41
The Roman Rural Settlement Project Preliminary results from the South-East: Settlement and Land-use Dr Martyn Allen
The Roman Rural Settlement
Project
Preliminary results from the South-East:
Settlement and Land-use
Dr Martyn Allen
Project aims and research framework
• A new study of the rural settlement of Roman Britain
• Focus is almost exclusively on the excavated evidence
• Examine temporal and regional variation in settlement evidence from
across England and Wales
• Chronological remit covers the later 1stC BC to the early 5thC AD
Roman Rural Settlement Themes
I. Settlement and land-use
II. The agricultural economy
III. Rural industries
IV. Material culture and identity
V. Ritual and religious practice
VI. Death and burial
Geographic context
of the South-East
region
The South-East region is defined
here as the modern counties of
Kent, Greater London, Surrey, East
Sussex, West Sussex, Hampshire,
Isle of Wight, Berkshire,
Buckinghamshire and Oxfordshire
0
100
200
300
400
500
up to 1949 1950-79 1980-89 1990-99 2000+
No
. R
om
an
ru
ral
site
reco
rds
fro
m t
he
So
uth
-Ea
st
year of report
Inception of
PPG16
The impact of developer-funded archaeology:quantity and distribution of reported Roman rural sites in the South-East
Up to 1949 Up to 1989
Up to 2013
Organisations
0 20 40 60 80 100 120
Oxford Archaeology
Wessex Archaeology
Thames Valley Archaeological Services
Archaeology South-East
Museum of London Archaeology
Canterbury Archaeological Trust
Oxford-Wessex joint venture
Pre-Construct Archaeology
John Moore Heritage Services
Surrey County Archaeology Unit
Cotswold Archaeology
Other commercial units
Local/Regional societies
Museums/Rescue Committees
Universities
State-run/English Heritage
No organisation/private individual
com
me
rcia
l (n
=6
5%
)
no
n-c
om
me
rcia
l
(n-3
5%
)
No. records
Changing perspectives of late Iron Age and
Roman rural settlement in the South-East
• Land at North Bersted, West Sussex
(Thames Valley Archaeological Services -
Taylor and Weale 2009)
• Total excavations covering around 20ha,
sited less than 750m NW from Wren
Crescent
• Late Iron Age to early Roman occupation
• Substantial evidence for poly-focal
settlement occupation, trackways,
enclosures, burial, and elite activity
• Potentially revolutionises understanding
of settlement on the coastal plain
The dataset
Composition of dataset by fieldwork typeCounty Excavation Evaluation Watching Brief Field Survey Fieldwalking Geophysics Total
Kent 120 40 6 1 167
Oxfordshire 115 26 14 1 156
Hampshire 93 19 14 2 128
Buckinghamshire 63 13 6 1 1 1 85
Greater London 54 19 3 76
West Sussex 48 19 1 2 1 71
Surrey 55 9 1 1 1 67
Berkshire 42 9 1 2 54
East Sussex 24 4 3 1 32
Isle of Wight 8 6 14
Total 622 164 49 9 5 1 850
0
20
40
60
80
100
Pe
rce
nta
ge
of
site
re
cord
s
Excavation/SMS Evaluation/Watching brief Non-invasive investigation
Area of land excavated,
and density of sites per hectareOxfordshire
10% Buckinghamshire
5%
Surrey
5%
West & East
Sussex
8%
Hampshire & IoW
12%
Berkshire
6%
Kent
34%
Greater London
20%
0.000.200.400.600.801.001.201.401.60
1000.6 hectares of land excavated
with evidence for Roman rural
settlement in the South-East
Number of sites recorded per
hectare of excavation by county
Roman records vs. NMR dataset*
• Distribution of project sites against the National Monuments Record index of all investigations
• General correlation between Roman rural sites and the local frequency of excavation
• Some gaps in settlement evidence are present, though different factors impact upon regional
distribution*NMR data courtesy of Tim Evans, ADS, University of York
0
2
4
6
8
10
Upper Thames
Valley and Berks
Downs (n=212)
Hampshire/South
Downs (n=108)
South Coast and
IoW (n=97)
North Downs and
Wealden
Greensand
(n=109)
Thames Estuary
and North Kent
Plain (n=98)
High and Low
