Full text of DOI:10.1111/1468-0289.00221

Economic History Review, LV, 2 (2002), pp. 262–298
English rural societies and
geographical marital endogamy,
1700-1837
By K. D. M. SNELL1
‘Better wed over the mixen, than over the moor.’2
‘The young man who left his village or town to seek a bride was considered
to be nothing more than an outlaw against the rules of decency.’3
t is now often argued that nuptiality and fertility were the main
I
determinants of demographic trends in England between c. 1538 and
1837.4 England’s leading historical demographer has written that ‘Mar-
riage could be said to have been in the past the great valve which regulated
population change.’5 Given the crucial demographic roles, respectively, of
female age at first marriage and of the proportions of women marrying,
much research has come to focus upon these, and upon the determinants,
customs, and social history of marriage. This has produced an increasingly
large literature and a thriving field of historical investigation. However,
as yet little attention has been paid to parochial geographical endogamy
and exogamy in the English past, even though these spatial aspects of
marital behaviour have considerable cultural and social importance, are
relevant to some key issues within anthropology, and clearly bear upon
English historical demography.6 This neglect in the subject literature
prevails despite the abundance of relevant information about residence
upon marriage in parish registers, alongside the much used recording of
vital events.
A further reason for pursuing such study is that parochial geographical
endogamy connects historical demography with research into local cultural
regions and their distinctiveness. Some areas of the country were notice-
ably more endogamous in their behaviour than others. The reasons for
this remain to be explored, but might include population size or density,
1 I am most grateful to Ros Davies, Brian Outhwaite, Alan Macfarlane, Charles Phythian-Adams,
Barry Reay, Simon Szreter, Tony Wrigley, and to anonymous referees for their helpful advice.
2 Grose, Provincial glossary, n.p., under ‘Cheshire’.
3 Findler, Folklore of the lake counties , p. 11.
4 The overwhelming influence here has, of course, been that of the Cambridge Group for the
History of Population and Social Structure. See in particular Wrigley and Schofield, Population
history of England; and Wrigley et al., English population history.
5 Wrigley, ‘Small-scale but not parochial’, p. 29.
6 My motivating interests here relate primarily to rural and community history. However, the
findings have implications for methods and arguments within historical demography, and these are
raised more fully in a demographic afterword, after the basic trends in parochial geographical
endogamy have been established.
 Economic History Society 2002. Published by Blackwell Publishers, 108 Cowley Road, Oxford OX4 1JF, UK and 350 Main Street, Malden,
MA 02148, USA.

english rural societies and marital endogamy
263
the sizes of parishes, the nature of regional topography, local xenophobia
or custom, or poor-law administration and the application of the settle-
ment laws. Both the significance of this subject for the persistence of
local cultures and the demographic and economic importance of such
cultural variation remain unappreciated. It is frequently said, for example,
that there was a custom in England for weddings to take place in the
wife’s parish (‘uxorilocal’ weddings), but for her then to move to her
husband’s parish.7 Yet almost no research has measured the extent of
this, or changes in it. Nor has there been consideration of the regional
variety that underlay it, or of the alternative practice of ‘virilocal’ wed-
dings.8
Analysis of geographical endogamy or exogamy also brings one to local
study of migration, which has flourished in recent decades, using a great
variety of sources. Included in such study has been analysis of ‘marriage
horizons’, or distances between spouses’ residences. The results have
underlined the predominantly local nature of marriage. In particular,
considerable light has been shed on marital horizons after 1837, during
the most crucial period of nineteenth-century rural depopulation. There
have been pioneering articles by Peel, by Constant, and by Perry.9 The
pattern of change they uncovered, while identified for certain districts
only, was one in which often high levels of parochial endogamy and
restricted marriage horizons declined rapidly from some point between
about 1830 and 1880.10 This was closely connected with out-migration
from the land. Constant found that in an adjoining area of Northampton-
shire and Huntingdonshire, in the century after 1754, the average distance
between spouses was very small indeed, being only a few miles. Marriage
distances grew from about 1850, a consequence in particular of the
railway, which re-oriented the outlook of parish inhabitants. He noted
the continued effects of physical barriers and ‘local prejudice’ as influenc-
ing marriage networks. He also indicated that marriage distances may
enlarge or contract over time, and that there is no reason to suppose a
linear historical increase in such horizons.11
In an important article covering 27 west Dorset parishes, between 1837
and 1936, Perry revealed that change away from localism and high
parochial endogamy occurred strikingly among the labouring poor from
7 See, for example, Wrigley, ‘Clandestine marriage in Tetbury’; idem, ‘Life-time mobility of married
women’; idem, ‘Effect of migration on estimation of marriage age’ pp. 88, 95, 96; Strathern, Kinship
at the core, p. 237; Robin, Elmdon, p. 27.
8 For this discussion, I shall use ‘marriage’ and ‘wedding’ as synonymous terms. With regard to
the terms ‘uxorilocal’, ‘virilocal’, ‘endogamy’, and ‘exogamy’, I am well aware of slightly different
(and often varying) usage in the anthropological literature, and I ask for forbearance among
anthropologists for my terminological reapplications here. A brief summary of changing usage of the
term ‘exogamy’ within anthropology may be found in Wolfram, In-laws and outlaws, pp. 169-80.
9 Peel, ‘Local intermarriage’, pp. 22-30; Constant, ‘Geographical background’, pp. 78-88; Perry,
‘Working-class isolation’, pp. 115-35. See also Maltby, ‘Easingwold marriage horizons’, pp. 113-14,
on localized marriage patterns; Mills and Mills, ‘Rural mobility in the Victorian censuses’.
10 Or see Strathern, Kinship at the core, p. 293, n. 41, on the decline in endogamous (Elmdon-
Elmdon) marriages registered in Elmdon church. These comprised 84% of all marriages in 1813-
22, falling to 74% (1843-52), 50% (1873-82), 19% (1943-52), and 12% (1953-62).
11 Constant, ‘Geographical background’, pp. 80-5.
 Economic History Society 2002

264
k. d. m. snell
the mid-1880s.12 He excluded many better-off classes to allow concen-
tration upon the labouring population, and was thus able to show the
decline of their ‘isolation’ within the countryside from about that time.
This resulted from the effects of broader education and literacy, extra-
parish school friendships, declining population and so lesser self-contain-
ment of villages, rising living standards, falling working hours, the railway,
the widening of courtship areas (beyond about four miles) after the
coming of the affordable Stanley safety bicycle (from 1885), and the
erosion of some features of local custom and prejudice. Although he
omitted from his analysis higher classes and occupations (whose average
marriage distances would have been greater than those for the labouring
poor), a memorable feature of his study was how very isolated and
endogamous the rural poor were prior to the 1880s. It would hardly be
possible to find higher levels of endogamy than those obtaining at the
start of his study—about 80 per cent of marriages then being intra-
parochial. This raised previously unasked but important questions about
whether those high levels had been long established prior to 1837.
It is worth highlighting additional findings from the few other related
studies in this area. Millard, analysing marriages for Stony Stratford
(Buckinghamshire), discovered that there was an ‘apparent retrenchment
of marriage horizons’ between 1794 and 1833.13 He felt that this might
be related to falling real wages and to aspects of poor-law administration.
A similar explanation, involving the settlement and poor laws, was
advanced by Eversley, when he pointed to evidence for rising parochial
endogamy in the eighteenth century.14 In his study of Bickenhill
(Warwickshire), Skipp argued for a shift towards very largely local mar-
riages in the period 1758-1847, away from marriages involving greater
distances between spouses which had been common earlier, when many
spouses were from such places as Coventry, Birmingham, Solihull, and
Coleshill.15 In an interesting study of 17 Shropshire parishes, Sogner
found that a tendency for both partners to come from the same parish
became more pronounced over 1711-60, most notably between 1711-20
and 1751-60.16 Hunter’s work on the estate village of Haynes
(Bedfordshire) showed extra-parochial marriages falling considerably as a
percentage of all marriages between 1754 and 1833, with few changes
in their distances.17 Yasumoto uncovered something similar in Methley
12 Perry, ‘Working-class isolation’, pp. 130-3.
13 Millard, ‘New approach’, p. 161.
14 Eversley, ‘Exploitation of Anglican parish registers’, p. 64. ‘There is little’, he added, ‘in the
English sociological literature about the pattern of marriages in the geographical or sociological
sense’ (ibid., p. 94, n. 42), a point that still remains valid. Eversley, ‘Survey of population’, analysed
15 parishes around Bromsgrove in north Worcestershire, and was able ‘to state quite clearly that
marriages between partners of different parishes were actually more frequent in the eighteenth
century than in the early nineteenth century’. (p. 412). His figures after c.1800 gave much higher
percentages of ‘home’ marriages than hitherto, especially in large parishes. He argued that ‘Greater
freedom of movement cannot therefore have contributed to increase of population by providing
more scope for marriage’: ibid., p. 413.
15 Skipp, Discovering Bickenhill, p. 39.
16 Sogner, ‘Aspects of the demographic situation’, pp. 132-3.
17 Hunter, ‘Marriage horizons and seasonality’, pp. 38-40.
 Economic History Society 2002

