Sowerby Bridge
Ossett
Wakefield
0
Thornhill
Kirkheaton
HUDDERSFIELD
200
Flockton
Emley
10
Almondbury
200
Km
Honley
2 0 0
R. Dearne
Barnsley
Figure 1. The location of Sowerby and Calverley within the worsted and
woollen areas respectively of West Yorkshire c. 1780–1830
Although little more than 10 miles apart, the townships had very
different forms of proto-industrial organization. Calverley had an early
and heavy dependence on the broadcloth industry which apparently did
27 Heaton, Yorkshire woollen and worsted; Hudson, Genesis.
28 Hearth tax: PRO E179/210/393, 16 Charles, Lady Day 1664. Population Book 1764, Sowerby,
SPL192, West Yorkshire Archive Service (WYAS), Halifax. Calverley Survey in the parish registers,
WYAS, Leeds. 1811 census lists, WYAS, Halifax.
Economic History Society 2000

712
pat hudson and steve king
not spring from poor landholding or particularly rapid population growth.
It may instead have reflected the active pursuit of economic and social
betterment, important with a land market which was almost completely
leasehold and tightly controlled. Landholding in the township was polar-
ized, with one family (the Calverleys, succeeded by the Thornhills in
1754) owning most of the land, only a very small middle stratum of
landholders, and the bulk of the population who rented small plots or
had little direct connection with working the land.29 Both the Calverley
and Thornhill families, but especially the former, were active in attracting
farmer-weavers to rent land, in order to maximize their rental income.
This was a further encouragement to textile manufacturing in the town-
ship.30 For much of the eighteenth century, over 70 per cent of the male
working population were engaged in the textile industry, predominantly
as clothiers.31 Even by the early nineteenth century, when clothworkers
were complaining vigorously about the changing character of their trade
and the marginalization of independent men, small-scale clothiers making
cloth in a family production unit alongside a small landholding (usually
rented and of between 9 and 11 acres) were the representative producers
in the township. Thus, while large ‘putting-out’ employers had begun to
accelerate ‘positive proletarianization’, there is evidence that upward social
mobility via small businesses and family networks of business remained
a possibility until well into the nineteenth century.32 Household pro-
duction units generally engaged family labour with one or two journeymen
and apprentices, although there was considerable seasonal flexibility and
agrarian by-employments were ubiquitous. Before the enclosures of com-
mons and waste land which occurred from the 1750s, additional income
and subsistence could be earned from pasturing animals, in particular
dairy cows and poultry on the commons. In the second half of the
eighteenth century such opportunities declined, releasing low cost female
labour for spinning.33 Technological changes, including the application
of water power to scribbling and carding and the slow introduction of
mechanized spinning at the end of the eighteenth century, progressively
removed some of the major textile work undertaken by women and
children in the clothier household. This encouraged the movement of
young women into Leeds and a diversification of female economic activity
especially into petty production of foodstuffs and drinks for local sale.34
By contrast, Sowerby was an upland area where many of the inhabitants
in the second half of the eighteenth century, if not earlier, became
29 For details of the social and landholding structures in the two townships, see Hudson, ‘Landhold-
ing and the organization of textile manufacture’.
30 King, ‘Nature and causes’ p. 68.
31 The occupational structure of the two townships is discussed in detail in Hudson and King,
‘Rural industrialising townships’. Parish register, probate, and other sources are used.
32 See King, ‘Migrants on the margin?’.
33 The release of cheap female labour after enclosure in West Yorkshire and the boost which this
gave to textile competitiveness in the region are explored in Quaide, ‘Great wheel and the goose’.
34 Poor law records testify to the local production of foodstuffs and drinks: see King, ‘Recon-
structing lives’. Calverley was the second most common place of origin for marriage partners in
Leeds who did not claim Leeds itself as their place of residence: King, ‘Nature and causes’ p. 261.
Economic History Society 2000

two textile townships, c. 1660-1820
713
proletarianized workers in the textile industry. After the 1760s the town-
ship predominantly produced worsteds, largely for export markets. This
was an industry organized on a ‘putting-out’ basis by a mixture of urban
merchants and local operators. Dependence on the textile sector during
the eighteenth century was higher than in Calverley, at around 80 per
cent of men in recorded employments. By the 1730s over 50 per cent
of male workers are recorded as weavers, a label which had entirely
displaced the term clothier by that decade, reflecting a real shift to
putting out at the expense of integrated spinning/weaving households and
independent producers.35 However, the description weaver covered a
wealth of different household situations. In some cases entire families
specialized in just one part of the production process, predominantly
weaving. This form increased with the mechanization of spinning at the
end of the eighteenth century. In other households family members were
involved in tasks at different stages of the production process, often
working for different employers and engaged in other seasonal and inter-
mittent work not connected with the textile sector. The landholding of
weaving households varied enormously from the landless to those farming
20 acres or more. These variations reflected different mixes of textile
specialization and agrarian by-employments from cottager-weavers to
those who farmed and leased land and employed neighbours and others
as textile workers. Landless weavers were on the increase in the eighteenth
century: a product of the positive, if precarious, proletarianization stimu-
lated by the spread of putting out. There was a bigger middling stratum
of landholders in Sowerby than in Calverley though the land which they
owned was worth much less in agricultural terms. Most middling land-
holders were weavers and/or textile putting-out merchants (described as
yeomen). Unlike in Calverley, farming as an occupation did not figure
in Sowerby in the eighteenth century. Although enclosure by act occurred
late (in the 1840s), use rights had been considerably reduced long before
and the pasturing potential of common lands was in any case limited by
the topography and soils of the Pennines. Income earning from spinning
textiles had figured centrally in the lives of most women in the township
since at least the sixteenth century, but slow mechanization of spinning
from the 1780s was beginning to undermine women’s waged work opport-
unities in the home in favour of spinning in larger workshops and
proto-factories.
In short, although both townships were dominated by textile manufac-
ture, textile-centred family economies formed a larger proportion of all
households in Sowerby than in Calverley. In the latter, textile households
still predominated over all others and they were generally more integrated
family work units than in Sowerby. However, Calverley had more alterna-
tive employments: (for men) in agriculture, building, and metal working
and (particularly for women) in the service sector. In both places there
was a marked degree of labour flexibility especially in the case of women
35 The proportion of weavers fluctuated between 51 and 73 per cent of occupied males after 1740
depending on source and time period. See Hudson and King, ‘Rural industrialising townships’.
Economic History Society 2000