Wealds (n=38)
Chilterns and
London Basin
(n=190)
pe
rce
nta
ge
of
Ro
ma
n r
eco
rds
ag
ain
st N
MR
da
tase
t
Sub-regional variation in Roman rural settlement data
• Frequency of known Roman rural settlements
appear varied across the South-East
• Results heavily impacted by the considerable level
of intervention in advance of modern development
– e.g. the ‘London factor’
• Does not allow for firm conclusions regarding
Roman settlement density, though the region is not
homogenous
Quantity of site type by county
*Consolidated nucleated settlement records
Small towns No. records
Springhead 4
Staines 14
Neatham 4
Villages
Abingdon 6
Ewell 6
Brentford 3
Hassocks 2
Old Ford 2
Shadwell 2
Croydon 3
Thatcham/Newbury 2
• Remarkable variation in the range of settlement types and features present
• Site categories still mask further variation in the settlement record, e.g. ‘farm’ constitutes a range of
settlements with differing forms, functions, and socio-economic status
• Each site record may consist of one or more site type, e.g. farms which develop into villa establishments or
nucleated settlements which include cemeteries and/or shrines
• However, nucleated settlements may consist of a number of site records, in order to observe spatial and
temporal variation within each*
Major category Site type Oxon Buck Berks Hants IoW W. Sussex E. Sussex Surrey Gr. London Kent Total
Settlement and
Associated Landscape
farm 82 49 33 68 5 30 12 26 44 90 439
villa 21 13 2 19 5 14 4 12 2 27 119
field system 30 11 18 16 3 18 1 5 20 38 160
village* 10 2 3 4 2 1 1 5 3 31
unwalled small town* 1 1 1 1 1 5
defended settlement 1 2 1 1 2 1 8
oppidum 3 2 5
agricultural/isolated
building 3 11 1 2 1 6 24
other military 1 1 2
sub-total 151 86 57 112 13 68 20 46 74 166 793
Religious/Ritual
shrine 8 6 5 7 2 6 2 15 50
funerary site 28 21 4 12 3 3 6 12 43 132
RC temple 3 2 1 1 3 1 3 8 22
sub-total 39 29 5 18 0 13 6 15 14 66 205
Communications/
Infrastructure
road 3 2 4 1 4 7 11 32
jetty/bridge 1 1 1 1 1 4 2 11
mansio 1 2 1 1 5
sub-total 4 3 1 5 0 3 1 5 12 14 48
Industry
mill 1 1 1 3
iron production 3 10 1 1 3 7 1 8 34
other metal production 3 3 1 1 1 9
pottery production 17 8 15 3 2 4 3 6 58
quarry 6 5 1 12 1 3 2 18 9 57
salt production 2 1 4 7
tile production 1 3 2 1 4 1 3 15
other industry 1 4 1 1 2 1 1 11
sub-total 30 27 6 35 2 11 13 13 25 32 194
Total 224 145 69 170 15 95 40 79 125 278 1240
Settlement chronology,
morphology and transformation
Late Iron Age settlement and the evidence for
continuity/change in during the conquest period
• 280 farming settlements (53%) occupied
during the late Iron Age, spread widely
across the region (green dots)
• 8% of LIA settlements abandoned and 7%
fundamentally changed in site type/form by
c.AD50
% LIA farms abandoned by
mid1stC AD
% LIA farms with major
changes in settlement
type/form
% LIA continuing unchanged
into late 1stC AD
continuity
change
New farming occupation sites in the conquest period,
and beyond
• 131 farm/villa sites (24.8%) originate
c.AD40-75
• Relatively wide distribution
• Clusters found close to major settlements
and/or arterial routes – possible factors in
stimulating new settlement?
0
5
10
15
20
25
30
35
40
pe
rce
nta
ge
of
farm
/vil
la s
ite
s
est
ab
lish
ed
(n
=3
64
)
Date of establishment of farming settlements
• Most farming sites established in the
1stC AD, predominantly in the second
half of the century
• Comparatively few new settlements
established from the later 2ndC AD
Frequency of farm/villa occupation
by period of use
0
100
200
300
400
500
No
. fa
rms/
vil
las
in u
se
period of occupation
• Increasing number of sites settled from the mid-late Iron
Age, and reduction after 2nd C AD
• Questions remain over the stimulus for increasing and
decreasing settlement, i.e. increasing population, trade
and exchange, etc.