english rural societies and marital endogamy
265
and York, and there were hints of this in Estabrook’s work on Bristol
and nearby Olveston.18 The evidence for changes in this direction—
towards increasing geographical endogamy after the mid-eighteenth cen-
tury—is as yet scanty, but it has been building up in a surprising and
suggestive way. This is most tantalizing, for if these hints from local
studies were more widely applicable, they would demonstrate an accentu-
ation of parochial endogamy during English industrialization when few
historians would expect to find this occurring.
I
As this distillation of earlier findings suggests, my concern here is with
geographical marital endogamy and exogamy, using the historical parish
as the basis of study. In other words, I am not considering endogamy
within status groups (so-called ‘homogamy’), or within occupations,19 or
kinship endogamy as studied by many anthropologists. Nor am I consider-
ing ‘marriage horizons’ as conventionally measured, since it is so usual
to find short distances between spouses’ residences (especially among the
labouring poor).20 Given the fundamental importance of the parish as
the local unit of administration until at least the new poor law (and of
its northern township equivalents in many cases), it is worth concentrating
upon it as the unit of analysis here. This importance of the parish is
universally recognized in historiography dealing with such diverse subjects
as the Anglican Church, registration of vital events, boundary mainte-
nance, highways, enclosure, local charities, the poor law and legal settle-
ment, or many facets of folklore, the cultural calendar, and local politics.
The centrality of the parish to local life and identity is indisputable. Such
a stress here has the benefit of relying upon a local sense of place that
was subjectively and administratively very real indeed to contemporaries.
It will serve us best for the analytical purpose in hand.
To explore these issues more extensively, I analysed 18,442 marriages
from the published registers of 69 parishes, scattered across eight coun-
ties.21 These comprise rural parishes in Derbyshire, Dorset, Lancashire,
Leicestershire, Norfolk, Northumberland, Oxfordshire, and Sussex. They
are listed in appendix A, alongside the documented years used for
18 Yasumoto, Industrialisation, urbanisation and demographic change, pp. 72-3; Estabrook, Urbane
and rustic England, pp. 94, 97, 99.
19 Compare the interesting discussion of high endogamy within ‘proto-industrial’ and some other
occupations during industrialization, argued for by Gillis, For better, for worse, pp. 116-18: ‘The same
tendency towards endogamy was evident everywhere domestic industry, mining, or fishing prevailed.’
20 See for example Sogner, ‘Aspects of the demographic situation’, pp. 131-3; Skipp, Discovering
Bickenhill, pp. 39-40; Maltby, ‘Easingwold marriage horizons’, pp. 113-16; J. R. Cole, ‘Marriage
horizons in east Kent, 1620-1640’ (unpub. essay deposited in the Centre for English Local History,
Univ. of Leicester, 1977); Millard, ‘New approach’; Harrison and Boyce, ‘Migration, exchange, and
genetic structure’, p. 133.
21 The places studied are listed in appendix A. Calculations were conducted on a parochial basis
from Anglican registers. A very large majority of these places are parishes. In a few cases, the place
was a township or chapelry within a parish, having its own Anglican register. To ensure consistency,
calculations for such places were still made on a parochial rather than a township basis, a township
register being treated here as a geographically unrepresentative sample of all marriages in that parish.
 Economic History Society 2002

266
k. d. m. snell
Figure 1. Places providing registers
Sources: See text and app. A
calculations. Their locations are shown in figure 1, from which it will be
apparent that I have picked clusters of parishes in disparate English
regions, to check for national homogeneity, or to allow any possible
regional patterns to emerge. I chose rural parishes rather than larger
market towns so as to dilate earlier historiographical findings, and to give
 Economic History Society 2002

english rural societies and marital endogamy
267
this study a rural coherence which it would lack if a fuller range of
parishes across the whole rural-urban spectrum was used. However, a
diversity of rural settlements in very differing regions was deliberately
chosen, from places such as Snoring Parva in Norfolk or Wootton
Fitzpaine in Dorset, to Urswick in Lancashire or Edlingham in Nor-
thumberland. Considerable variation in population size and acreage
existed across such places, their 1841 populations for example ranging
between 59 and 1,433. With regard to the religious complexion of these
places, and its possible influences upon courtship, marriage or registration,
58 per cent of them had a Nonconformist or Catholic chapel at some
point during the nineteenth century. This is close to the figure that one
might expect (51 per cent) from analysing such English rural parishes in
the 1851 Census of Religious Worship;22 and allowing for some additional
chapels documented in the other and later sources consulted, it is very
closely representative. There were no Quaker meeting houses, with the
potential to register marriages separately. Given the expansion of ‘old
dissent’ from the mid-eighteenth century, only a few old dissenting places
of worship could have existed in these places a century earlier. Religious
dissent would have had little effect upon marriage registration, especially
after Lord Hardwicke’s 1753 Act. Nonconformity mainly affected baptism
and burial registration.23 Further tests suggest that the cultural reper-
cussions of Nonconformity had little influence on my calculations of
endogamy.24 I started this study from 1700, because some of the parish
registers provided inadequate documentation of residence in the later
22 The latter ‘expected’ figure is calculated from 1,726 English rural parishes with populations
under 1,500 in 1851, using the enumerators’ returns to the 1851 census of religious worship. These
data are outlined in Snell and Ell, Rival Jerusalems. After our period of 1700-1837, a range of
sources (the census of religious worship, nineteenth-century county directories, and the 1874 edition
of Wilson’s Imperial gazetteer) showed a Catholic chapel (or school) in Eastwell, Melling, and Pyrton;
an Independent chapel in Ardingly, Cocking, South Creake, Piddletown, Rampisham, and Stoke
Abbot; a Baptist chapel in Knipton, Morley, Gilmorton, Queniborough, Smalley, Standlake, and
East Stower; a Wesleyan Methodist chapel in Bothal, Breadsall, Chickerell, Downham, Elvaston,
Hanborough, Holme Hale, Houghton on the Hill, Morley, Pennington, Piddletown, East Stower,
Swithland, Syderstone, Thurmaston, and Whalton; and a Primitive Methodist chapel in East Bar-
sham, Chickerell, South Creake, Helhoughton, Holme by the Sea, Holme Hale, Horningtoft,
Northmoor, Pyrton, Queniborough, Runham, Snoring Magna, Snoring Parva, Standlake, and Syder-
stone. There were also unspecified dissenting chapels at Birstall and Stoke Abbot. Most of these
were built after 1801, and many of them after 1837 (when my period of analysis ends), although
their later building may hint at earlier congregations.
23 From the operation of that act in 1754, until the start of civil registration of marriages in 1837,
‘the Church of England had a monopoly of legally valid marriages and a legal duty to administer
them, with the exception of the marriages of Jews, Quakers, and members of the royal family’.
Wrigley and Schofield, Population history of England, pp. 29, 73. See also Wrigley et al., English
population history, pp. 87-91; Wrigley, ‘Clandestine marriage in Tetbury’, p. 20.
24 My interest here relates not to the adequacy of marriage registration, but to the possibility that
some Nonconformists had different senses of place from Anglicans. Methodist circuits, for example,
traversed the Anglican (or non-evangelically led) parish, and the various layers of Methodist organiza-
tion could open up different prospects of potential marriage partners. This might separate some
Nonconformist courtship and marriage horizons from those of Anglicans. Such a view should be
borne in mind, and might be expected to conduce to rising exogamy over time, as Nonconformity
expanded. However, removing the parishes with the most nineteenth-century dissenting chapels had
little effect on the county trends to be shown here, and the parishes where Nonconformity was
strongest showed similar trends in endogamy to other parishes. This inclines me to believe that
religious dissent is not an important issue for these marriage calculations.
 Economic History Society 2002

268
k. d. m. snell
seventeenth century.25 Given the introduction of civil registration in 1837,
and the work done by others on the post-1837 period, I ended the
research at that date, my main interest here being in the lesser studied
eighteenth and early nineteenth centuries.
It is necessary initially to clarify what was meant in parish registers
when a person was said to be ‘of’ this, or another, parish. It was, of
course, a very common statement indeed, becoming general after Hard-
wicke’s Act, although widely found before then as well. The phrase ‘of
this parish’ is found on countless gravestones too, although changing in
frequency over time, a phenomenon that I analyse elsewhere.26 The
meaning of this information in registers has often seemed unclear. His-
torians have wondered whether ‘of this parish’ meant that the person
was legally settled there, or was ‘from’ there in some loosely understood
sense, or was born there, had parents or a guardian resident there, or
was resident there in the short or long term.27 One could, of course,
‘belong’ to a place (or to different places) in manifold ways, and personal
ambivalence over ‘belonging’ in the past was probably more complex
even than is true today, particularly given the permutations of legal
settlement, ‘irremovability’ status before and after the nineteenth-century
‘Irremovability’ Acts, and related legal and customary entitlements. How-
ever, Hardwicke’s Act (1753) and contemporary discussions were very
clear about what ‘of this parish’ meant, and it is worth dwelling upon
this meaning.
Hardwicke’s Act laid down that
all banns of matrimony shall be published in an audible manner in the parish
church, or in some public chapel. . .of or belonging to such parish or chapelry
wherein the persons to be married shall dwell. . .upon three Sundays preceding
the solemnization of marriage, during the time of morning service, or of
evening service. . .and whensoever it shall happen that the persons to be
married shall dwell in different parishes or chapelries, the banns shall in like
manner be published in the church or chapel belonging to such parish or
chapelry wherein each of the said persons shall dwell%the marriage shall be
solemnized in one of the parish churches or chapels where such banns have
been published, and in no other place whatsoever.28
In such cases of dual parishes, the act was neutral on which parish
should hold the wedding. No clergyman was obliged to publish banns
unless the persons to be married, seven days before such publication of
banns, delivered to him notice in writing stating ‘the house or houses of
their respective abodes within such parish, chapelry or extraparochial
25 In some cases, this inadequacy continued into the early eighteenth century. Yearly data were
analysed manually on plots distinguishing the four possible types of marriage, and in no case did I
begin analysis of a register until I was confident that information was being supplied on this matter.
Some sub-periods have been omitted for the same reason. (See below, n. 52.) The years used are
shown in appendix A.
26 Snell, ‘Gravestones, belonging and local attachment in England, 1700-2000’ (unpubd. paper).
27 Among other discussions, see Pain and Smith, ‘Do marriage horizons accurately measure
migration?’.
28 An Act for the Better Preventing of Clandestine Marriages, 26 Geo. II, c. 33, s. I (italics
in original).
 Economic History Society 2002