714
pat hudson and steve king
and children and a tendency for their proto-industrial participation rates
and roles to contract (in the face of both rapid population growth and
technological change) at the end of the period, after half a century of
marked expansion.
In both townships, poverty increased at the same time as industry
flourished. Poor law data for Calverley suggest that expenditure started
to rise soon after enclosure in 1755 and peaked at the turn of the
nineteenth century, a function both of inflation and of the increase in
numbers on relief.36 This said, the level of relief offered was meagre,
particularly to the elderly and infirm. Average pensions of less than 2s.
per week were one-third or half lower than the average level of pensions
offered by many southern counties.37 Moreover, such pensions were
offered to relatively few people, a reflection of the fact that relatives were,
and were expected to be, active in the provision of welfare.38 Fragmentary
returns for the mid-eighteenth century indicate that poor relief spending
in Sowerby also increased consistently as the worsted industry consoli-
dated its presence, and somewhat faster than population growth in the
township.39 By the time poor relief returns were being made to Parlia-
ment, after 1802-3, Sowerby was spending 2s. 3d. per pauper, 16 per
cent above the figure for Calverley.40 These broad differences persisted
well into the third decade of the nineteenth century, when they diverged
further in the run-up to the new poor law.41 Differences also emerge
when we look at other aspects of welfare. Middling groups in Sowerby
appear to have responded favourably to parish-level initiatives on chari-
table donations to relieve suffering arising out of trade crises, while the
Calverley middle class appear to have been both less prominent and
much less responsive collectively to poverty.42 The level of activity in
placing hundreds of pauper children in apprenticeships in Sowerby during
the eighteenth century has no counterpart in Calverley where boarding
was more common.43
It is difficult to discern how far the differences in welfare structures
36 King, ‘Reconstructing lives’, p. 324.
37 There were some exceptions but it appears that the poor law regimes of the south and east
were relatively generous compared with those of the north and west: Smith, ‘Ageing and well being’.
38 It is notable that those without relatives (through blood or marriage) tended to come to the
poor law earliest in the life-cycle and remained on communal relief the longest. Relatives are here
defined and identified using the standard methodology employed by Wrightson in ‘Kinship in an
English village’. This excludes the refinements suggested by Reay, ‘Kinship and the neighbourhood’
or Cooper and Donald, ‘Households and hidden kin’.
39 SPL series, WYAS, Halifax; ‘Marshall papers’, WYAS, Leeds. For a review of earlier docu-
mentation on the poor law which now appears to have been lost, see Gott, History and antiquities
of Halifax.
40 This widens when allowance is made for differential administrative costs. For a discussion of
filtering procedures used to separate administration from other items in poor law accounts, see
Hopkin, ‘Poor law strategies in Yorkshire’.
41 For these figures, see Rose, ‘New poor law in the West Riding’. See also Hopkin, ‘Poor law
strategies in Yorkshire’.
42 Smail, Origins of middle class culture; King, ‘Reconstructing lives’; dole book for Halifax parish,
SPL 164, WYAS, Halifax.
43 Despite a well preserved set of poor relief accounts for Calverley, only a handful of pauper
apprenticeship indentures have been discovered whereas there are over 500 for Sowerby for the
eighteenth century: SPL, 108, WYAS.
Economic History Society 2000

two textile townships, c. 1660-1820
715
between the two places reflected differences in wealth distribution, the
scale and intensity of poverty, differences in the ease with which the poor
could be absorbed within the household economies of their relatives, or
differences in philosophy on the part of those administering relief.44
Population appears to have grown more consistently in Sowerby than in
Calverley during much of the eighteenth century, reflected in the relative
movement of vital event totals shown in figure 2. In both townships
proto-industry was accompanied by growing wage dependency and per-
iodic economic insecurity, associated with trade fluctuations. Artisans in
Calverley maintained small plots of land which could help to cushion
them from the worst excesses of market instability but there is no doubt
that the numbers of individuals and families existing on the margins of
economic viability were increasing in both places. This was reflected both
in the poor relief figures and in rising mortality rates.
In order to explore the dynamics of population change, in these very
different proto-industrial townships, they have been subject to family
reconstitution.45 This process yielded 1,664 family histories and 6,000
partial or full individual life histories for the township of Calverley, and
3,950 family histories and 11,000 partial or full individual life histories
for Sowerby. From these we are able to calculate aggregate demographic
variables, for both communities, comparable with other studies and to
explore the demographic experience of sub-groups and individuals within
a wider quantitative framework. To this end we have linked family and
group life histories to a range of economic, social, and institutional
information derived from sources including poor law material, and tax
returns, wills and inventories, depositions, leases, business records, news-
papers, and the early nineteenth-century census returns. Thus demo-
graphic behaviour can be related much more closely to its economic,
social, and cultural context than is usually the case.
The key problem with reconstitution studies is that they focus upon
the experience of the least mobile sections of the population. Debate on
the representativeness of this group in terms of demographic (and other)
characteristics has been fierce and has spawned a number of technical
procedures for making allowances for the fact that it is rarely known at
which date a person moves from observation.46 Ultimately, the impact
44 Factors determining levels of poor relief and poor relief policy in Calverley are the subject of
King, ‘Reconstructing lives’, and patterns of relief in both townships are further explored in Hudson
and King, Industrialization.
45 We are alert to the persistent doubts about the technique of family reconstitution and the
subset of the population whose demography it identifies (see, e.g., Razzell, ‘Growth of population’;
idem, ‘Conundrum’; Ruggles, ‘Migration, marriage and mortality’; idem, ‘Limitations of English
family reconstitution’). The comparative approach is probably one of the most illuminating ways of
using reconstitution results because conclusions are not based solely on the absolute accuracy of
specific applications: comparisons are useful in themselves as long as a standard methodology
prevails. The demographic estimates presented in this article were calculated using standard sources
and methods of family reconstitution. Elsewhere we have reported on the process of ‘enriched
reconstitution’ which allows a more sensitive analysis of demographic behaviour: King, ‘Historical
demography’.
46 See principally, Ruggles, ‘Migration, marriage and mortality’, although Wrigley has provided a
powerful rebuttal of this critique: ‘Effect of migration on estimation’.
Economic History Society 2000

716
pat hudson and steve king
Figure 2. Baptisms and burials (nine-year moving averages) from Church of
England and nonconformist registers for Sowerby and Calverley, 1680–1820
Source: Sowerby and Calverley Church of England parish registers and microfilm of Sowerby Independent and
Methodist Register, West Yorkshire Archive Service, Leeds and Halifax. Calverley Moravian registers microfilm,
Yorkshire Archaeological Society. Calverley Independent and Baptist registers, PRO. A small number of vital
events for Sowerby and Calverley are also included from the parish registers of neighbouring churches and chapels
and from overlapping nonconformist circuits.
of migration upon demographic measures such as infant mortality or age
at first marriage and, more especially, on measures of concentration
(employed below) is unknowable. However, in common with most proto-
industrial communities, both Calverley and Sowerby saw a decline of
out-migration as the intensity of industry increased, enabling the reconsti-
tution to capture a growing proportion of the total population. The
percentage of marriage partners for whom both baptism and burial records
exist in Calverley rises from 27 per cent to 44 per cent comparing 1700-
49 with 1750-99. The comparative figures for Sowerby are 22 per cent
Economic History Society 2000