• Pattern masks regional variation
Settlement forms:
Unenclosed/Open
settlement
• Evidence for occupation not obviously
contained within a system of enclosure
• Enclosures may still be present on site,
but do not bound primary areas of
domestic activity (e.g. Moor Hall Farm &
Imperial Sports College)
• Large area of excavation generally
needed to demonstrate open settlement
Imperial Sports College, Harlington
Settlement forms:
Enclosed
settlement
• All or majority of domestic
activity contained within 1 or 2
enclosures
• Internal space not sub-divided to
a significant degree
Settlement forms:
Linear/Developed
settlement
Surrey County Archaeology Unit
• A complex of conjoined
enclosures
• Internal area often extensively
sub-divided
• Multiple areas used for domestic
activity
• Trackways and field-systems tend
to be incorporated within
settlement system
Reconstruction of a linear/developed Romano-
British farm showing divisions of internal space
Domestic
Stock
enclosure
Crop
processing
Industry
Ritual?
linear/
developed
14%
enclosed
28%
unenclosed/
open
4%
unclassified
54%
no. farm sites classified
by settlement form = 450
Cranford
Lane,
Harlington
Thames Valley Park,
Reading (W164)
Frequency of farmstead
sites in south-east by
settlement form
NB. some sites may include more than one
settlement form due to transformation
Westhampnett,
Area 5
Long term trends in settlement morphology
0.0
10.0
20.0
30.0
40.0
Pe
rce
nta
ge
of
site
s w
ith
cla
ssif
ied
se
ttle
me
nt
form
by
pe
rio
d o
f u
se
unenclosed/
open (n=21)
enclosed
(n=126)
linear/
developed
(n=61)
change changerelative
continuity• Long term trends in the frequency of
settlement forms show an increase in
linear farms and a decrease in
enclosed farms from the late Iron Age
to the late Roman period
• Open settlements are largely invisible
after the late Iron Age, whilst the
frequency of enclosed settlements
reduces c.10% across the same period
• Linear farms overtake enclosed as the
dominant settlement form by the 4thC
AD
Settlement transformation patterns
0
10
20
30
40
50
60
becomes
enclosed
becomes
linear
becomes a
villa
farm to field-
system
hiatus in
occupation
nu
mb
er
of
site
s
type of transformation where previously existing settlement
was observed
• Villas highly visible archaeologically – most (though
not all) have pre-villa settlement phases
• Suggests significant changes in wider land-use and
local socio-economic networks
• Other forms of settlement transformation likely to
be under-represented compared to villa
developments
• Pattern suggests that settlement transformations
may have been more widely occurring Barcombe villa, West Sussex
• Moor Hall Farm, Greater London
• Late Iron Age open settlement develops over
1st C AD into an enclosed farm
Settlement enclosure
Linear settlement
development
0
10
20
30
40
50
from
unclassified
farm
from open
settlement
from enclosed
farm
virgin
settlement or
unclear origins
no
. si
tes
• At least 13 sites demonstrate clear development
from other settlement forms to linear farms
• Other linear farms either have unclear origins or
were ‘virgin’ settlements
• Transformations to linear settlements happen
throughout the Roman period, but mostly commonly
occur during the late 1st C AD (64%)
Variation in linear settlement forms
Rowbury Farm, Hampshire:agglomeration of multiple enclosures
Wavendon Gate, Buckinghamshire:major internal divisions of space
Distribution of linear farms
• Clear predominance in the north and east of the region
• No clear association with major urban settlements
Thames Valley hinterland distribution of linear
farms
• Mostly found on the superficial geology, the sand and gravel river terraces
• Due to large-scale quarrying excavations revealing larger settlements?