english rural societies and marital endogamy
269
place as aforesaid, and of the time during which they have dwelt,
inhabited or lodged in such house or houses respectively’.29 With regard
to residence, this therefore implied a week’s notice plus three weeks for
the calling of banns. Further, no licence for marriage was to be granted
to solemnize any marriage in any other church or chapel than that
belonging to the parish or chapelry ‘within which the usual place of
abode of one of the persons to be married shall have been for the space
of four weeks immediately before the granting of such licence’. There
was further reference in the act to ‘the usual place of abode of one of
the parties’.30 The wording ‘of this Parish’, or ‘of the Parish’ of. . ., was
specified as the form for the entry.31
Throughout Hardwicke’s Act, it was abundantly clear that residency
was meant. No alternative was even hinted at.32 This should not surprise
us, for such intent was common long before 1753. Canons 62 and 102
concerning regular church marriage had stated that marriage should be
in the parish of residence of the bride.33 Marriage was also supposed to
be in place of residence under marriage licences.34 Reforming efforts by
the lower clergy in Convocation in 1714-15 had included an attempt to
ensure that licences should allow marriage only in the place of residence
of one of the couple, and efforts to bring this about had continued
thereafter.35 Such efforts may not have been very successful,36 but the
point here is that the presumed option had long involved residence, not
any other concept such as birth, or place of legal settlement under the
post-1662 settlement laws. From 25 March 1754, and allowing for
occasional evasion of Hardwicke’s Act or local circumstances such as the
temporary closure of a church for repair, queries over the status of a
chapelry, or the handling of extra-parochial places,37 there can generally
speaking be little further doubt about what was intended.
Some contemporaries added comments to parish register entries to
clarify their own meaning. It is possible to find clear explanatory phrases
29 Ibid., s. II.
30 Ibid., s. X. The other stipulations of the act (e.g. about under-age marriages, and the marriages
of Quakers and Jews) do not bear on this issue. Hardwicke’s Act became operative from 25 March
1754. It brought about a situation in which the residency requirements for marriage were tighter
than those for baptism or burial.
31 Ibid., s. XV.
32 For excellent discussion of residency requirements and practice under the act, see Outhwaite,
Clandestine marriage, pp. 80-4, 87, 98, 113, 116, 135-7.
33 Stone, Road to divorce, pp. 66, 97; Outhwaite, Clandestine marriage, pp. 8-9; idem, ‘Sweetapple
of Fledborough’, pp. 40-2.
34 Outhwaite, Clandestine marriage, p. 9. In some places, many licences broke the rules about
residency. See ibid., pp. 35, 47-8, 66, and idem, ‘Sweetapple of Fledborough’, pp. 40-2.
35 Stone, Road to divorce, pp. 110, 116-17, 121. On problems affecting the residency stipulations
of the 1753 act, especially in city parishes, see ibid., pp. 129, 135. On residency and marriage in
this pre-1754 period, see Outhwaite, Clandestine marriage, pp. 9, 11-12, 15-16.
36 Stone, Road to divorce, p. 100, and see below, fig. 2.
37 Hardwicke’s Act (s. I) laid down that when both or either of the persons to be married ‘shall
dwell’ in an extra-parochial place, lacking a church or chapel where banns were normally published,
then the banns were to be published in a church or chapel for a parish or chapelry adjoining the
extra-parochial place, and the clergyman publishing the banns should proceed as if either of the
persons to be married ‘dwelt’ in that adjoining parish.
 Economic History Society 2002

270
k. d. m. snell
such as ‘himself of this parish by residence’,38 or ‘residing in this parish’,39
or ‘both servants in this parish’,40 or ‘one of our antient and prinple [sic]
farmers’,41 or ‘dter to the antient race of that name here’,42 ‘of an old
race here’,43 or ‘native of Norton’.44 Specific mention of residency, which
is found before 1754 as well as afterwards, indicates an understanding
similar to that found in Hardwicke’s Act, and it is probable that in the
other examples quoted above the person was also resident.
In a few cases an entry can be more ambiguous, as in the case of
‘George Turner, a mariner, an inhabitant of no parish’,45 or ‘William
Hutton, a travlar, of no parish’.46 Examples such as these can perhaps
be taken as not locally resident. Thomas Sutton was said to be ‘of
Ashborne, but now of Kedleston’.47 Samuel Merry was ‘residing in this
parish, but of the parish of Kidlington’.48 William Wood was ‘of Hugnas-
ton, residing in this parish’.49 In these cases residence seems apparent,
but something else is being alluded to. In some instances, clearer allusions
were made to legal settlement status, and this was most probably the
case where a person was referred to as ‘a sojourner’, or the couple were
said to be ‘both sojourners’.50 In such cases it is almost certain that the
person so labelled was parochially resident, but legally settled elsewhere,
probably in receipt of a settlement certificate, for that usually accompanied
the use of the word ‘sojourner’ in the eighteenth century. This is more
explicit in the case of ‘William Morgan, certificated to this par. from the
par. of Dry Sandford, and Residing in Yarnton’.51 However, I have found
such reference to legal settlement status in only a handful of cases out
of over 18,000 marriages. With regard to legal settlement, one does not
find the attendant entries that one would expect if this was the main
issue prior to Hardwicke’s Act. Mention of a certificate (probably a
settlement certificate) is very rare indeed; and if settlement was being
referred to one would expect more often to see further qualifications
being made or mentioned (40 days’ residency, certificates, rent paid,
38 Simpson, ed., Derbyshire parish registers: marriages, vol. XIII, Smalley, 29 Nov. 1789, p. 151.
39 Ibid., Elvaston, 29 Oct. 1794, p. 106.
40 Haswell, trans., Parish registers of Brougham, p. 45.
41 Wood, ed., Registers of Long Houghton (Northumberland), 21 Oct. 1710, p. 82.
42 Ibid., 16 June 1719, p. 85.
43 Ibid., 5 July 1726, p. 87.
44 An example cited in Pain and Smith, ‘Do marriage horizons accurately measure migration?’,
p. 45.
45 Wood, ed., Registers of Long Houghton (Northumberland), 1 Dec. 1771, p. 98.
46 Ibid., 7 Sept. 1779, p. 100.
47 Simpson, ed., Derbyshire parish registers: marriages, vol. XIII, p. 180 (Kedleston, Derbyshire), 1
Jan. 1763. It is my impression that estate or ‘closed’ villages (such as this one) were often more
observant in the wording used.
48 Phillimore and Oldfield, eds., Oxfordshire parish registers: marriages, vol. II, Yarnton, 1815, p. 151.
For some similar but much earlier examples, from the diocese of Canterbury in 1475-1600, see
O’Hara, Courtship and constraint, p. 125.
49 Simpson, ed., Derbyshire parish registers: marriages, vol. XIII, Elvaston, 29 Oct. 1794, p. 106.
50 See for example ibid., p. 180 (Kedleston, Derbyshire), 26 Aug. 1765. There is also very
occasionally a mention of ‘lodgers in the parish’, as in Yarnton (1809), in Phillimore and Oldfield,
eds., Oxfordshire parish registers: marriages, vol. II, p. 150.
51 Ibid., Yarnton, 1821, p. 152.
 Economic History Society 2002

english rural societies and marital endogamy
271
paternal settlement, and so on). These are virtually always lacking in
marriage registers.
While it is possible to dwell on interesting examples in the hope of
general illumination, most entries in registers are plainer and simply state
‘of this parish’ or ‘of [place]’. Standardization of entry improves from
1754, as prescribed by the new act, but there is unlikely to have been
much difference in the meaning of such wording before and after 1754.
I did not use entries when such reference to place was lacking in a
register over a substantial number of years (such periods sometimes
coincide with particular clergymen’s incumbencies). There are periods
(mainly before 1754) when it is necessary to drop marriages from analysis
because no information was being given.52 However, for the (published)
registers analysed here, I was struck by how good these details generally
were both before and after 1754. Given the unequivocal wording of
Hardwicke’s Act, and the generally reinforcing nature of the more explicit
register entries, my strong presumption is that residence was almost
always meant by contemporaries. This may have been anything from
long-term residence, amounting to family ties of considerable longevity,
down to quite short-term residence, such as a period of recent service
or brief employment. It is necessary to stress this possible variation. Such
evidence does not necessarily tell us anything very certain about mobility,
or where a person ‘came from’: one person may have been resident in a
parish all his/her life, while another may have led a highly picaresque
rambling existence before recently becoming resident. Both such cases
would normally and simply be said to be ‘of this parish’, a phrase that
cannot be taken to imply geographical stability. It is therefore necessary
to be very careful indeed in generalizing about trends in residential
stability, or labour mobility, from parish register material. Many familial,
cultural, and other factors influenced where vital events were registered,
and the registration of those events may be a poor indicator of the nature
of migration. It is possible that residence upon marriage better reflects
adult residence than registration of baptisms and burials (given many
customs affecting baptism or burial);53 but marriage itself was also a
52 The years used for analysis of parishes are shown in appendix A. The data from each register
were inspected on yearly dot-type plots, in which the four possible types of marriage were visually
distinguished, and periods of 10 or more years with no mention of external places were generally
excluded. When graphically shown, periods of omitted residential entries usually stand out very
clearly against the running patterns for each parish. In my 69 parishes, such periods affect 12
registers (mainly in Leicestershire, Dorset, and Norfolk), usually in the early eighteenth century.
Excluding such parishes does not alter the findings (county graphs including them and excluding
them were almost identical) and the omitted periods have no influence. Periods omitted because
external residency was not being registered are indicated in appendix A. In normal periods, when
people were being regularly entered as resident in another place, names with no residential wording
were interpreted as being ‘of this parish’. In an era in which insider-outsider divisions, however
defined, were crucial to so much of parochial life, when the clergy were intimately involved with
vestry administration, and when such register entries were often used as a leading avenue of enquiry
in settlement cases and disputes, I suspect that these entries are really quite reliable. I have, however,
used published registers, which are often of higher quality than unpublished ones, and careful
scrutiny of the quality of the register is an essential prior condition for this kind of research, as for
other demographic purposes.
53 And see above, n. 30.
 Economic History Society 2002