two textile townships, c. 1660-1820
717
and 39 per cent, indicating a significant increase in ‘stayers’ in both
townships by the second half of the eighteenth century.47 Furthermore,
as Wrigley has suggested, those who left were often not unrepresentative
of those who stayed.48 It is possible to trace a proportion of those who
left Calverley for Leeds in the marriage registers of the town. This
suggests that out-migrants had very similar marriage age profiles (in terms
of mean and standard deviations) to those who stayed.49
Nonconformity is also an issue because Anglican registers may capture
a biased sample in terms of demographic behaviour. In our study, the
databases underpinning the reconstitution samples contain the surviving
nonconformist registers (which include the largest groups of nonconform-
ists in the area: Independents, Baptists, and Moravians). Including the
nonconformist populations in the reconstitution makes a negligible differ-
ence to the profile or behaviour of vital variables. These groups constitute
only a small percentage of the life events encompassed by family reconsti-
tution. Even in late eighteenth-century Calverley, which fell under the
influence of the sizable Moravian community at Fulneck, nonconformist
events make up only 9 per cent of all births recorded in family life histor-
ies.
Section II exposes our findings in comparative context and Section III
begins to explain them. What follows has important implications for
future demographic analysis at local and national levels because of the
emphasis which we place upon the impact of subgroup behaviour within
broader, more stable, demographic regimes.
II
Important differences in the demographic framework of the two town-
ships, and between the townships and a sample of other reconstitutions,
are revealed in figures 3-9 and table 1. Mean female age at first marriage,
emphasized as the key driver of English demography at national level
after 1750, appears to have only limited relevance to the demographic
systems of Calverley and Sowerby in the long eighteenth century. Mean
female first marriage ages in Calverley were consistently low and stable
over the long eighteenth century. The number of cases is too small to
offer definitive explanations, but stability of mean marriage ages in the
face of proto-industrial development seems to have been relatively com-
mon on the continent.50 The Calverley study thus poses important ques-
tions for proto-industrial and demographic historians of England. Sowerby
is more consistent in following the ‘proto-industrial pattern’ of some
decline in female marriage ages suggested by Wrigley et al., but even
47 King, ‘Migrants on the margin?’, p. 288, which gives more details of residential stability for
Calverley; reconstitution results for Sowerby.
48 Wrigley, ‘Effect of migration on estimation’.
49 Reconstitution results.
50 See, e.g., Hendrickx, In order not to fall; Trompetter, Agriculture, proto-industry and Mennonite
entrepreneurship; Gutman and Leboutte, ‘Rethinking proto-industrialization’; Cerman and Ogilvie,
European proto-industrialization; nn. 18-21 above.
Economic History Society 2000

Table 1. The distribution of female first marriage ages in Calverley and Sowerby
Calverley
Sowerby
3rd quartile
9th decile
Mean
1st decile
1st quartile
Years
1st quartile
1st decile
Mean
9th decile
3rd quartile
25
25.2
25
25.1
25
24.7
28.6
28.9
28.5
28.2
28
23.2
23.6
23
23.2
22.9
22.8
18.5
18.9
18.2
18.3
18
20.2
20.1
20
20.4
20
1680-99
1700-24
1725-49
1750-74
1775-99
1800-24
20.7
20.8
20.4
20
19.7
19.5
19.1
19.1
18.6
18.2
17.9
17.6
24.6
24.3
23.8
23.6
23.4
23.1
31.8
31.4
30.4
29.6
29
27.5
27.1
26.2
25.3
25.1
24.9
27.7
17.8
19.7
28.1
Note: Sample sizes: Calverley 650, Sowerby 1365. Distributed as follows in chronological order by sub-period: Calverley: 40, 65, 90, 137, 156, 162; Sowerby: 65, 140, 200, 270,
332, 358.
Source: Family reconstitutions. For comparable data see Wrigley et al., English population history, pp. 146-7

two textile townships, c. 1660-1820
719
26 studies
Gedling
Shepshed
Calverley
Sowerby
1800-37
1750-99
1700-49
1650-99
20
21
22
23
24
25
26
27
28
Age (Years)
Figure 3. Mean female first marriage ages in comparative perspective
Note: sample sizes: Calverley 762, Sowerby 1,650
Source: Family reconstitutions, and Wrigley et al., English population history, pp. 184–5
here the fall was only of the order of 18 months over the period 1680-
1820.51 As figure 3 suggests, the fall in female marriage ages in most
other studies, including those of other proto-industrial areas such as
Gedling and Shepshed, was significantly more pronounced although the
base from which the decline started in Sowerby was lower.52 This obser-
vation is surprising given the rapid rise of the putting out system,
51 For a model of the significance of falls in the age at marriage to measures such as the gross
reproduction rate, see Wrigley et al., English population history. Even on a conservative estimate, up
to one half of this variation might be explained by biases in all reconstitution data along the lines
suggested by Ruggles, ‘Migration, marriage and mortality’, although Wrigley has provided a powerful
rebuttal of this critique: ‘Effect of migration on estimation’.
52 Low marriage ages characterize the small-scale reconstitution of the village of Countesthorpe,
near Shepshed: Carpenter, ‘Peasants and stockingers’. Continental studies have variously emphasized
the degree of dependency upon money wages and the nature of the sexual division of labour as the
key variables determining whether proto-industrial areas would experience falling female ages of
first marriage. See Spagnoli, ‘Industrialisation, proletarianisation and marriage’; Gullickson, ‘Proto-
industrialization’. One could also argue, of course, that the relationship between proto-industry and
demographic change is dependent upon threshold levels of proto-industrial dependency: that the
major decline in both Sowerby and Calverley may have occurred earlier than 1660, perhaps in the
sixteenth century when textile manufacturing first began to engage a significant proportion of the
population. We have tested this for Calverley by taking the family reconstitution back to the early
seventeenth century on a selective basis. No decline in marriage ages is discernible for the seventeenth
century. Marriage licence evidence supports this, despite the fact that in the early seventeenth
century less than 30% of male occupations were associated with textiles in the township compared
with over 70% a century later.
Economic History Society 2000