• Following the main river
alignments, the Thames, the
Kennet, and the Colne
• Also found along the route of
Akeman Street
Thames Valley
Region
Geographic spread of villa
construction• 1st C AD villa construction focussed on north Kent and the Sussex
coastal area
• Villa construction spread far more widely by the 3rd C AD,
particularly across the North Downs into Hampshire and through
Goring Gap
• Clear switch from localised development (l.1st C AD) to wider
spread development (3rd-e.4th C AD)
• Few concentrations close to major urban settlement, perhaps
apart from Chichester
Villa chronologies
• 64.7% villas with known date of earliest construction (out of 119 villas
in total)
• Large proportion built during the late 1st C AD with most new sites
appearing in the 2nd C AD
• After the early 2nd C AD villas normally develop from pre-existing farms
• Highest proportion of villas occupied during third and fourth centuries
AD
-100
0
100
200
300
400
500
Fis
hb
ou
rne
Pa
lace
An
gm
eri
ng
vil
la
Ba
x F
arm
, T
eyn
ha
m
Th
urn
ha
m
'Sp
es
Bo
na
', L
an
gst
on
e
Ecc
les
Co
mb
e E
ast
En
d
No
rth
fle
et
vil
la
Fa
ve
rsh
am
Min
ste
r-in
-Th
an
et
Be
dd
ing
ha
m
We
st P
ark
, R
ock
bo
urn
e
So
uth
wa
y,
Ne
wh
av
en
Ma
no
r H
all
Ro
ad
, S
ou
thw
ick
Ch
ich
est
er
Ha
rbo
ur,
Fis
hb
ou
rne
Da
ren
th
Ch
urc
hfi
eld
s, S
no
dla
nd
Fa
rnin
gh
am
II
Sa
nd
wic
h v
illa
Fra
nks
Ha
ll, F
arn
ing
ha
m I
II
Sh
ak
en
oa
k F
arm
De
ert
on
Str
ee
t, T
eyn
ha
m
Bra
din
g
Co
bh
am
Pa
rk
Bu
ry C
lose
, F
aw
ler
Dit
chle
y P
ark
Alf
red
's C
ast
le,
Ash
bu
ry
Sta
nto
nb
ury
(M
K3
01
)
Ba
ncr
oft
vil
la
Go
rin
g v
illa
, W
est
Du
rrin
gto
n
Ea
st W
ea
r B
ay
, F
olk
est
on
e
Wym
bu
sh (
MK
21
1)
Sta
nto
n L
ow
Go
sde
n R
oa
d,
Litt
leh
am
pto
n
All
en
's F
arm
, P
laxt
ol
Big
no
r R
om
an
vill
a
Cro
fto
n,
Orp
ing
ton
Sa
un
de
rto
n (
He
mle
y H
ill)
Se
lse
y R
oa
d,
Sid
lesh
am
Lati
me
r
Hig
h W
yco
mb
e
Ma
ntl
es
Gre
en
, A
me
rsh
am
Hu
ll P
lace
, S
ho
lde
n
Th
e M
ou
nt,
Ma
idst
on
e
Lull
ing
sto
ne
Co
x G
ree
n,
Ma
ide
nh
ea
d
Sh
av
ard
s F
arm
, M
eo
nst
oke
Wa
lto
n-o
n-t
he
-Hil
l
Ho
ug
hto
n D
ow
n,
Lon
gst
ock
Be
dd
ing
ton
Sp
ars
ho
lt R
om
an
vil
la
Ke
sto
n
Th
ruxt
on
Ba
tte
n H
an
ge
r, W
est
De
an
Sh
ide
, N
ew
po
rt
Ra
psl
ey,
Ew
hu
rst
Ga
teh
am
pto
n F
arm
, G
ori
ng
Ba
rco
mb
e
Ch
ilto
n F
ield
s
Wh
ite
Be
ech
, C
hid
din
gfo
ld
Fu
lle
rto
n
Ca
stle
Hil
l, W
itte
nh
am
s
Did
cot
We
st (
Gre
at
We
ste
rn P
ark
)
Bro
ad
Str
ee
t C
om
mo
n,…
Ha
le R
oa
d,
Fa
rnh
am
Litt
le C
ha
rt
Wig
gin
ton
De
nch
wo
rth
Ro
ad
, W
an
tag
e
Du
nk
irt
Ba
rn,
Ab
bo
tts
An
n
Ye
wd
en
Ro
ma
n v
illa
Ba
rto
n C
ou
rt F
arm
Bra
md
ea
n
Ch
ilg
rove
1,
Bri
ck K
iln
Fa
rm
Gra
tele
y S
ou
th,
Gra
tele
y
Oa
kri
dg
e V
II,
Ba
sin
gst
oke
Pit
lan
d F
arm
, U
pm
ard
en
Ch
ilg
rove
2,
Cro
ss R
oa
ds
Fie
ld
BC
/AD
date of villa construction period of occupation
Settlement development:Beddington, Surrey
Late Iron Age settlement (enclosed?) Early Roman co-axial field-system
Mid-Roman villa Late Roman villa
Patterns of wider land-use
Distribution of field-systems
• More widespread distribution than linear farms
• Apparent clustering of ‘field-system areas’: Thanet; south Kent; coastal
plain; Thames estuary; Middle Thames; Upper Thames Valley
Field-system forms
aggregated
7%
co-axial
23%
both
3%
uncertain
form
67%
no. sites with evidence
of field-system = 161
Co-axial fields
• consistent alignment of fields
• number of fields follow the same land boundary
• evidence for planned development
• little attention paid to local topography
Aggregated fields
• piecemeal/organic development
• not necessarily on alignment
• tend to pay attention to local topography
Fullerton, Hampshire
Highstead, Chislet, Kent
Field-system chronologies:when were field-systems in use and how long for?