272
k. d. m. snell
custom-influenced and positioned event,54 and clearly there are grounds
for caution in extrapolations from findings about geographical endogamy.
These points are inescapable, notably in a society with so many servants,
apprentices, and other mobile people, and they may vitiate some
incautious studies of migration using marriage registers. Even so, limi-
tations and caveats of similar kinds are inherent in all sources which
partially document personal mobility in the past, and they should not
prevent research into geographical endogamy using the enormous quantity
of parish register evidence.
II
For present analytical and presentational purposes, and to aid analysis of
change over time, marriages may be categorized as follows:
A
F
B
G
E
All registered marriages in a parish.55
‘Foreign’ marriages, where both partners are resident elsewhere.56
Bride-as-outsider marriages (the bride being resident elsewhere).
Groom-as-outsider marriages (the groom being resident elsewhere).
‘Endogamous’ marriages, where both partners are resident in the
parish of marriage.
Developing from these categories, for present purposes I shall define
‘uxorilocal’ marriages as those weddings taking place in the wife’s parish
of residence, in which the husband may or may not be resident elsewhere.
(These comprise categories G plus E.) ‘Virilocal’ marriages are those
taking place in the husband’s parish of residence, in which the wife may
or may not be resident elsewhere (categories B plus E). ‘Exogamous’
marriages comprise categories F, B and G. For some analytical purposes,
however, it may be desirable to exclude ‘foreign’ marriages from that
exogamous category, and this would entail taking only B plus G.57
Changes in marriage behaviour over time may manifest themselves in
shifts in the relative numbers of any of these five categories. It is these
shifts that I hope to illuminate here between 1700 and 1837.
When I began to explore this subject, I anticipated finding a decline
in endogamy over this key early period of industrialization, and an
54 See my ‘Parish registration and the study of labour mobility’, pp. 29-43, on custom-related
mobility for registration of vital events.
55 I included all registered marriages by banns or licence, that took place in the church or chapel
documented by the Anglican register. Banns without a subsequent marriage, and mention of banns
or licences for marriages elsewhere, were ignored. Five of the Northumberland registers (for Hebburn,
Edlingham, Whalton, Longhoughton, and Ingram) also gave entries (from 1754) for banns called
in these places of worship for which the marriages were solemnized elsewhere. Dual calling of banns
was required under Hardwicke’s Act (section I) when the persons to be married ‘shall dwell in
divers parishes or chapelries’. These Northumberland entries were usually for locally resident men
who were marrying in another place, and I have excluded them from my calculations. No other
registers supplied this information.
56 The term ‘foreign’ marriages was used by contemporaries in the eighteenth century to describe
such marriages, and that usage is adopted here.
57 For example, one might wish to calculate exogamous marriages over time as a percentage of
all marriages excluding foreign marriages: ((B + G)/(A − F)) × 100.
 Economic History Society 2002

english rural societies and marital endogamy
273
Figure 2. ‘Foreign’ marriages as percentage of all marriages
Sources: See text and app. A
opening up of rural marriage patterns. I envisaged an argument that
linked rising exogamy with greater migration (associated with industrializ-
ation and large-scale military mobilization), some presumed relaxation of
the settlement laws, agricultural innovation and regional specialization,
greater reliance upon solely wage-dependent labour, an expansion of
cottage industry and its supply and marketing networks, the effects of
extra-parochial Nonconformist organization, and other related matters.
There were many indications in the historiography and theoretical litera-
ture that would support such expectations.
However, my findings were quite contrary to this. The results were
unexpected and remarkably consistent across the diverse range of parishes.
To display them, I have combined the parish figures for each county,
and expressed these by decade.58 It is helpful to start with ‘foreign’
marriages, where both partners were resident outside the parish. The
dramatic changes here are shown in figure 2. All counties show the same
major reduction in ‘foreign’ marriages, to very low percentages of all
marriages after 1754. One sees the very considerable proportion of such
marriages before Hardwicke’s Act, usually comprising between 10 and
58 This was done for analytical convenience, to distinguish counties as a guide to regions, and to
avoid small number problems of the kind common in family reconstitutions.
 Economic History Society 2002

274
k. d. m. snell
30 per cent of all marriages. In some individual parishes the figures can
be much higher. Hardwicke’s Act, designed to restrict clandestine mar-
riage by imposing tighter controls (with severe penalties) over residency
requirements and the marriage of minors, shows itself to have been highly
effective over all counties. Some ‘foreign’ marriages were found after
1754, but they were very few; and they were probably often the result
of considerations such as the temporary closure of churches,59 or absentee
clergy, necessitating marriage in a place where neither partner was nor-
mally resident. Such marriages were rare after 1754, and the parish data
indicate that this change came about swiftly following the act.
In some cases registers contain entries resembling that for Chickerell,
Dorset: ‘Here end all Foreign Marriages’.60 It is worth stressing the very
large number of ‘foreign’ marriages conducted in some parishes before
1754, which was tantamount almost to a local clerical industry in some
places. Of the chosen 69 parishes, Askerswell (1719-53), Urswick (1723-
48), Downham (c. 1724-54), Halton (1727-54), Tatham (1706-37),
Swithland (1721-34), Syderstone (1729-54), North Barsham (1700-20),
Tatterford (1743-53), Snoring Parva (1723-39), Edlingham (1710-22),
and Alton Pancras (1700-18) all showed sizable numbers of such mar-
riages in the years indicated. The problem, if such it was, was not one
that was growing inexorably until 1754, for peaks often occurred some
time before then; and some parishes show very little sign of the practice.
In others, however, the large numbers indicate that certain clergy were
deliberately supplementing their (often inadequate) income with fees from
additional ‘foreign’ marriages, which they were clearly keen to conduct.
It can be demonstrated that high proportions of foreign marriages were
associated with low values of ecclesiastical livings, markedly so in some
of the Dorset and Derbyshire parishes. Scanty clerical income over this
period, particularly in small rural parishes, was a significant problem for
many benefices. The foundation of Queen Anne’s Bounty in 1704 aimed
to augment such income, but it was very slow and long-term in effect.
Gains in the clergy’s economic position in parishes enclosed by Parliamen-
tary Act had yet to occur. In addition, clerical non-residency was more
of a problem at this time than later; and for many couples this would
have necessitated marriages in places where they were non-resident.
Presumably many of these marriages were ‘clandestine’ (as perceived by
Hardwicke’s Act and by modern social historians)—but the issue is not
only a matter of demand from couples intent on surreptitious marriage.
The use and abuse of the licence system also needs to be considered, as
well as the supply of clerical services and competitive fees. Inadequate
incomes often underlay clerical absenteeism, and until 1754 ‘foreign’
59 In 1830 marriages were allowed in nearby churches if a church was in disrepair or abandoned
for demolition, and this certainly occurred earlier also. See 11 Geo. IV, c. 18.
60 Phillimore and Nevill, eds., Dorset parish registers: marriages, vol. VII, Chickerell, 9 March
1754, p. 46.
 Economic History Society 2002

english rural societies and marital endogamy
275
marriages were one possible remedy for the clergy, so strikingly shown
here.
Figure 3. Endogamous marriages as percentage of all marriages
Sources: See text and app. A
The most significant long-term change cannot be adduced as easily to
Lord Hardwicke. This was the rise in endogamous marriages (category
E: where both partners were from the same parish). This is shown in
figure 3, where endogamous marriages are expressed as a percentage of
all marriages. This is perhaps the most surprising finding. The change
was from between approximately 45-60 per cent endogamy, rising to
about 60-80 per cent. It was a consistent and steady rise across all these
varied counties. (The converse of this was a long-term decline in exogam-
ous marriages as a percentage of all marriages: from some 40-55 per
cent, down to approximately 20-40 per cent). As a development during
the heyday of early industrialization, it will be contrary to many historical
expectations, although it is a change that accords well with initial sugges-
tions in the historiography, as cited earlier. Hardwicke’s Act seems to
have been less instrumental here, for its date marks no sharp break: the
change was certainly under way well before 1753, and it continued
thereafter. This shift to very high levels of endogamy by the 1830s (when
more than 80 per cent of all marriages in some parishes were
endogamous) is very surprising. It lays the groundwork for the subsequent
 Economic History Society 2002

276
k. d. m. snell
decline in endogamy from such high levels which was documented from
various points after the mid-nineteenth century, by authors including
Constant, Perry, and Peel. Contrary to some presumptions, the high level
of endogamy in the early to mid-nineteenth century was not a longstand-
ing historical condition, stretching timelessly and unchangeably into the
pre-industrial past. Like many other demographic phenomena, it had
been subject to noticeable change earlier, and in particular the rise shown
here between 1700 and 1837.
Figure 4. Endogamous marriages as percentage of marriages (all counties)
Sources: See text and app. A
It has been shown that Hardwicke’s Act curtailed the numbers of
‘foreign’ marriages. This affects the total marriages from which these
percentages have been calculated, and thus may tend to lessen the
predominance of endogamous marriages before 1754. Figure 4 tests for
this effect, by combining all 69 parishes, and expressing endogamous
marriages as a percentage of all marriages minus foreign ones.61 This is
shown alongside the combined percentages (all parishes) for all marriages,
calculated as in figure 3. It will be seen that the trend is diminished
when foreign marriages are eliminated, but that the overall movement is
still unmistakably upward over the long term.
61 That is, using the marriage categories above: (E/(A − F)) × 100.
 Economic History Society 2002

english rural societies and marital endogamy
277
Figure 5. Bride-as-outsider marriages as percentage of all marriages
Sources: See text and app. A
One of the types of marriage listed above was the ‘bride-as-outsider’
category, with the wedding taking place in the groom’s parish of residence.
These marriages are shown in figure 5, as a percentage of all marriages.62
The movement was downwards here, as a long-term change, not very
obviously tied to Hardwicke, with quite marked declines for the parishes
in some counties, such as Norfolk, Dorset, or Northumberland.63 In
other words, it was becoming less and less common to find ‘bride-as-
outsider’ marriages. Secondly, and associated with this, it can be shown
that ‘bride-as-outsider’ marriages were diminishing markedly relative to
‘groom-as-outsider’ marriages. Taking all 69 parishes together, there were
about three groom-as-outsider marriages to every one bride-as-outsider
marriage in the early eighteenth century. But this ratio had moved to
about 7:1 by the early nineteenth century, the change being especially
concentrated between 1750 and 1770. Indeed, in the combined parishes
62 The removal of foreign marriages from the total marriages tends slightly to accentuate the
downward trend, but has very little other effect.
63 Here, as elsewhere, too much heed should not be paid to short-run fluctuations of the data,
for these are largely attributable to the data sizes for these parishes, and if further data were added
for more parishes such fluctuations would tend to be smoothed out. The overall patterns of change
would probably be little affected.
 Economic History Society 2002

