British

Archaeology

The voice of archaeology in Britain and beyond

Cover of British Archaeology 94

Issue 94

May/June 2007

Contents

news

Unique decorated jet lozenge from Suffolk matches Stonehenge gold

Heritage white paper praised: but who will pay for it?

Classic jadeite axe may leave UK

What was Roman interest in Silbury Hill?

New dates

In Brief & Phase 2

features

Ringed with the wrecks of slave ships: The Atlantic slave trade
Buy, sell, trade, drown. Jane Webster asks what archaeology can bring to the story

Churches face East, don't they?
Church alignments. Do churches face sunrise on saints' days? Ian Hinton surveys

Excavating Dover's Medieval seasfarers
Keith Parfitt and Barry Corke discover a fishing community

on the web

Recommended websites
Online glossaries and the CBA's Community Archaeology Forum wiki (CAF)

letters

Views and responses

CBA correspondent

Campaigns, comment and communications from the CBA
Adapting archaeology: Gill Chitty looks at climate change and archaeology in the UK

 

ISSN 1357-4442

Editor Mike Pitts

features

Churches face East, don't they?

Armed with a Silva compass fixed to a piece of wood with brass screws, over 10 years Ian Hinton surveyed almost 1,750 churches in England and Wales. He resolved an old belief – but uncovered a new mystery.

When I explained the purpose of my research, the most frequent comment from incumbents, churchwardens and keyholders was a question: "Churches face east, don't they?" Antiquarians over the previous couple of centuries, however, have proposed that churches – that is, the end with the main altar and "east" window – are aligned with everything but east, including Jerusalem or sunrise on the first day of building, but most frequently with sunrise on the feast day of the patronal saint to which the church is dedicated. After surveying 1,747 sites – every rural medieval parish church in 13 counties in England and Wales – I can offer the definitive solution to this question. So what do they face?

East-west mystery

I surveyed details of the church, the size and topography of the churchyard, the surrounding landscape and the elevation of the eastern horizon. I omitted only churches described in volumes of Pevsner's Buildings of England as having had the naves rebuilt, as the buildings' alignments may have changed at that time.

The mean alignment of all the churches in the survey is 86.0° true, four degrees north of east, but this simple figure masks a significant variation east-west across the country. Churches in Cornwall have a mean alignment of 80.4°, compared with the mean in Kent of 92.4°. Grouping the survey into three areas by longitude (Table 1), at one standard deviation the difference between the closest parts of the alignment ranges of churches in the east and west of the country is over 6°. In the west, three quarters of all churches are aligned to the north of east, whereas in the east churches are equally divided between those aligned to the north and south of east.

Table 1 Church alignment by longitude in England and Wales

Area of countryNo.RangeMeanStd DevRange at
1 std dev
% north
of east
West (2°west +)41848–126°82.1°±1.1°81.0–83.2°75
Central 1.01°E–1.99°W79238-128°85.4°±0.8°84.6–86.2°66
East (1°east +)53756-120°90.1°±0.8°89.3–90.9°50
OVERALL174738-128°86.0°±0.3°85.7–86.3°63.5

If the aim was to align churches eastwards, why was there this difference? Individual churches in the east were set out over a range in alignments of 64°, centred within 2° of east, with an overall mean direction of 90°. In the west, the range was slightly greater (78°), but centred within 3° of east, whereas the mean alignment was 8° further north at 82°. Churches in the centre of the country fit neatly in the middle of these results, with a mean alignment of 85.4°.

It appears that the further west a church is located, the more likely it is to be aligned to the north of east. Did the builders have a different focus for alignment in the west? Their overall accuracy was similar to builders in the east of the country, with 81% of western churches aligned within 15° of east, compared with 84% of churches in the east, so why were so many more churches aligned to the north of east?

The greater density of churches in Norfolk allows closer analysis in a single county. The overall relationship between east and west continues here. Dividing the county so that an equal number of churches appear in each half (Table 2), shows a similar variation in the mean alignment between churches in the west (87.5°) and in the east (90.3°) and produces another statistically significant result at 1 standard deviation. Similarly, the proportion of churches that is aligned north of east follows the same pattern – an increasing proportion of churches aligned to the north of east as the analysis moves westwards.

Table 2 Church alignment by longitude in Norfolk

Area of countryNo.RangeMeanStd DevRange at
1 std dev
% north
of east
West (1.10°E or less)27964–128°87.5°±1.3°86.2–88.8°62
East (1.11°E +)27056–120°90.3°±1.2°89.1–91.5°50
OVERALL54956–12888.9°±0.988.089.855.9

The fact that this pattern appears to stretch from Cornwall to Denmark (where Niels Abrahamsen surveyed 204 churches with a mean alignment of 94.2°), and can even be repeated at sub-county level in the UK, makes it seems most unlikely that it was brought about by the unconnected acts of a large number of church builders. It is most improbable that this has happened by chance, but what links them? There appear to be no other factors about the churches in this survey which contribute to this difference.