720
pat hudson and steve king
agrarian change, and the speed of proletarianization in Sowerby in the
eighteenth century.53
Aggregate indicators are significant but it is the dispersal of the marriage
ages, rather than their mean, which may well hold the key to characteriz-
ing and explaining nuptiality. As Goldstone saw, and as Wrigley et al.
confirm, a central feature in the falling age at marriage nationally was
the replacement of a core of late-marrying women with a core of early-
marrying women, accompanied by a narrowing of the variability of female
marriage ages.54 Sowerby would appear to line up with such national
trends and to match the experience of other proto-industrial communities
in this respect. The standard deviation of female marriage ages fell
over time, to around the mean standard deviation of all 26 Cambridge
reconstitutions, and duplicates that calculated for Shepshed.55 Identifying
the behaviour of quartiles and percentiles in marriage ages for Sowerby
illustrates that shifts in the mean were largely a result of changing
behaviour at either end of the spectrum (table 1). Calverley, by contrast,
with a stable standard deviation of 3.0, and an interquartile range of 7,
as well as relative stability of the mean throughout the period, really stands
out.56 Male marriage ages are generally accepted to be less important to
the dynamics of local demographic systems but it is instructive to note
that male age at first marriage was relatively stable over the long eight-
eenth century in both townships although it seems to have been more
concentrated in the age spectrum in Sowerby than in Calverley (figure
4).57 This contrasts with the Cambridge Group sample where male
marriage ages fell overall by around 3 years with proto-industrial Gedling
and Shepshed experiencing a more pronounced fall.
These observations about nuptiality complement the marked variations of
local experience found in the 26 Cambridge reconstitutions and add weight
to a continental literature which calls for a reinterpretation of the demographic
53 The limited decline in mean female marriage age and the low base place the township at the
extreme end of the Cambridge Group’s ‘proto-industrial’ typology or even outside it.
54 Goldstone, ‘Demographic revolution’. This phenomenon may have been particularly marked in
handicraft communities. At the turn of the nineteenth century the gap between the highest and
lowest mean female marriage ages in the reconstitution sample was four years. This was more than
the fall in aggregate marriage ages over the course of the long eighteenth century: Wrigley et al.,
English population history.
55 Levine, Family formation p. 161. See also Wrigley et al., English population history.
56 The number of cases for this township is relatively small and generates all of the potential
problems of manipulating data based upon a sub-sample of the population. Yet extensive testing of
these findings, and the linkage on which they are based, suggests that this picture of low, stable,
and concentrated marriage ages is more than simply a product of random variation or incorrect
linkage. See King, ‘Nature and causes’.
57 In short, women in the artisan township of Calverley (where proletarianization occurred later
and more slowly) generally married within a restricted age band around the early twenties, while
their husbands showed rather more variation in marriage ages. In Sowerby, it was the grooms who
had the most concentrated marriage age experience, which might suggest that a ‘normal’ age at
marriage for men was accompanied by a need to find a wife even if she was considerably older or
younger than the man concerned. A continuing tendency for Sowerby grooms to draw their brides
from within the township or somewhere in close proximity may have intensified this experience. For
more on marriage horizons, see Hudson and King, ‘Rural industrialising townships’.
Economic History Society 2000

two textile townships, c. 1660-1820
721
26 studies
Gedling
Shepshed
Calverley
Sowerby
1800-37
1750-99
1700-49
1650-99
0
5
10
15
20
25
30
Age (Years)
Figure 4. Male ages at first marriage in comparative perspective
Note: sample sizes: Calverley 784, Sowerby 1,635
Source: Family reconstitutions, and Wrigley et al., English population history, pp. 184–5
implications of proto-industrialization.58 The general, rapid, and pronounced
falls in marriage ages suggested by Wrigley and Schofield and highlighted
by Levine, Mendels, and others in proto-industrial areas, have little place in
these two textile townships. However, as figure 5 suggests, the picture is not
yet complete. Celibacy in Sowerby more than halved between the late
seventeenth century and the early nineteenth, with the fall most pronounced
in the later eighteenth century.59 This contrasts with the wider sample of 26
reconstitutions which experienced a more limited decline on trend in the
1800-25
1750-99
Sowerby
Calverley
1700-49
1680-99
0
2
4
6
8
10
12
Percentage celibate
Figure 5. Female celibacy rates in Calverley and Sowerby
Source: Family reconstitutions. For method of calculation see text.
58 See, most recently, Hendrickx, ‘In order not to fall’; Trompetter, Agriculture, proto-industry and
Mennonite entrepreneurship.
59 Methods used to calculate female celibacy rates from reconstitution data involve identifying
women who reach the age of 45 without ever being married. The figures are at best ‘fuzzy’ and
provide only an approximation. Male celibacy rates cannot be calculated. This is unfortunate given
the concentrated male marriage age range in Sowerby which might suggest that for men the ‘rather
never than late’ maxim has considerable pertinence after 1750. For this maxim, particularly in
relation to spinsters pre-1700, see Weir, ‘Rather never’.
Economic History Society 2000

722
pat hudson and steve king
female celibacy rate. The experience of Calverley could not be more different,
suggesting that not only were marriage ages low, but for much of the early
part of the long eighteenth century, celibacy rates were low as well. By the
early nineteenth century, the township represents the high side of recorded
female celibacy rates in proto-industrial areas (where rates had fallen) but
was still below the ‘national’ mean. Overall, changing nuptiality in Calverley
cannot have been responsible for the population growth in the township. In
Sowerby, by contrast, a gentle (though not equitably distributed) fall in
female age at first marriage combined with an increasing rate of marriage
(especially after 1750) to contribute to the upward drift of population. Such
results begin to highlight the importance of micro-studies in exposing different
paths and patterns of demographic development between proto-industrial
communities.
Fertility trends provide further evidence of variation. The certainty of
1981 that marital fertility levels were roughly stable and uniform across
English time and space has been replaced with a realization that total
marital fertility rose during the long eighteenth century.60 The proto-
industrial communities in the Cambridge sample all experienced an
upturn in marital fertility after the middle of the eighteenth century. The
same was not true for rural areas or market towns. Figure 6 presents
birth interval data as an indicator of fertility changes in our townships
in comparative perspective. For both places, birth intervals suggest that
marital fertility was high and rising compared with other proto-industrial
and rural communities. In Calverley marital fertility rose particularly
strongly and it, rather than nuptiality, appears to have been the stimulus
to natural population growth. Our township studies support the con-
clusion that industrial areas stood at the heart of marital fertility rises in
the later eighteenth century with one important proviso. The key feature
of aggregate fertility change was an increase in the proportion of birth
intervals of 12-18 months and a fall in the proportion of 36 months or
more, as indicated in figure 7. This resulted from growth in size of a
subset of women who had consistently short birth intervals (around 15
per cent of those for whom we have complete fertility life cycles in the
two townships) and a further shortening of their birth spacing in the
middle of the child-bearing years. It did not result from a generalized
reduction of birth intervals for the majority. This might, of course, reflect
a reduced tendency for certain families to have incomplete local church
baptisms but it is notable that a similar concentration of high fertility
experience is to be observed among the Moravians of Calverley for whom
registration appears to have been rigorous.
In the national figures and in Sowerby (figure 8), illegitimate fertility
complements this picture. Sowerby saw bastardy levels rise considerably
such that by the early nineteenth century over 8 per cent of all births in
the township were illegitimate. This was generally in line with the experi-
ence of other proto-industrial communities, but ahead of the experience
in rural areas. As with other demographic indicators, illegitimacy was
60 Wilson, ‘Natural fertility’; Wrigley et al., English population history.
Economic History Society 2000

two textile townships, c. 1660-1820
723
Figure 6. Birth intervals in Calverley and Sowerby in comparative perspective
(Figure continues overleaf)
Note: sample sizes: Calverley 2,736, Sowerby 5,377
Source: Family reconstitutions. For comparative material see Yorkshire Archaeological Society, ‘Family Reconsti-
tution Collection’, and Don, ‘Reconstitution of Paxton’.
associated with a finite subset of the population (about 8 per cent of the
potential pool of bastard-bearing women between the ages of 18 and 29
for any 20-year cohort) and with repeated illegitimacy experience within
this group.61 Infant mortality among illegitimates was high in Sowerby
in common with the findings of other studies.62 Because of this the
61 This concentration of illegitimacy within a finite subset has been identified for some time. See,
for example, Laslett et al., Bastardy.
62 Crudely estimated, this averaged 394 per 1,000 compared with 120 per 1,000 of legitimate
infants. See also Wrigley et al., English population history, pp. 219-22.
Economic History Society 2000