278
k. d. m. snell
of some counties, this ratio moved from about 3:1 to well over 10:1. In
the Northumberland parishes, it went from about 1:1 in 1700-9, to
around 20:1 between 1790 and 1819. There were steep rises also in
Norfolk and Sussex, although in these three counties these ratios were
declining from about 1800 (Sussex) or 1810 (Northumberland and
Norfolk). In such cases one wonders about the effects on male residence
of wartime mobilization and de-mobilization. Over the entire period, the
trend of this ratio was strikingly upward in every county.
Figure 6. Uxorilocal marriages as percentage of all marriages
Sources: See text and app. A
Another way to address these changes is to consider ‘uxorilocal’ mar-
riages, defined here as weddings where the bride was resident in her
parish of marriage (marriage categories G plus E). The trend here is
shown in figure 6. Much of this encompasses the upward movement of
endogamy (category E), and it was a development situated mainly in the
mid-eighteenth century, presumably closely related to Hardwicke’s Act.
One also sees here a growing uniformity of practice over time and regions.
From the 1760s, about 95 per cent of all marriages were of this type.
As an historical generalization, it would, of course, be an exaggeration
to say that uxorilocal weddings were the norm, particularly in the early
eighteenth century. However, ‘giving away the bride’ in her own parish,
usually followed by neolocal residence in the groom’s (increasingly
 Economic History Society 2002

english rural societies and marital endogamy
279
identical) parish, was seemingly becoming more common.64 Going further,
some readers may already be trying to detect regional contrasts in the
data for these different counties. No doubt further quantitative research
would help to document local and regional differences in practice (for
instance in uxorilocal weddings), and would lay the groundwork for a
more refined exploration of regional differences, which might also incor-
porate regional literary evidence. There is, after all, massive and surpris-
ingly untapped potential for such work in English and Welsh parish regis-
ters.
III
To develop the analysis further, and as an aid to explanation, regression
and correlation tests were conducted to see what factors were associated
with high endogamy. The results showed that parish population size, and
population density, were positively associated with high percentages of
total marriages being endogamous.65 The larger the settlement, the higher
the percentage of endogamous marriages. The point here is that marriage
partners could be found more easily in larger or more densely settled
parishes, whereas residents in smaller parishes more commonly had to
seek partners resident elsewhere.66 This has been suggested by French
64 Where the spousal parishes were different, this change may point to higher valuation of men
and less reluctance to lose women, although the anthropological literature is uncertain about this
implication. However, a rising trend of endogamy would counteract this, the women not being lost
to their community. A further matter to consider is that throughout this period a woman took her
husband’s legal settlement upon marriage, even if she had never been there before. This had strong
implications for her post-marital residence. In the later eighteenth- and early nineteenth-century
context of rising poor-relief expenditure, this could more frequently involve adverse reactions to
‘foreign’ married women, legally imposed upon a community that did not know them. It would,
therefore, be possible to speculate (assuming frequent synonymity of legal settlement and residence
after marriage) that geographical endogamy, and thus a growing identity between maiden settlements
and those of grooms (i.e. a continuity of maiden to marital legal settlement), would help to obviate
the frequent difficulties or hostility which many newly wed women would otherwise have faced
during a period of high relief costs.
65 Using parishes as the units of analysis, the Pearson correlation coefficient between total
population (1841) and the percentage of marriages that were endogamous (for each parish over its
documented period) was 0.386 (p = .003). Very similar results were obtained with 1811 and 1831
population totals, and with Spearman (rank) correlation. The equivalent Pearson coefficient using
population per sq. km. in 1831 was 0.425 (p = .001). In other words, in bivariate analyses such a
variable explains about 15 to 18% of the variability in parochial endogamy, measured (with due
precaution) as a percentage. In regression equations, population size and population density used
together produce values for R squared of about 0.24 (F = .000). If all ‘foreign’ marriages are
eliminated from consideration, percentages of endogamous marriages are re-calculated from the
resulting totals, and these procedures are repeated, the results are little altered. All findings were
re-examined with this precaution in mind, but the results differed very little from those reported here.
66 See also Bryant, ‘Demographic trends in south Devon’, p. 133, which argues that larger parishes
had more stable populations. I do not wholly agree with this deduction from Bryant’s data.
 Economic History Society 2002

280
k. d. m. snell
data,67 and is certainly true for these English parishes.68 It can also be
demonstrated for other and earlier English data.69 Additional factors
positively associated with endogamy were the population growth rate
(1811-41), average numbers of marriages per annum, the proportions of
families engaged in trade (1831), measures of religious denominational
diversity (calculated from the 1851 religious census),70 the value of
67 Flandrin, Families in former times, pp. 34-5, 47. This cites late seventeenth- and eighteenth-
century figures from villages in the Vallage suggesting that the larger the village, the higher the
degree of endogamy. In very small villages, the percentages could be as low as 31 to 34. In others
they rose to over 70%, up to 93% in one case, sometimes associated with violence against outsiders
trying to court village women (which was also often documented for England and Wales). In some
mountain villages (e.g. Caillac) topographical conditions could compound this and induce very high
figures indeed. Such ‘startlingly high’ figures led to very dense kinship ties. It might be expected
that the English rural changes towards higher endogamy would also intensify village kin ties and
produce high surname densities (which certainly obtained in some Norfolk parishes). This would
have many implications for community formation and consciousness. Furthermore, while the caveats
made above regarding links between ‘place of origin’, migration, and residency upon marriage should
not be forgotten, it is worth pointing out that high kinship densities could well conduce to lower
marriage ages. As King has suggested, they ‘would tend towards lower female marriage ages through
more, and more certain, courtship opportunities’. King, ‘English historical demography’, p. 150.
Such densities would have enhanced the roles of relatives and ‘friends’, facilitating contacts based
on trust and knowledge, providing more local support for young couples, and so on. (I use the
term ‘friends’ in the sense of the English labouring poor: to denote both ‘family’ and closely linked
people, whether they be kin or those in close upbrought or dependent alliance.)
68 This finding could be extended to remark on another area of debate. It is an interesting
historical paradox that the most ‘closed’ or ‘estate’ parishes had to bring in many outsiders to marry
(by virtue of their deliberately ‘closed’ nature and small populations), while the more ‘open’ or
populous parishes could rely more upon their own populations to obtain marital partners. Such an
effect was contrary to some of the dispositions towards closure that brought ‘closed’ parishes into
being. Labour shortages in ‘closed’ parishes may also have conduced to similar ends, and temporary
in-coming labour was one way in which potential marriage partners resident elsewhere might be
met. Of my 69 parishes, Kedleston, Helhoughton, Hanborough, and Langton Herring had property
held ‘in one hand’, and Breadsall, Elvaston, East Stower, Rampisham, Swyre, Downham, Birstall,
Eastwell, Queniborough, East Barsham, North Barsham, Holme-by-the-Sea, Toftrees, Bothal,
Edlingham, Ingram, Longhoughton, Pyrton, and Cocking had property held ‘in a few hands’. I use
here the landownership classifications of Wilson, ed., Imperial gazetteer, which are favourably assessed
as evidence in Snell and Ell, Rival Jerusalems, app. E, pp. 440-8. This distribution of the 69 parishes
into categories of landownership, as judged from this source, is exactly representative of broader
patterns for the most ‘closed’ parishes (5.8%), and shows fewer parishes than expected being ‘in a
few hands’ (28%, as compared with an expected 54%). See ibid., tab. 7.2, p. 225.
69 Mitson, ‘Significance of kinship networks’, pp. 60, 74; Phythian-Adams, Re-thinking English local
history, p. 35; O’Hara, Courtship and constraint, pp. 123, 132-3, 135-7. I have re-analysed Mitson’s
figures for 11 parishes (1561-1710) along the lines of my presentation, using an unweighted mean
of her population totals from five sources (1603-1743). They show much the same positive relation-
ship (Pearson’s r = 0.599) between population size and parochial endogamy as do my parishes.
Curiously, their positive scatter is parallel to, but noticeably below, a linear trend for my (later)
data, suggesting significantly lower endogamy (about 20% lower) for any given population size in
1561-1710 compared with 1700-1837. (Any poorer registration earlier would produce the loss of
some information about differential residency, and so would predispose to seemingly higher endogamy
at earlier dates.) If this is a valid evidential and historical finding, it has very interesting implications
indeed for the structure of communities and kinship densities over time, and it may suggest a
longer-term rise in endogamy commencing well before 1700. It is also necessary to consider the
possibility of even earlier cycles, and I have in mind the downward shift in endogamous courtships
documented by O’Hara for Kentish parishes (1475-1594), in her Courtship and constraint, pp. 127-
9. As with Mitson, however, O’Hara’s mean figure of 47% of courtship cases being co-resident
parochially (if translated into ‘endogamous marriages’) is well below the later figures from my work.
This throws into sharp relief the very high parochial endogamy of the period c. 1780-1840.
70 For further discussion of such measures, see Snell and Ell, Rival Jerusalems, app. C,
esp. pp. 436-7.
 Economic History Society 2002

english rural societies and marital endogamy
281
parochial property, values of the clerical living,71 and average per annum
poor relief (1832-6). Property values per caput, and the proportion of
families engaged in agriculture (1831), were negatively correlated with
endogamy. The sex ratio was (perhaps surprisingly) irrelevant. In other
words, the more these rural parishes leaned towards ‘urban’ character-
istics, judged by such measures, the higher their percentages of endoga-
mous marriages. All these results follow predictably from the influence
of population size and density, and in statistical tests these two variables
take precedence in explaining the relative ease, or difficulty, experienced
by people in finding marriage partners resident in their own parish. There
must have been many further cultural and social variables having influence
here, most of them less amenable to statistical exploration. At present
their precise roles are largely hidden from the historian.
Figure 7. Parochial endogamy and population size
Sources: See text and app. A
It is possible to extend this analysis, and to shed interesting light on
a further regional dimension, by considering the distribution that emerges
when parish endogamy percentages are plotted against population size
on a scattergram. Figure 7 shows this, using parish populations in 1841,
the counties of each parish being symbolically distinguished, and a lowess
curve fitted.72 The tendency for endogamy to rise steeply up to a popu-
71 The latter two variables are from Wilson, ed., Imperial gazetteer, and are only loosely dependable
as they are not contemporaneous with the marriage data.
72 A few places of non-parochial character were omitted from this scattergram. For the lowess
curve, see Norusis, SPSS for Windows, pp. 558, 645-6.
 Economic History Society 2002