Construction, such as the range of sizes and floor plans, in the east and west of the country, and in the east and west of Norfolk, is similar; topographical variables, such as the slope of the site they are built on, are similar; they are built in churchyards of similar size and restrictions which might affect their alignment, such as proximity to a boundary or boundaries. It has been shown that differences in magnetic declination (the difference between magnetic north and true north) across the country in medieval times could not have had an effect on church alignment. Had it done so, the results would have been reversed, with higher alignments in the west of the country.

Saintly alignment

The churches I examined are dedicated to over 150 different saints. Much of the analysis that follows concentrates on the most common dedications and those with a single feast day. Churches dedicated to St Mary have been excluded in most cases because, without knowing which of the six feast days was originally celebrated at specific churches, the introduction of additional possible sunrise points for so many churches complicates the issue.

The results are summarised here by grouping churches together whose patronal saints' feast days occur in the same season, dividing the year into four equal segments centred on the solstices and equinoxes. If the individual church alignment was affected by a respect for the feast day sunrise point, then the alignment of those churches dedicated to a saint with a feast day in the depth of winter would tend to cluster towards the south-east; churches with summer dedications would tend to the northeast; and those with equinoctial dedications towards the east. In reality, as Table 3 shows, the mean alignment for the churches in each season is within half a degree of the total mean of 86°, effectively showing no systematic variation at all. (The closeness of these results confirms the significance of the variation in alignments across the uk noted earlier.)

Table 3 Medieval churches by season of their patron saint

Saints' seasonNo.RangeMeanStd Dev% north of east
Winter32654–12086.4±1.362
Summer45947–11886.1±1.061
Equinoctial54838–12885.5±1.066
Sub total133338–12885.9±0.663
No saints day4848–11184.2±3.863
St Mary36656–11686.7±1.164
TOTAL1747 86.0±0.363.5

Higher horizons delay sunrise, making the actual point of sunrise appear further to the south (as the sun moves southwards as it rises). To take this into account, the horizons of 849 churches in four counties, Cumbria, Bedfordshire, Oxfordshire and Norfolk, were measured and the actual sunrise position calculated.

Only one in five churches (20%) is aligned within 15° of its actual saint's day sunrise, whilst half (49%) are aligned more than 30° away from the sunrise point. On the other hand, more than four of every five churches (83%) are aligned within 15° of east, with only eight churches (1%) aligned more than 30° away from east. These figures are confirmed by the graph (page 29), which shows the pattern in Norfolk, with the majority of saints' day sunrises at the extreme ends of the church alignment range, or outside it completely. It has been suggested that church rededication and calendar shift nullify this sort of analysis, but no amount of fiddling with details will ever make Norfolk church alignments match the saints' day sunrises.

Churches of different dedications patently do not align with different sunrise points, they all align roughly eastwards. Read together with the summary "saint's season" results, it is definite that the idea that churches were aligned with sunrise on their patronal saints' day is not true now. Whether a large number of these churches have been rededicated so that they are still aligned towards their original patron saints' sunrise is not known. However, if that were the case, then those churches would have to have been dedicated to saints whose feast day is close to the equinox (sunrise due east, ie 90°), such as St Michael, since the majority of churches are aligned within 10° of east. It would also mean that churches dedicated to some of the most popular saints, whose feast days are close to the solstices – St Andrew, St Nicholas, St John the Evangelist, St John the Baptist and St Peter – could never have been aligned towards their sunrise. It seems most unlikely, on this evidence, that churches ever faced their patronal saints' sunrise.

Whilst churches face generally eastwards, they do not face east specifically. Something must have caused the church-builders to align their churches differently across the country, but neither setting out errors, nor any of the physical, topographical or environmental elements surveyed here, can explain it. This is still an open question: suggestions are welcomed.

Details of the English survey can be found in Antiquaries Journal 86 (2006), 206–26. Ian Hinton took early retirement and is now completing a PhD at the University of East Anglia where he is a continuing education tutor in landscape history and archaeology.

CBA web:

British Archaeology

Jan/Feb 2005
Mar/Apr 2005
May/Jun 2005
Jul/Aug 2005
Sep/Oct 2005
Nov/Dec 2005
Jan/Feb 2006
Mar/Apr 2006
May/Jun 2006
Jul/Aug 2006
Sep/Oct 2006
Nov/Dec 2006
Jan/Feb 2007
Mar/Apr 2007
May/Jun 2007
Jul/Aug 2007
Sep/Oct 2007
Nov/Dec 2007
Jan/Feb 2008
Mar/Apr 2008
May/Jun 2008
Jul/Aug 2008
Sep/Oct 2008
Nov/Dec 2008
Jan/Feb 2009
Mar/Apr 2009
May/Jun 2009
Jul/Aug 2009
Sep/Oct 2009
Nov/Dec 2009
Jan/Feb 2010
Mar/Apr 2010
May/Jun 2010
Jul/Aug 2010
Sep/Oct 2010
Nov/Dec 2010
Jan/Feb 2011
Mar/Apr 2011
May/Jun 2011
Jul/Aug 2011
Sep/Oct 2011
Nov/Dec 2011
Jan/Feb 2012
Mar/Apr 2012

CBA Briefing

Fieldwork
Conferences
Noticeboard
Courses & lectures
CBA Network
Grants & awards

CBA homepage