724
pat hudson and steve king
Figure 6. continued.
impact of illegitimacy itself upon population growth rates was limited.
But the pre-nuptial pregnancy (PNP) rate (42 per cent of all first births
by 1810) which grew alongside illegitimacy in Sowerby was significant
enough to impact upon total fertility (which rose by over one-fifth between
the early and late eighteenth century) and upon population growth rates.
These experiences make the history of illegitimacy in Calverley very
intriguing. In contrast to most other studies, illegitimacy fell over the
course of the eighteenth and early nineteenth centuries and never reached
Economic History Society 2000

two textile townships, c. 1660-1820
725
%
100
90
80
70
60
50
40
30
20
10
0
36+ Months
24-36 Months
18-24 Months
12-18 Months
9-12 Months
1700-09 1710-19 1720-29 1730-39 1740-49 1750-59 1760-69 1770-79 1780-89 1790-99
Decade
Figure 7. Distribution of birth intervals for births 1–8 (Calverley and Sow-
erby combined)
Note: sample size: 8,571
Source: Family reconstitutions
Birstall
Colne
Shepshed
Sowerby
1800-20
1750-99
1700-49
1680-99
Calverley
0
2
4
6
8
10
12
14
16
Percentage of first births illegitimate
Figure 8. Calverley and Sowerby illegitimacy in comparative perspective
Source: Family reconstitutions. For comparative material see King, ‘Nature and causes’
more than 2 per cent of all births after 1700.63 As in Sowerby, the vast
majority of illegitimate births were confined to a small subgroup while
PNP was much more widely dispersed. PNP rates in Calverley rose from
less than 10 per cent to 36 per cent of all births by the early nineteenth
63 There is little evidence to suggest that this is a reflection of inadequate recording of illegitimate
children, especially since at various points in this period the registers provide an extraordinary level
of detail on the registration process. Informal marriage could, of course, upset this picture since it
Economic History Society 2000

726
pat hudson and steve king
century.64 This evidence of the limited distribution of illegitimacy experi-
ence, especially in Calverley, and the much more widespread incidence
of PNP, is difficult to reconcile with the commonly stated thesis that
illegitimacy in the eighteenth century was largely precipitated by the
breakdown of post-conception marriage plans because of economic crises
and resultant out-migration.65 Illegitimacy experience appears to have
been too concentrated to be a result of generally experienced economic
factors alone.
The other significant pillar of the demographic systems in our proto-
industrial townships was mortality. Mortality levels, trends, and spatial
variations were dominated, as elsewhere, by the dynamics of infant
mortality.66 Once again the townships had contrasting experiences, as
figure 9 suggests. Infant mortality levels in Calverley were above the
average for rural areas in the ‘national’ sample, but well below the mean
of all 26 family reconstitutions. And while the township was roughly on
a par with the proto-industrial village of Gedling, it lagged well behind
infant death rates in Sowerby. Here, infant mortality rose very significantly
in the later eighteenth and early nineteenth centuries, reaching almost
200 per thousand. Of the 26 Cambridge Group reconstitutions, only
Birstall and Banbury approach these levels at this time. The significant
point about both townships, however, lies less in the level of infant
mortality than in the trend. Birstall, Gedling, Shepshed, Calverley, and
Sowerby experienced more or less consistent upward movements in infant
mortality during the rural industrialization process. This contrasts with
the rise in 1700-49 and consistent falls thereafter which can be seen in
the ‘national’ sample. The conclusion that proto-industrial areas ‘bucked’
the national mortality trend is reinforced by the results of our research.67
This finding contrasts with some of the continental literatures which
have stressed the positive relationship between rural industry, improving
incomes, health, and longevity, albeit in the nineteenth rather than the
eighteenth century.68 But, as is shown below, mortality experience had
is uncertain how births to parents in this sort of union would have been viewed. We cannot estimate
this directly, but in Calverley there was a slight increase during the eighteenth century in the number
of people for whom their own baptisms and the baptisms of their children, but no marriage, could
be traced. This may have reflected consensual unions given that all surrounding parish registers
were searched for this part of the study. In Sowerby there was a rather stronger increase in the
number of life-cycles of this sort. In common with other studies, we can simply note this phenomenon
rather than control for it. Gillis, World of their own making; Lemmings, ‘Marriage and the law’.
64 King, ‘Nature and causes’.
65 Levine, ‘Industrialization and the proletarian family’, p. 185; Tilly and Scott, ‘Women’s work
and family economy’.
66 Adult mortality levels are in any case very difficult to calculate with family reconstitution data
and involve the implementation of a number of assumptions about the timing of migration on the
part of those at risk. These issues are explored in King, ‘Profitable pursuits?’ and idem, ‘Historical
demography’. See also Ruggles, ‘Migration, marriage and mortality’.
67 Relatively high infant mortality in our townships, and more generally in proto-industrial areas,
also carried over into relatively heavy mortality in early childhood. This is reported in King,
‘Profitable pursuits?’.
68 Vardi, Land and loom; Vandenbroeke, ‘Proto-industry in Flanders’. On one level our mortality
rate findings endorse Wrigley and Schofield’s emphasis upon marriage and fertility as the engines
of population growth in the eighteenth century. There is, however, a contrary implication: if such
high proto-industrial mortality rates were repeated throughout most proto-industrial areas, life
Economic History Society 2000

two textile townships, c. 1660-1820
727
Paxton
Birstall
Shepsted
26 studies
Calverley
Sowerby
1800-24
1750-99
1700-49
1680-99
0
20
40
60
80
100
120
140
160
180
200
Rate per thousand
Figure 9. Calverley and Sowerby infant mortality in comparative perspective
Note: sample sizes: Calverley 649, Sowerby 1,491
Source: Family reconstitutions. For comparative material see Yorkshire Archaeological Society, ‘Family Reconsti-
tution Collection’; Wrigley et al., English population history; Don, ‘Reconstitution of Paxton’
much to do with kinship support networks, very localized environmental
conditions, variable access to food supplies, domestic arrangements, and
weaning habits. These factors, affecting restricted groups in the popu-
lation, appear to have been much more important than the nature of the
dominant economic sector alone in accounting for changes in mortality.69
Death rates in both Sowerby and Calverley increased over time, but
not because there was a significant increase in risk for the majority of
the population; in fact mortality experiences for most families were stable
or improving. What caused the identifiable increase in average mortality
levels was an increase in the number of families susceptible to multiple
infant deaths and an increased mortality incidence among susceptible
families. Thus in Calverley between 1680 and 1820, 330 families lost
649 infants but 319 (50 per cent) were in just 102 families with four
families experiencing six or more infant deaths. Background mortality
was thus of the order of 60-80 per thousand by 1800 while foreground
expectancy must have been improving very markedly indeed in areas elsewhere in order to bring
the national averages up sufficiently to indicate mild improvement. With rural to urban and
agricultural to proto-industrial migration, an increasing proportion of the national population over
the course of the eighteenth century were exposed to high and often deteriorating mortality regimes
which suggests that national averages might be disguising substantial levels of improvement in some
regions. Renewed research on regions, localities, or population subsets experiencing improving life
expectancy during industrialization is long overdue. See Razzell, ‘Growth of population’
69 Hudson and King, Industrialization.
Economic History Society 2000