282
k. d. m. snell
lation of nearly 400, and then to level off, should be noted. This shows
the threshold population size needed to stabilize endogamous practice.
Very similar results are obtained with 1811 and 1831 population data,
or with earlier measures involving average marriages per year or decade,
or with population density as the independent variable.73 The logical
coherence of these findings is also reassuring for such use of residency
data from parish registers.
Interesting regional characteristics are also immediately apparent from
figure 7. The most conspicuous is the case of Norfolk, whose rural
inhabitants in the past have indeed sometimes been described (for various
reasons) as having rather ‘stay-at-home’ proclivities. One hesitates to
reinforce such stereotypes, but it happens to be true that of all these
parishes those from Norfolk show very much the highest endogamy
relative to population size.74 In general they are situated together in
figure 7 well above the parishes of other counties, all being on or above
the fitted line, in positions higher than their population sizes might
warrant. The rural parishes of Dorset come somewhat lower in the
endogamy stakes, but endogamy was still relatively high there—as one
would expect from Perry’s striking findings on working-class rural isolation
in that county.75 Parishes in Oxfordshire also feature high endogamy. At
the other extreme, certain parishes in Derbyshire and Lancashire reveal
levels of endogamy which are among the lowest relative to population.
These were more industrial, higher waged counties, ones in which farm
service persisted in relatively unchanged forms compared with the low-
wage rural south, and in which the numbers leaving their parishes of
birth were higher than in the south. We should envisage parameters of
courtship which had become rather different across these broad regions.
Furthermore, figure 7 clearly shows low-wage counties as having higher
endogamy than high-wage ones. The indications are that higher levels of
indigenous mobility, associated with the survival of service, and with
proximate cottage and factory industrialization (the pull of which helped
farm service to survive in midland and northern regions), may have had
a significant effect upon these geographical marriage patterns, differen-
tiating the Derbyshire and Lancashire parishes from those in Norfolk,
Dorset, Oxfordshire, and some other very agricultural and southern areas.
In other words, high endogamy appears to be linked to the decline of
service (an institution that had structured high youth mobility), relatively
stagnant labour markets, low wages, high structural and seasonal unem-
ployment, and perhaps resulting plebeian cultures that were hostile to
interloping competitors. On the other hand, ‘low’ endogamy or ‘high’
exogamy obtained where such adverse conditions were less pronounced—
that is, in the rural midlands and north.
73 With the latter, endogamy rises (to about 63%), stabilizing when a level of approximately 50
people per sq. km. is reached. Something comparable is also found when acreage is used as the
independent variable: the lowess curve rising sharply to a parish acreage of about 1,800, and then
almost completely levelling off at 55-63% endogamy.
74 The same is true in relation to population density (people per sq. km. in 1831).
75 Perry, ‘Working-class isolation’.
 Economic History Society 2002

english rural societies and marital endogamy
283
It is always interesting to marry quantitative research with literary
description, and to this end the high endogamy of the rural Norfolk
parishes warrants further comment. Norfolk, like Dorset, was notorious
for very low wages, high poor-law dependency, and associated problems
of surplus labour. Such conditions fostered sentiments that were protective
of local resources for the labouring poor, and probably induced further
resistance against outsiders seen to be rivals for local work, welfare, and
wives. This may have heightened endogamous tendencies within these
parishes. There was certainly considerable mobility. Yet in rural Norfolk
from the late eighteenth century it may have involved longer-distance
movement away from home in search of work (to Norwich or the East
Anglian towns, London, or further afield), rather than to immediately
surrounding parishes which shared surplus-labour and poor-law problems,
and from which one might otherwise expect most marriage partners to
come. Many commentators remarked upon fixity and stolid tendencies
among Norfolk labourers. Very few people in Norfolk took up the assisted
migration schemes under the new poor law, many fewer than in Suffolk.
The Revd. C.D. Brereton (of Little Massingham, Norfolk), one of the
most perceptive of poor-law pamphleteers, wrote in the 1820s of how
The marriages among the peasantry are now very much confined to their
respective villages, or villages immediately adjoining each other. The close
intermarriage of near relatives . . . is of course more frequent. We have in this
parish four old men, none of whom are natives of this village . . . . This is
the case of the oldest inhabitants in the adjoining villages . . . they migrated
in early life, in quest of employment and improvement. Whereas now . . . where
they have been born, there they usually (to use the common phrase) belong;
and there they seem to think they are to . . . exist and die.76
Immobility was even sometimes encouraged as parish policy. Rider Hag-
gard, for example, commented that ‘I have noticed all my life that
Bradenham people cling very much to their own village’, and he noted
that farms to let were generally taken ‘by a native and not by a foreig-
ner’.77 He also quoted from a printed parish form for Necton (as used
in the 1830s), which spoke of the need ‘to encourage those who do
marry to marry those women who belong to their own village’. The aim
of this was apparently to reduce ‘the superabundant population’, being
part of a larger attempt to restrain marriage, which included rewards and
penalties respectively for late and early marriage.78 A.G. Street in the
early twentieth century wrote of how
‘Norfolk’s farming
community . . . does not approve of interlopers from other parts . . . it
would take at least three generations before one really belonged’.79 Similar
evidence could be provided for rural Dorset, although it is slightly weaker
76 Brereton, Agricultural labourers, pp. 33-4.
77 Haggard, Rural England, vol. 2, p. 502.
78 Ibid., pp. 502-3. I have not found elsewhere such explicit instruction to marry women from
the same parish. (Necton and East and West Bradenham are in central Norfolk, between Swaffham
and East Dereham.)
79 Street, In his own country, p. 168, extracted from his Country calendar (1935).
 Economic History Society 2002

284
k. d. m. snell
than for Norfolk. For example, a commentator in 1847 spoke of how
‘In Dorsetshire we very much vegetate where we are born, and live very
close indeed.’80 Such comment should not be endorsed in any simple
way—and I have sceptically discussed crude stereotypes of ‘Hodge’ else-
where.81 Even so, it is of much interest that the quantitative findings
here, from a source undoubtedly open to some interpretative ambiguity
in connection with migration, should reinforce such literary comment.
IV
There is no doubt that Hardwicke’s Act explains some of the changes
discussed here, notably the sharp decline in ‘foreign’ marriages. This
bears witness to the effectiveness of that 1753 act in bringing marital
behaviour under tighter control.82 In this case, the change was swift,
dramatic, and permanent, perhaps more so than many demographic
studies have indicated, although it is possible that consensual unions
became more numerous as a consequence. The act also presumably
returned the marriages of some inhabitant couples (defined as ‘foreign’
elsewhere, and perhaps marrying in towns where licence surrogates were
likely to live) back to their own parishes. In this regard it would in effect
have augmented the endogamy rate in their home parishes. Furthermore,
the act seems to have strengthened the custom of uxorilocal weddings.
Even so, many of the changes documented here were gradual and long-
term, beginning well before the mid-eighteenth century, and continuing
long afterwards, and with regard to them the role of this act seems less
certain. The rise in endogamy was one such long-run trend, and so
probably needs explanations beyond Lord Hardwicke.
There are evidential issues of relevance here, concerning improvements
in parish registration and specification of residence, and the possibility
that such improvements may have contributed to the patterns of apparent
change shown. A number of points may be made in this connection. All
data and graphs were analysed with and without the inclusion of ‘foreign’
marriages, and where their omission made a significant difference readers
have been alerted to this. Any such improvement would have been largely
sudden, commencing with the widespread introduction of Hardwicke Act
registers;83 but this does not seem apparent in the steady, century-long
shift to higher endogamy. It may be that better documentation of resi-
dence, or perhaps a clearer understanding among contemporaries that it
was residence rather than any other concept that should be entered in
registers, would have tended to increase the likelihood of extra-parochial
80 Sixth Report from the Select Committee on Settlement, and Poor Removal, XI (1847), p. 157.
81 Snell, Annals of the labouring poor, pp. 1-15.
82 On the effectiveness of the act, see also Outhwaite, ‘Sweetapple of Fledborough’, pp. 36, 45.
The ramifications of the act for marriage within propertied elites are discussed in Lemmings,
‘Marriage and the law’.
83 Some of the parishes studied showed this introduction, older style registers being abandoned
from 1753/4 and replaced by new ones, sometimes with comment by incumbents to mark the replace-
ment.
 Economic History Society 2002

english rural societies and marital endogamy
285
residence being noticed after 1753 than before. Yet, as noted above, it
seems clear that incumbents had long before adopted residence as the
ruling presumption. In addition, any such documentary changes would
probably induce an apparent shift to higher exogamy (i.e. partners resident
in different parishes being more commonly noted), rather than higher
endogamy. It would thus work against the rising trend of endogamy
documented here. While such possibilities cannot be ruled out, notably
from 1754, it seems more likely that the apparent trends reflect something
real historically. Furthermore, if ‘residency’ entries in registers were fickle,
random, and of little documentary value, we would surely not be finding
the very clear-cut and intelligible relationships between endogamy and
population size that were revealed in figure 7.
Accordingly, I shall proceed on the assumption that we are indeed
witnessing something that was historically substantive, and which calls
for an explanation in demographic and socioeconomic terms. This is
uncharted territory, with no historiographical signposts, and so what
follows is a tentative suggestion of causation. Future historians will be able
to refine or extend this account, for such work invites many comparative
possibilities. The degree of endogamy certainly varied between parishes
and regions, as we have seen, and I have covered only rural parishes.
Yet the major changes appear to have been shared by very disparate
regions. It is therefore necessary to think about possible causes that were
quite widespread.
The number of marriages in every decade was rising steadily in these
parishes, as it was more widely. There is no doubt that most of these
places were growing in population. Given the relationship between popu-
lation size and endogamy, the rising population of settlements would
itself tend to engender higher endogamy, making it easier to find partners
within the parish. This must be part of the explanation. Yet this ‘mechan-
ical’ explanation seems insufficient by itself, for the following reasons.
First, the 69 parishes were divided into sub-groups, and those eight
parishes experiencing population decline between 1811 and 1841 were
analysed separately.84 While this is not a reliable guide to their eighteenth-
century demography, it was nevertheless the case that these eight parishes
also experienced significant rises in endogamy, much as did the entire
group of parishes. Secondly, such a mechanical explanation would mainly
affect the smallest parishes, with populations under about 400. At larger
population sizes, further increase in population appears to have little
association with higher endogamy (see figure 7). Endogamy (without
foreign marriages) in these parishes rises from about 60 per cent to about
70 per cent (1700-1837). Extrapolation from figure 7 suggests that it
would have taken a doubling of population size in the smallest 28 parishes
to achieve this (and such a doubling certainly did not occur for all small
84 Cassington, Eastwell, Edlingham, Halton, Kedleston, Tatterford, Toftrees, and Whalton.
 Economic History Society 2002