0.0
25.0
50.0
75.0
1st C BC 1st C AD 2nd C AD 3rd C AD 4th C AD 5th C AD
pe
rce
nta
ge
of
fie
ld-s
yst
em
sit
es
(n=
16
1)
Period in use
• Many lack accuracy in dating with regards to
construction and abandonment (c.16% cannot
be dated beyond ‘Roman’)
• Periods of use are more clearly identifiable,
particularly on sites with changing layout and
developments
0
10
20
30
40
50
60
70
</=100 years 101-200 years 201-300 years >301 years
nu
mb
er
of
fie
ld-s
yst
em
sit
es
Length of time field-system in use
• Field-systems normally in use for relatively
long periods
• Most appear to be in use between 100 and
200 years, forming distinct chronologies
Field-systems and associated settlement forms
• Most sites with field-systems either
associated with farms or have no directly
associated settlement recorded
• Nucleated settlements most frequently
associated with field-systems
• Reinforces the rural character of villages
and small towns via associated pastoral
and/or agricultural activity
*nucleated settlement sites amalgamated
0
20
40
60
80
farm villa nucleated
settlement*
other
association
no association
no
. fi
eld
-sy
ste
m s
ite
s
associated site type
0.0
10.0
20.0
30.0
40.0
farm villa nucleated
settlement*
other
association
no association
% f
ield
-sy
ste
m s
ite
s
associated site type
Field-systems and nucleated settlement
Excavation at Syon Park, Brentford, Greater London (MoLA 2009)
• Trackway linking roadside settlement at Brentford, along the London-Silchester
Road, to Thames palaeochannel
• Co-axial field-system developed alongside the trackway, supporting the settlement
Identifying activity in the fields: livestock
Excavation of a waterhole at Runfold Farm, Farnham.
Surrey Archaeological Unit
0
3
6
9
12
15
18
farm farm with
field-system
field-system village villa other
no
. si
tes
wit
h e
vid
en
ce o
f
wa
terh
ole
s
enclosed
21%
linear
47%
unclassified
32%
sites with waterholes (n=34)
• Waterholes dug for livestock water-
supply
• Common association of waterholes with
farms with related field-systems
• Waterholes more common on sites of
linear settlement form (more than twice
the frequency of enclosed farms)
Identifying activity in the fields: hay meadows
ID site county site type
16039 Farmoor Oxfordshire farm/field-system
11002 Penfold Lane, Rustington Bypass West Sussex farm/mill
10015 Staines, former Central Trading Estate Surrey small town
15040 Clatterford Roman villa Isle of Wight villa
15018 Dunkirt Barn, Abbotts Ann Hampshire villa
15015 Fullerton Hampshire villa/field system
16037 Denchworth Road and Mill Street, Wantage Oxfordshire villa/field-system
16067 Appleford Oxfordshire village/field-system
9019 Westhawk Farm, Ashford Kent village/field-system
Sites with evidence for hay meadows
Yellow rattle Rhinanthus minor
• Range of site types with evidence for hay meadows
• Villas appear well represented
• Yellow rattle remains found in some environmental assemblages
• Hemi-parasitic species commonly associated with hay meadows
• Seeds effectively spread by traditional hay-making practices
Identifying activity in the fields: cereal cultivation
0
10
20
30
40
50
1stC BC 1stC AD 2ndC AD 3rdC AD 4thC AD
%
period in use
percentage of field-system sites with evidence
for cultivated plant remains (n=161)
0
10
20
30
40
50
hulled wheat free-threshing
wheat
barley other minor
crop species
%
cereal taxa present
percentage of field-system sites with
cultivated plant remains by taxon (n=161)
• Chronological frequency of field-system sites
with archaeobotanical evidence is similar to
overall pattern of field-system use
• However, difficulties found in generating
good environmental assemblages from field-
system contexts
• Impact on plant taxa present due to selection
and processing practices
Concluding points
• Developer-funded archaeology has had a clear and profound impact on our
understanding and perspective of Roman rural settlement in the South-East region
• Rural settlement pattern does not appear uniform across the region
• Settlement forms are clearly varied, whilst generally distinctive
• Concentrated development of different settlement type appears to occur in
different places, i.e. villas on south coast and north Kent; developed farms in
Thames Valley
• Evidence for further land-use and field-systems are wide-spread though clustering
of the latter occurs
• Future work may be directed towards further identification of the form and layout
of field-systems, and on identifying and securely dating the activities carried out
within them