728
pat hudson and steve king
mortality among the small number of high mortality families was as high
as 240.70 In Sowerby the figures are even more stark. Eight hundred and
sixty-nine families lost 1,491 infants but 150 families lost at least three
infants and 40 families lost six or more. As in Calverley, the numbers of
these high mortality families grew in the later eighteenth century over
and above the number which would have been expected simply because
of rising population. That this is not simply a matter of better detection
of deaths can be demonstrated by using the Moravian families of Fulneck
(in Calverley) as a control: there is a near exact measure of mortality
concentration in flawless registration data. The susceptibility of families
in both Sowerby and Calverley appears to have been largely unrelated to
family size: in the former, 8 per cent of all infant deaths were among
families with nine or more births but these families accounted for 17 per
cent of all allowable births during the eighteenth century. Similar obser-
vations are found in Calverley: even within family size groupings the
concentration is clear: between 1750 and 1820, 50 per cent of all infant
mortality in completed families of 10+ children was concentrated in just
9 per cent of all families of this size. For families with a completed size
of six children, 50 per cent of all deaths were concentrated in just 11
per cent of all families. Many families with eight and more children lost
none at all. For example, John Hainsworth, a Calverley clothier, and his
wife Hannah had nine children between 1804 and 1816 and lost only
one child (aged 8 months), in 1811. In Sowerby the background rate for
the last quarter of the eighteenth century was 105:1000 while the fore-
ground rate was 310, demonstrating that Sowerby experienced higher
infant mortality than Calverley across the range. The concentration of
high mortality in certain families is not a novel observation: it has been
found in Swedish and other continental data.71 Nor is it a feature which
emerged in our townships only in the eighteenth century: it simply
appears to have become more intense at that time. The finding is
significant because of the implications it has for analysing the causes and
consequences of infant death.
Thus all of the major demographic variables—illegitimacy, mortality
and to a lesser extent but still to a pronounced degree, fertility and
nuptiality—shifted at the aggregate level in our townships in the eighteenth
century because of proximate change in the number of families who were
experiencing extremes of behaviour and because there was an increasing
intensity of extreme experience. Such families had always been present
but, especially in the late eighteenth century, their experience came
disproportionately to shape aggregate patterns. This raises important
70 Isolating background and foreground mortality rates in these terms brings the acute problem
of statistical significance with small sample size. Thus, calculating a rate for background mortality
among those families in Calverley which experience few if any deaths in the period 1775-1800
would involve just 19 deaths. The tendency of demographers to concentrate on rates is ill suited to
this sort of analysis. Reviewing the plain numbers is sufficient indication of concentrated experience
to warrant closer investigation.
71 Knodel, Demographic behaviour; Brandstrom and Sundin, ‘Infant mortality’. For more on
Calverley see King, ‘Dying with style’.
Economic History Society 2000

two textile townships, c. 1660-1820
729
questions about the prevalence and impact of subgroup experiences more
generally, especially during the period of industrialization.
III
Changes in aggregate indicators appear to have resulted from fundamental
changes in the behaviour of finite, relatively small, subsets of the popu-
lation while the demographic activity of the majority remained roughly
stable and apparently undisturbed by economic transformations. If chang-
ing social, economic, and institutional forces were the most important
causal factors in population change, they appear to have been felt most
keenly by a small proportion of the township populations and/or by
individual birth and marriage cohorts. This paints a rather different gloss
on the whole issue of demographic motivation from that provided by
macro-theories of change or by theories which try to explain behaviour
and motivation in relation to particular typologies of communities such
as ‘the proto-industrial.’72
Before proceeding to offer some initial analysis of the observed concen-
trations of demographic experience, it is necessary to comment upon the
statistical significance of the level of clustering which we have found and
upon the extent to which different subgroups overlapped. It is difficult
to compare the degree of concentration of extreme demographic behaviour
in Sowerby and Calverley with other studies as these are few and no
reliable benchmark measures exist for ‘normal’ clustering. It is possible,
however, to get some clue by comparing infant mortality concentrations
for our townships with those found by Brandstrom for the Swedish town
of Jokkomok, which he saw to be related largely to genetic propensities.73
Taking Sowerby and Calverley combined, the infant mortality concen-
tration appears to have been at least 40 per cent higher than in Jokkomok.
An alternative approach to infant mortality clustering lies in the use of a
binomial probability distribution to allocate elements of an indicative
infant mortality rate to families of different sizes, allowing us to estimate
what proportion of families one would ‘normally’ expect to account for,
say, 50 per cent of all deaths. Using this approach and taking the infant
death rate for the two townships combined in the key period 1750-99,
we would expect 50 per cent of all infant deaths to be concentrated in
17 per cent of families rather than the observed concentration into 9 per
cent of families.74 There are fewer obvious ways of measuring the signifi-
cance of concentration found in the other demographic variables, but the
evidence suggests higher levels than would be expected on the basis of
inherent population characteristics (such as genetic propensities) alone.
72 As in Wrigley et al. English population history.
73 Brandstrom and Sundin, ‘Infant mortality’.
74 This method was advised by Jim Oeppen and we were assisted in making the necessary
calculations by Emmett Sullivan. Our thanks to each. As a rule of thumb, the higher the notional
death rate applied, the greater would be the expected concentration of mortality. The infant death
rate for the two townships combined for this period (1750-99) was 171 per thousand. Details of
these calculations are obtainable from KingȰBrookes.ac.uk
Economic History Society 2000