286
k. d. m. snell
rural parishes),85 and that even when allowing for eighteenth-century
growth about half the parishes would probably be little affected by this
cause.86 It is therefore necessary to suggest additional causes.
From various points after the mid-eighteenth century, poor-law expen-
diture also began to grow in almost all regions, usually accelerating very
sharply in the 1790s and early nineteenth century. Real per caput expendi-
ture on the poor also usually rose, especially from about 1813 in the
south and south midlands. Very large numbers of the labouring classes
became more reliant upon the poor law, and that law permeated local
labour markets. This brought the labouring poor more often into touch
with the settlement laws, as parishes ascertained their settlements with a
view to their removal or the claiming of non-resident relief for them.
Huge numbers were being examined and removed to their settlements
from the later eighteenth century. This was above all a southern situation,
but very many northern parishes also experienced such problems. The
poor, of course, associated strongly with their parishes of legal settlement.
Such parishes were where they ‘belonged’, as many letters from them
said,87 and legal settlement—defined by all contemporaries as the right
to poor relief—was their final recourse and safeguard. The two most
common means to gain a settlement—farm service and full apprentice-
ship—were changing away from their traditional legal forms in many
areas from the mid-eighteenth century, making it harder to gain legal
settlements in one’s own right, enlarging the proportions who therefore
took paternal settlements, and perhaps discouraging movement away from
home parishes. Legal settlement was something that landowners and
employers (as ratepayers) were often keen to restrict, especially when
relief expenditure rose sharply. I have discussed this and its consequences
elsewhere.88 This legal framework, and this context of rising relief costs,
provided the perfect breeding ground for intensified forms of local xeno-
phobia: against interlopers, ‘foreigners’, outsiders who might threaten to
usurp portions of the ‘limited good’ available to ‘core’ and settled
85 Given the well-known problems with earlier sources, it is hard to arrive at satisfactory population
figures for the early eighteenth century. The Cambridge Group’s 404 parishes, as used by Wrigley
and Schofield, Population history of England, include only three of my 69 parishes. These are Ardingly,
Cowfold, and Standlake. Analysis of their vital events, using appropriate crude birth rates, suggests
that their respective populations between 1700 and 1812 went from about 316 to 468 (Ardingly),
203 to 653 (Cowfold), and 622 to 493 (Standlake). In other words, one should not assume that
such rural parishes shared national growth patterns. I am grateful to Ros Davies for these calculations
using the Cambridge Group’s data.
86 The use of average numbers of marriages p.a. for the early eighteenth and early nineteenth
centuries produced very similar lowess curves to figure 7, and also suggested that over half of these
parishes would be largely unaffected by this ‘mechanical’ effect.
87 For an excellent collection of such letters, see Sokoll, ed., Essex pauper letters.
88 Snell, Annals of the labouring poor, pp. 69-84, 234-9, 334-43, on the decline of farm service
and traditional apprenticeship, notable in many parts of southern England, and the resultant growth
of the proportions who failed to gain a settlement in their own right. Those changes conduced to
inter-generational continuity in parishes, and thus may be related to the rising trends in endogamy
outlined here. The points I made there about changing social relations and senses of community
(ibid., pp. 334-43) are augmented by these findings on endogamy.
 Economic History Society 2002

english rural societies and marital endogamy
287
parochial inhabitants.89 There had, of course, always been efforts to
restrict access to communal resources, by manorial court action, local
by-laws, settlement law, crowd action, and many other means. Yet issues
of entitlement, and access to and exclusion from resources, became
extremely acute in many rural parishes over this period. Among the
examples were disputed rights to traditional forms of employment, agi-
tation over soaring prices in years such as 1795-6, 1799-1801, or 1809-
13, low wages, poor-law and charity administration, reduced access to
remaining commons and wastes, gleaning, the controversy over enclosure
as a form of discipline and exclusion, let alone the violence against Irish
harvesters, Luddism, the destruction of threshing machinery, and many
other protests (e.g. ‘Bread or Blood’ in 1816, Captain Swing in 1830-1)
that were defensive of employment. Many of these were set against the
recurring and extended background of European wars. As Wrigley has
pointed out, the country was on the verge of a Malthusian crisis, and,
as he and Schofield have shown, it was not by any means yet clear that
productivity gains would be sufficient to avert this.90 Most labouring
people in the countryside were aware of acute competition for limited
resources and means of livelihood.
Such near-famine conditions and high parish dependency probably
consolidated local rivalries, aggravated attempts to stop settlements being
gained, and in many rural areas induced hostility to interlopers who were
felt to detract from local well-being. The labouring poor faced with these
crises are often described as coming to see themselves more in class
terms; but these social perceptions were still chained to conservative
expectation and sanguine ideas of parochial administration.91 John Clare’s
The parish: a satire, to take one among many examples, was intensely
bounded locally, vehement though its social grievances were.92 Through
much of rural England, a broader labouring sense of union gained fuller
expression only in the 1870s, bringing together labour unions which
hitherto had often been extremely localized.93 Against this background of
localized identification, compounded by many local rivalries,94 a growing
suspicion of extra-parochial outsiders coming in to marry was to be
anticipated. In the rural conditions of the late eighteenth and early
89 The discussion here derives partly from the anthropological theories of ‘limited good’ most
associated with Foster, ‘Peasant society’; idem, ‘Anatomy of envy’; and Strathern’s ideas of ‘core’
belonging, outlined in her Kinship at the core. See also Snell, ‘Culture of local xenophobia’.
90 Wrigley, People, cities and wealth, pp. 234-41; idem, Continuity, chance and change, pp. 60-7;
Wrigley and Schofield, Population history of England, chs. 10-11.
91 Snell, ‘Deferential bitterness’, pp. 177-81.
92 Clare, The parish.
93 Scotland, Methodism and the revolt of the field, p. 12.
94 See for example Ellis, Down a cobbled street, p. 30. Such rivalries—as between Clovelly and
Appledore, or Sheringham and Cromer, Bala and Llangollen, Pudsey and Yeadon, Walsall and
Darlaston, Bruton and Batcombe, Elmdon and Chrishall, north and south Gower, and countless
other examples (including parishes within towns, as in Derby, or between the parishes of St Anne
and St Giles in London)—were extremely widespread across the country, fuelled by many prejudices
and periodically flaring into youth-group factional violence. There is little evidence that early
industrialization diminished these local hostilities, and it may well have aggravated them. See Snell,
‘Culture of local xenophobia’.
 Economic History Society 2002

288
k. d. m. snell
nineteenth centuries, it is perhaps unsurprising to find parochial endogamy
becoming more frequent. For women had long been seen as a ‘local
perquisite’, their courtship by outsiders often meeting with doubts or
even violence. Contemporaries remarked upon this, as they also did in
France, and such views have been widely documented in later community
studies touching upon notions of ‘core’ belonging.95 Suspicion of inter-
lopers, particularly newly married men coming in to reside from else-
where,96 was another corollary of the socioeconomic conditions increas-
ingly pressing upon the rural poor after the mid-eighteenth century.
Among the 69 parishes studied here, parochial endogamy was at its most
extreme in rural parishes in Norfolk and Dorset, and to a slightly lesser
extent in Oxfordshire and Sussex. While some of these rural areas may,
for a variety of reasons, have been culturally inclined to such behaviour,
these were counties with among the lowest wages, worst living standards,
and highest relief expenditures in England. They were also counties where
there had been considerable decline of farm service in the later eighteenth
and early nineteenth centuries, and this, like their low real wages, set
them apart in crucial ways from many other counties in the north and
midlands, where mobile farm service usually persisted, where rural wages
were much higher, and where the levels of endogamy (see figure 7) were
significantly lower.
Linking my findings to those of Perry, Constant, and Peel for the later
nineteenth century, it therefore seems that rural endogamy peaked in the
most adverse period of rural living standards: from about 1770 to 1840.
This occurred in all regions, but it was most extreme in the rural south.
In all likelihood, those threatening living conditions intensified parochial
resistance against outsiders among single labouring men seeking brides.97
Southern rural labour markets and poor-law administrations especially
95 Strathern, Kinship at the core, pp. 170-1, 176-9 on ‘stealing women’; Evans, Askthe fellows who
cut the hay, p. 239; Lawson, Progress in Pudsey, pp. 74-83; Rees, Life in a Welsh countryside, pp. 83-
4; Stevenson, Popular disturbances in England, pp. 51-2. Women themselves sometimes objected to the
exogamous courting behaviour of men from their parish. For example, the women of Ravenstonedale
(Cumbria), who ‘have for many years past been much injured and abused by the illegal practice of
our neighbouring parishioners’, issued a proclamation in 1776 to all single men of their parish,
telling them that they would be made to pay a fine or imprisoned if they married a woman from
another parish. Findler, Folklore of the lake counties , p. 11.
96 Brides who married in their own parishes and came to live in their husbands’ parishes were
more easily tolerated, for they were less threatening to parochial resources. In addition, many parish
authorities during this period were eager to marry their local women to men from elsewhere, as this
removed those women from parochial responsibility and risk (wives took their husbands’ settlements
under settlement law, and bastards took their place of birth as settlement. A settled and pregnant
single woman could not be removed anywhere.) These considerations may underlie reinforced
‘customs’ of uxorilocal weddings and virilocal marital residence during this period, although it is
too early yet to give this a regional underpinning.
97 Social tensions were probably not eased by a situation in which parish poor-law authorities—
well versed in Malthus and bastardy fears—were rather differently disposed than single men towards
unmarried women, seeing them as potential burdens (or suppliers of burdens) on the parish poor
rates, and having every wish to see such women married out. In such social tensions over marriage,
the rising endogamy of the period may suggest that in this rural arena, if in few others, the interests
of the early nineteenth-century labouring poor predominated. Here, as with the causes of rising
endogamy and its implications for the changing composition of communities, there are further links
connecting historical demography with the historiography of the labouring classes.
 Economic History Society 2002