730
pat hudson and steve king
Marriage age
Marital fertility
Infant death
: Overlap between low marriage ages and high marital fertility
: Overlap between infant death concentration and high fertility/low marriage ages
: Overlap between all three variables
Figure 10. Venn diagram representing case study overlaps between female age
at first marriage, marital fertility (completed families), and high infant mortality
experience, 1700–1800
Note: cases: Female age of first marriage, 1817; fertility (completed families) 2,790; infant mortality 1,279. Case
numbers where all three variables overlap: 396. Case numbers of overlap between high infant death and low
marriage age but not marital fertility: 137. Case numbers of overlap of high marital fertility and high infant death
but not low marriage age: 179. Case numbers of overlap between low marriage age and high marital fertility but
not high infant death: 132
Source: Family reconstitutions
The degree of overlap between subgroups experiencing radical change
in different aspects of demographic experience is important for our study
but difficult to estimate precisely. Cohort evidence for both townships
highlights significant congruence between subgroups particularly regarding
PNP and high marital fertility. Of the group of women whose consistently
low birth intervals served to push up aggregate fertility levels in Sowerby,
42 per cent had experienced PNP. Case studies provide additional evi-
dence of a similar high degree of overlap in both townships between
female age at first marriage, fertility (completed families), and infant
mortality, as illustrated in figure 10. It appears that between a quarter
and one-fifth of the population in each township experienced marked
change in two or more aspects of demographic experience and that
this restricted group was responsible for shifts observable in aggregate
vital rates.
We thus need to explain the marked shifts in demographic experience
Economic History Society 2000

two textile townships, c. 1660-1820
731
of small, identifiable, and largely overlapping subgroups. At the same
time, we must also suggest why background demographic variables
remained relatively stable in the townships. Extreme shifts require atten-
tion because of their impact upon arithmetical averages and measures of
dispersal which are so much the focus of existing causal analyses. Change
in marriage behaviour is a good exemplar here. By looking at occupational
variations, it is possible to obtain clues to understanding the shifting core
of extreme marriage behaviour which resulted in the emergence of a
subset of young marriers in Sowerby, and the increasing rarity of late
first marriage. Demographic patterns in Sowerby were disproportionately
influenced by just one occupational group, weavers, although there were
important divisions of experience within the group. Children from the
wealthiest and most prosperous textile families appear to have married
consistently late in the eighteenth century.75 Those from proletarianized
families married consistently early. Falling female marriage ages in Sow-
erby are largely to be explained by cumulative changes in the marriage
behaviour of children from increasingly proletarianized weaving families,
especially during the initial stages of commercial worsted expansion after
the 1760s (table 2b). The growing core of early marriers may thus have
been precipitated by positive, but often insecure, female proletarianization
whilst the marriage ages of the bulk of young people in the community
remained fairly stable.
In Calverley, clothier families dominated marriage patterns although,
again, there were wide differences of experience. Males and females
falling down the occupational scale at marriage appear to have been
among the earliest marriers in the township.76 There was no consistent
analogue in higher marriage ages among those whose fortunes were stable
or rising, except in the experience of the children of migrants and those
marrying husbandmen. In-migrants married slightly later than the mean
and were more likely to experience upward mobility at marriage. The
female children of farming families appear to have married very early
indeed, but the sons of farmers and the brides of farmers, who were not
themselves drawn from farming families, appear to have married late
(table 2a).77 These extremes suggest that the same dynamic of prolet-
arianization may have been at work in Calverley as in Sowerby but the
process and its demographic impact appear more sluggish. It is likely to
have been slowed and modified by the family-unit structure of the
industry which could absorb some of the surplus proletarianized labour,
especially young women, and, perhaps, more importantly, by the ease of
migration into Leeds. The impact of proletarianization on marriage age
is also likely to have been diminished by the existing norm of relatively
75 This was assessed by comparing the families concerned with the size of poor rate payments
and landholdings. More detailed investigation of these connections appears in Hudson and King,
Industrialization.
76 For this analysis, occupational ascription in any source within four years of marriage was used.
Our assessment of what constituted ‘falling down the occupational scale’ is based on our study of
occupational and earnings hierarchies in the two townships.
77 For much more detail on this see King, ‘Nature and causes’.
Economic History Society 2000

Table 2. Female age at first marriage according to occupation of father and groom in Calverley and Sowerby
(a) Calverley
Groom 1700-49
Father
Groom 1750-99
Waged
other
Service
Waged
textile
Craft
Farmer
Clothier
Clothier Farmer
Craft
Waged
textile
Service
Waged
other
22.4
24.9
n.a.
22.3
n.a.
22
23.6
23.2
22.0
n.a.
23.4
23.6
30
23.0
25.0
24.1
21.9
24.0
21.7
20
24.0
23.5
22.4
23.5
22.5
n.a.
24
23.0
22.0
23.2
n.a.
23.2
n.a.
39
22.4
24.6
22.8
22.9
22.2
22.9
110
Clothier
Farmer
Craft
Waged textile
Service
Waged other
Cases
22.6
24.9
22.3
22.3
23.0
22.8
201
23.2
22.4
23.0
n.a.
22.4
n.a.
60
23.6
23.0
22.1
n.a.
22.9
n.a.
33
22.8
n.a.
24.0
21.9
22.0
21.7
37
23.0
23.6
22.9
23.6
23.0
n.a.
21
21.8
n.a.
n.a.
20.9
n.a.
21.7
23
19
(b) Sowerby
Groom 1700-49
Father
Groom 1750-99
Service
No
occupation
Waged
other
Service
Craft
Farming Weaving
Weaving Farming
Craft
Waged
other
No
occupation
24.3
25.6
24.8
22.6
21.8
24.6
40
23.7
—
25.3
23.7
22.1
22.9
22
24.2
24.0
26.0
23.0
—
—
47
24.0
25.6
26.0
24.0
22.0
23.0
43
24.2
26.0
27.0
22.8
—
24.0
21
23.7
25.9
25.2
23.0
21.6
23.4
140
Weaving
Farming
Craft
23.3
25.5
25.4
22.8
20.6
23.4
259
25.0
25.8
26.5
25.0
—
26.0
27
23.9
25.6
26.1
23.7
—
24.0
48
23.0
24.2
25.0
23.0
22.7
22.3
63
22.1
—
—
22.0
22.4
21.0
111
21.9
25.0
25.2
21.2
23.6
21.6
53
Service
Waged other
No occupation
Cases
Source: Family reconstitutions

two textile townships, c. 1660-1820
733
Table 3. The density of kinship in Calverley and Sowerby
Sowerby
Calverley
1750-1799
1700-1749
Density of kinship
1700-1749 1750-1799
30
30
20
20
40
28
16
16
Related to no other families
Related to 1 other family
Related to 2 other families
Related to 3 or more families
34
29
17
20
29
28
24
20
Source: Family reconstitutions. Figures indicate the percentage of families in the townships who are related to 0,
1, 2, and 3 or more other families within the same township either by blood or by marriage. Both vertical and
horizontal connections are included. The kinship network is curtailed at the level of first cousins. The figures are
based upon 19,000 discrete kinship relations.
early marriage. This may explain why it was the subset of late marriers
whose behaviour was most affected. Young women made proactive
choices to marry or to migrate: it was increasingly unattractive for single
and proletarianized women to remain in the township.
That female marriage ages were more variable than male in both
townships may be explained by the greater economic insecurities of
young women whose wage-earning opportunities in both agriculture and
domestic manufacturing were being squeezed in the later eighteenth
century. Marriage as an economic necessity for a proletarianized subgroup
of women was further endorsed by the low levels of poor relief provision
in both townships and by relief policies which placed single able-bodied
women as the lowest priority.78 Denser networks of kinship in Calverley
than those found in Sowerby (table 3) may have provided some support
for adolescent girls and may thus have protected some from the economic
necessity of very early marriage.79 In both townships, however, lowering
of the age of marriage, declining celibacy, and rising illegitimacy may all
be largely explained by the reactions of those most affected by economic
necessity and the behaviour of proletarianized, economically marginal
young women appears to have been crucial in determining extremes of
marital experience.
To some extent the restriction of significant shifts in fertility to an
identifiable subgroup whose demographic experience was changing in
other ways springs from two well-known demographic interlinkages. First,
as one would expect, there was a feedback between high infant mortality
and short birth intervals. Secondly, and more significantly, those who
were pre-nuptially pregnant were predictably highly represented in the
high fertility subgroup. However, these two links are only proximate and
partial elements of explanation. The high fertility subgroup may have
been a result of biological variation in fertility (including susceptibility to
still births), or of greater frequency of penetrative and/or premarital sex
for certain couples. That the subset of short birth interval mothers
78 Idem, ‘Reconstructing lives’; idem, ‘Nature and causes’; Sowerby settlement examinations and
certificates, SPL 94.
79 Poor relief policies and the positive role of kinship support in Calverley are fully explored in
King, ‘Reconstructing lives’.
Economic History Society 2000

734
pat hudson and steve king
increased in the later eighteenth century is also compatible with changes
in breast feeding which are likely to have accompanied changes in
women’s work. The growing self-exploitation of the family economy and
the move to workshops and some factories for spinning, which separated
the home from the workplace, probably made breast feeding more difficult
and early weaning essential. This might explain the overlap between high
fertility and high mortality subgroups. Additionally, a decline in the
number of stillbirths may have played a significant role in generating
higher fertility in certain families though not as great as that suggested
by Wrigley for the country as a whole. For the two townships combined,
endogenous infant mortality (death in the first month of life and hence
largely a result of poor maternal nutrition or genetic factors) fell from
96 per thousand births (live and dead) in 1700-24, to 64 in 1800-24
which compares roughly with Wrigley’s national figures of 100-125 per
thousand at the end of the seventeenth century and 40 per thousand by
the early nineteenth century.80 Finally, high and rising fertility was most
strongly associated with those weaving and clothier families who were
not major land or wealth holders and who did not have strong kinship
links. A disproportionate number were also first-generation migrants. This
suggests the importance of a range of economic factors in encouraging
high reproduction rates in certain population subgroups, with the impact
of proletarianization featuring very strongly.81
These sorts of influences also appear to have been important in ident-
ifying the high mortality subgroup. High and deteriorating mortality levels
in our townships may potentially be explained by the usually invoked
effects of industrialization: trade fluctuations, job insecurity, work con-
ditions, poor public health and hygiene, increased exposure and reduced
resistance to disease, or by poor childcare practices. But if these general
factors were important why was their impact so unevenly distributed?
The high infant and child mortality experiences of certain families were
linked to low birth intervals and hence to high fertility, as we have seen.82
More significantly, the clustering of high infant deaths in certain families
in both townships appears to have been closely linked to economic and
geographical position. Case studies show a close association between high
infant mortality, illegitimacy, economic marginality, migrant status, and
the absence of local kinship. Lack of kinship networks, particularly asso-
ciated with inmigrants, appears to have been a more important factor
than occupation, status, or other indications of economic marginality per
se. Just as important, in Calverley in particular, was the association
80 This calculation was made after eliminating dummy variables from the analysis, making the
figures compatible with Wrigley’s national-level estimates for endogenous infant mortality: Wrigley,
‘Explaining the rise in marital fertility’. Elimination of the dummy variables does, however, reduce
sample size for our township study and makes the exercise and the comparison somewhat question-
able.
81 Calverley and Sowerby, provisional reconstitution results.
82 Mortality extremes also appear to have been clustered in families sufficiently strongly to suggest
the sort of genetic explanation identified by Brandstrom and Sundin, ‘Infant mortality’. Wrigley
suggests that genetic factors in infant mortality experience became less important in the second half
of the eighteenth century in England, which may be reflected in our figures.
Economic History Society 2000

two textile townships, c. 1660-1820
735
between very poor mortality experience and residence in the low-lying
and damper riverine areas of the township which were near to the main
thoroughfares and thus more susceptible to the spread of infections.83
The ‘artificiality of the conventional separation of demographic behav-
iour into fertility, mortality and nuptiality’84 is highlighted by these
findings, in particular by the stress which we place on the impact of
overlapping sub-groups of extreme experience. A key question regarding
major shifts in demographic behaviour must be how proletarianization
and associated population mobility impacted differentially upon men,
women, and families in terms of job security, employment on the margins
of waged work, the domestic role, child nurturing, and the pressures
and time constraints upon women. The proactive responses to such
circumstances on the part of families and individuals, in particular young
women, also appear to lie at the heart of demographic changes in the
two townships.
The relative stability of background variables of nuptiality, fertility,
illegitimacy, and mortality in Sowerby and Calverley is as important as
the changes and also requires explanation. Rural industry enabled more
people to live out their lives in the community rather than migrate away
and hence both to have larger and stronger kinship networks and, other
things being equal, to be influenced more by communal norms and
practices than might otherwise have been the case. Table 3 which sets
out estimates of kinship density over time for both places suggests
that demographic life-cycles, processes of proletarianization, courtship,
experiences of poverty, and continuity and change in intergenerational
relationships were played out against a backdrop where individualized
decision making, so central to ideas about the cultural impact of mod-
ernization, capitalism, and proto-industry may have been subject to an
increasingly powerful variety of familial and social constraints.85 This
factor in itself may have endorsed the relative stability of demographic
behaviour among the majority of township residents despite economic
transformations and increased in- and out-migration.86
In addition, although Seccombe has made an important contribution
to our understanding of demographic change by distinguishing between
positive and negative proletarianization, we contend that proletarianization
can also be seen to act in a neutral manner.87 Where a drift to landlessness
and wage earning is a function of demographic increase itself and is
accompanied by a growth in manufacturing earning opportunities, it
becomes significant in stabilizing rather than changing the economic and
83 King, ‘Dying with style’; Hudson and King, ‘Industrialising townships’; King, ‘Profitable pur-
suits?’.
84 Wrigley, ‘Explaining the rise in marital fertility’, p. 460.
85 See Hudson and King, ‘A sense of place’ and, for an exploration of these factors in a different
context, King, ‘Chance encounters’.
86 The increased residential stability of core groups in the populations of industrial regions in the
later eighteenth century was noted by Deane and Cole, British economic growth, pp. 106-22. The
industrial West Riding as a whole experienced relatively little in- and out-migration between 1750
and the early nineteenth century.
87 Seccombe, Millennium of family change.
Economic History Society 2000