english rural societies and marital endogamy
289
disadvantaged single men, for married men had precedence for employ-
ment among employers who were also rate-payers. This was a factor in
the decline of service. In addition, solidarity among such men was likely
to have been more apparent in larger, more (supposedly) ‘open’ parishes.
Xenophobic parochial factionalism in such contexts, probably buttressed
by ‘high’ kin densities, thus augmented ‘mechanical’ explanations of rising
endogamy as linked to settlement size, that I have outlined. It took the
rising rural standard of living, broadened horizons, and declining village
populations of the subsequent half century to weaken such endogamy
and its associated attitudes in most parishes, to reassert the greater
outwardness of rural marital behaviour seen in the early eighteenth-
century evidence, and then, during the twentieth century, to reduce
such extreme levels of rural endogamy to something of an historico-
anthropological curiosity.
Let me end with two contrasted examples of literary evidence, from
very different periods, which sum up this change. It is always necessary
to be extremely cautious when reading about supposed generational
changes. Nevertheless, caution spurs curiosity, and if these statements
are taken as broad generalisations, both of them seem to contain some
truth, and to reflect general shifts that are now apparent in the quantitative
evidence. I earlier quoted a Norfolk clergyman, the Revd C.D. Brereton,
writing from Little Massingham on the once mobile elderly men, non-
native to his parish, whom (in the 1820s) he contrasted with the more
endogamous, stay-at-home tendencies of the younger generation, whose
‘marriages [are] very much confined to their respective villages’.98 Now
let us move forward a century or so, and compare the courting youth of
a later period. By the close of the interwar period in the twentieth
century, the rural commentator and broadcaster F.G. Thomas was writing
in The changing village about older ‘ghetto-like’ rural inhabitants who
‘show very little desire to move. They seem almost as immobile as their
parents . . . they will tell heroic tales of journeys in their youth in a cart
drawn by a donkey.’ By contrast, ‘A new mobile generation is in the
country . . . infatuated by mobility. The pedal cycle, the motor bike and
pillion seat, are increasing. . . . They go “courting” outside the village.’99
University of Leicester
DEMOGRAPHIC AFTERWORD
Discussion of geographical, kinship, occupational or denominational endogamy
in the English demographic literature is scarce. This is partly because kinship
endogamy is of little importance in an English context of considerable geographi-
cal exogamy and mobility, social mobility and comparative flexibility of marital
choice, neolocal residence, and complex structure of marriage.100 Kin inter-
marriage was rare in England. Even occupational endogamy was not very con-
98 See above, n. 76.
99 Thomas, Changing village, pp. 19-20.
100 Macfarlane, Marriage and love in England, pp. 250, 254-8.
 Economic History Society 2002

290
k. d. m. snell
spicuous, especially when compared with situations in India or some other
countries.101 Religious denominational endogamy existed to some extent, but the
examples of this were exceptions to normal practice.102 This has all contributed
to relatively little attention being paid to the subject.103 Nevertheless, the argu-
ments of this article have demographic implications that are worth mentioning.
The effectiveness of Hardwicke’s Act in curtailing foreign marriages is striking,
and the decline of such marriages affects the demographic use of parish regis-
ters.104 The very large numbers of ‘foreign’ marriages in many rural parishes
may induce further caution in the use of baptism-marriage-burial ratios to study
population trends. The numbers of these marriages in some early eighteenth-
century registers can bear little relation to the intrinsic demography of the parish,
given the foreign marriages, and the latter were largely from nearby parishes. As
Outhwaite pointed out, Hardwicke’s Act restored these to their canonically
rightful parishes.105 It is often said that such marriages (when termed
‘clandestine’) were even more of a problem in many urban areas than in
rural parishes.
It is worth underlining that much depends upon whether the parish incumbent
lost or gained financially through foreign marriages—on whether the parish
imported or exported such marriages, and the extent to which they were viewed
as clerical currency. Even the location of non-foreign, exogamous weddings may
have been influenced by the fees payable. Foreign marriages, Anglican regis-
tration, and its fees relate crucially to the forsaken topics of clerical incumbency,
income, and the economic history of the Church of England. As Adair remarked,
with reference to the diocese of Ely in the seventeenth century, some ecclesiastical
authorities probably regarded clandestine marriage as ‘a nice little earner’.106 In
rural areas there appears to have been a clear inverse relation between the value
of the clerical living and the number of foreign marriages before 1754. Yet
lateral connections between these historical subjects (with the exception of
101 Quite high occupational endogamy can nevertheless be found in English contexts, for example
in fishing villages, among agricultural labourers, weavers, framework knitters, lead and coal miners.
(This could also involve parochial geographical endogamy, and regional variations in it: some lower-
class workers having more marital choice locally than was true for people in higher status occupations
seeking a bride of similar social standing.) In some cases (e.g. in some textile trades) such endogamy
was related to the need for reciprocal skills between partners and high output from family units of
production. In other cases (e.g. agricultural labourers) this could sometimes be a kind of default
endogamy, women from other occupational families preferring not to marry them, a situation
becoming more pronounced in nineteenth-century England, as with women in some regions of
twentieth-century Ireland. Yet England almost never attained the levels of occupational endogamy
found for some occupations in parts of the Ivory Coast, or China, or among the Masai, or in India
where caste was so intimately tied to occupation.
102 The most striking British examples (comparable to the South Dakota Hutterites) were the
sixteenth-century Familists, the Quakers, and to a lesser extent some General Baptists in the early
eighteenth century. Some of the Methodist denominations (such as the Primitive Methodists) inclined
towards denomination endogamy in certain times and places. In general, partners in Nonconformist
marriages were commonly of the same denomination. It is also worth noting that some early
eighteenth-century ‘foreign’ marriages, as discussed above, may have been of dissenters refused
marriage elsewhere.
103 Information about parishes of residence was not collected for the 404 parishes used in the
Cambridge Group’s aggregate analyses. It appears to be available for some of their parish reconsti-
tutions, although its use was perhaps not envisaged, and the consistency of its transcription and
entry needs to be checked carefully.
104 On the ‘immediate impact’ of Hardwicke’s Act in reducing defective marriage registration, see
also Wrigley and Schofield, Population history of England, pp. 30, 65, and for suggestions of slightly
delayed effectiveness in London, ibid., p. 86.
105 Outhwaite, ‘Sweetapple of Fledborough’, pp. 37, 44-5.
106 Adair, Courtship, illegitimacy and marriage, p. 141.
 Economic History Society 2002

english rural societies and marital endogamy
291
Outhwaite’s work) remain almost wholly neglected by church historians and
demographers as fields of enquiry.
The suddenness of change from 1754, as shown here, may provide reassurance
that continuing declines in marriage age were not a result of gradually improved
registration thereafter of the marriages of minors, whose marriages may some-
times have been omitted hitherto, but were to be more tightly controlled under
the 1753 act.
The decline of foreign marriages also bears upon the ‘reconstitutable minority’
in parish demographic reconstitutions, for it affects the local area from which
couples’ marriages were being registered. In many parishes, the field from which
couples were being registered may have diminished somewhat over time, given
the termination of ‘foreign’ marriages and the gradually enlarging proportion of
endogamous marriages. Exogamous marriages are often lost for family reconsti-
tution purposes, involving many absent male and female marriage ages. Many
women for whom a marriage age is available move out of the parish thereafter,
and the social, regional, and temporal specificity of customs affecting this are
little understood. With significant foreign marriages, and/or high exogamy, such
problems in demographic research become more acute than in situations of high
endogamy. A perfect situation for family reconstitution is total endogamy and
immobility, as transient persons usually become lost with family reconstitution.
High endogamy brings more people into the frame of the reconstitutable min-
ority. The growth of geographical endogamy probably permits fuller completion
of family reconstitution forms over time. To some extent this may counteract
the detrimental effects of growing religious Nonconformity on registration of
baptisms and burials.
Marriage ages may be calculated only for those born in a parish who stayed
to marry there, that is for the ‘stayers’. Ruggles has argued that late marriages
may have a greater chance of being excluded from reconstitutions, because the
chances of migrating before marriage were greater for those who married late
than for those who married early, as the former group were at risk of migrating
for longer.107 Calculations of mean age at first marriage may thereby be under-
estimated. His theoretical argument and its micro-simulation estimates have been
questioned by Wrigley, who has shown empirically (using family-reconstitution
and 1851 census data) that such a ‘migration effect’ on marriage age was
probably very slight, and that marriage ages appear to have been nearly identical
between ‘movers’ and ‘stayers’ in English parish populations.108 However, it is
possible that, for many English regions, 1851 data might not fully characterize
patterns of migration and/or exogamy in the eighteenth century, when unmarried
yearly service and traditional apprenticeship had been widespread, and had mass
repercussions for people’s legal settlements.109 It remains feasible that shifts in
migration behaviour, perhaps reflected in or influenced by changes in parochial
endogamy, could alter balances between late and early marriages in family
reconstitution, such as to influence estimates of mean age at first marriage. For
example, a rise in endogamy (as documented here) might result in growing
107 Ruggles, ‘Migration, marriage and mortality’. On demographic features of migrants and non-
migrants, see also Lesthaeghe, ‘On the social control of human reproduction’.
108 Wrigley, ‘Effect of migration’. Wall, ‘Real property, marriage and children’, p. 463, referred to
unpublished work on 10 parishes by Wrigley which suggested that bridal marriage age may not have
been affected by whether the groom was foreign to the parish.
109 On the decline of yearly service and of traditional seven-year apprenticeships, see Kussmaul,
Servants in husbandry, ch. 6, esp. pp. 97-8; Snell, Annals of the labouring poor, chs. 2 and 5, esp. pp. 74-
80 and 234-41; idem, ‘Apprenticeship system’.
 Economic History Society 2002