THE PREHISTORY OF BRITAIN AND IRELAND pt.2

A NEW BEGINNING
TWO MODELS
This chapter discusses the beginnings of the Neolithic period in Britain and
Ireland and is concerned with the time between approximately 4000 and
3300 BC. That simple statement raises many problems (Figs. 2.1 and 2.2).
In order to work out when the Neolithic started, it is necessary to establish
what that term means today. It is a period label which has been inherited
from the past and has had rather different connotations from one generation to
another. The adjective ‘Neolithic’ was originally devised to describe a particular
kind of technology based on the use of ground and polished stone, although
it was soon appreciated that the finished artefacts were often found with
ceramics. This definition became less important once it was discovered that
these innovations occurred at the same time as the adoption of domesticates.
Gordon Childe (1952) even spoke of a ‘Neolithic Revolution’, a termwhich he
intended to evoke the important social and economic changes associated with
the development of farming. Still more recently, attention has shifted to the
idea that the Neolithic period also saw profound changes in human attitudes
to the world and that these were reflected in the construction of monuments
and the use of a more complex material culture (Hodder 1990).
Each of these definitions has a certain validity. The problem is how their
different components might be integrated with one another. Thus the oldest
characterisation of the Neolithic period works well in England and Scotland,
but in Ireland, and possibly in Wales, ground stone artefacts had already been
made by hunter gatherers (Costa, Strenke, and Woodman 2005; David and
Walker 2004). Monuments do seem to be a new development (Bradley 1998b),
but again they cannot characterise the entire Neolithic phenomenon. In most
regions they were not built at the very beginning of this period, and there
are some areas in which they were never adopted on a significant scale. For
example, in southwest Ireland, there is evidence for Neolithic houses, but the
conventional range of stone-built tombs is missing and part of the population may have been buried in caves (O´ ’Floinn 1992). That also happened in regions
like northern England which had a tradition of building mortuary monuments
(Barnatt and Edmonds 2002). In the same way, recent fieldwork in Ireland, and
to a lesser extent in Scotland and Wales, has produced convincing evidence of
houses and cereal growing, but in England this is relatively rare (Grogan 2004a;
Darvill 1996). It seems as if different definitions of the Neolithic period apply
to each of these islands, and yet their material culture shares many features in
common. In particular, the ceramic sequence begins with a series of finely made
undecorated vessels which are normally grouped together as Carinated Bowls.
They were used until about 3700 BC alongside another ceramic tradition
that may first have developed in Brittany (Herne 1986; Cowie 1993; Sheridan
2003a). The development of regional styles of decorated pottery did not happen
on a significant scale until later. In much the same way most parts of Britain
and Ireland include some of the same kinds of stone artefacts.
A more serious problem involves the question of explanation. How did the
insular Neolithic originate? Did it begin among the indigenous hunter gatherers
who are described in the conventional terminology as ‘Mesolithic’? How
was it related to what was happening during the same period in Continental
Europe, and to what extent did it develop a distinctive character of its own?
In many ways the evidence from Ireland comes closest to the accepted
definition of the Neolithic period (Cooney 2000). Here there were changes
in the nature and distribution of settlement and in the character of material
culture. Substantial houses are widely distributed and are commonly associated
with cereals and domesticated animals. Field systems are known in the west of
the country, and over its northern half there is a dense distribution of mortuary
monuments. In southern England, on the other hand, the situation is less
clear-cut. There are local differences between the distributions of Mesolithic
and Neolithic occupation sites, but this is not always the case. In regions like
the wetlands of the Somerset Levels (Coles and Coles 1986: chapter 4), wild
resources were still exploited, and there may also have been some continuity in the ways in which stone artefacts were made. In the south the evidence for
domestic buildings is very limited indeed (Darvill 1996). Cereals are usually
found in small quantities along with evidence of wild plant foods (Moffet,
Robinson, and Straker 1989; Robinson 2000), and the material remains of the
period are dominated by a series of stone and earthwork monuments which
seem to have played a role in public events and in the treatment of the dead
(Bradley 1998b).
Such stark contrasts do not extend into every area, but they have given
rise to an important debate between the proponents of an ‘Irish/Northern
British’ model which sees the Neolithic period as a radical break with the
past (Cooney 2000: chapter 7), and a ‘Southern English/Wessex’ model
which stresses continuity, acculturation, and a mobile pattern of settlement
( J. Thomas 1999). Scottish prehistorians have resisted the influence of the
English interpretation and have taken exception to the idea that whatever happened
in the rich archaeology of the south and east must have established the
pattern for both islands (G. Barclay 2001).
These differences extend to the ways in which the archaeological record has
been interpreted. Irish prehistorians consider that at the start of the Neolithic
period an agricultural economy was introduced by settlement from overseas.
Every aspect of the archaeological record seems to have changed quite rapidly,
suggesting that the indigenous population soon adopted a new way of life.
This model is much the same as that applied to parts of mainland Europe,
where it seems as if the northwards spread of farming was accomplished by the
settlement of new land (Whittle 1996: chapters 6 and 7). That process had been
checked in the early fifth millennium BC, but later it could have resumed. To
the west it proceeded without a break, and it may be significant that there is
a general resemblance between the first megalithic monuments in Ireland and
along the Atlantic fac¸ade of Britain and those in northern and western France
(Sheridan 2003b). These areas may have been among the sources of a new
population. It seems possible that this development began before a Neolithic
way of life was established elsewhere.
By contrast, the ‘Wessex’ model has been influenced by research in Northern
Europe, where the Neolithic period began at virtually the same time as it
did in Britain ( J. Thomas 1988). Again it involved the creation of impressive
monuments to the dead, but in this case the remains of houses do seem to
have survived. Economic changes took place much more gradually, and in
south Scandinavia it seems that at first the settlement pattern was based on
domesticated animals. Arable farming on any scale did not appear until later.
It seems likely that the early exploitation of domesticates ran in parallel with
the continued use of wild resources. The commonest interpretation is that the
indigenous people adopted elements of a new way of life through a long period
of contacts across the agricultural frontier to the south (Fischer 2002). This is
suggested by the presence of pottery and imported axes on Mesolithic sites and may have been related to the development of increased social complexity
among the local population. The idea is certainly supported by the evidence
of cemeteries. In this case the movement of ideas may have been as important
as the movement of farmers. The southern English model is explicitly based on analogy with the sequence
in Scandinavia (J. Thomas 1988), and it is true that the Neolithic of these regions
could have had some features in common: an initial emphasis on mobility and
the use of domesticated animals; a settlement pattern which in some regions
has resulted in the survival of rather insubstantial houses; the formal deposition
of elaborate artefacts in pits or similar features; and the construction of various
kinds of earthwork monuments including long mounds and, eventually, enclosures.
The problem is that these comparisons apply to the Neolithic end of
what was a much longer sequence. How does the British and Irish Mesolithic
compare with its North European counterpart?
The Scandinavian model depends on several key features: a long period of
contact with areas to the south, shown by the movement of nonlocal artefacts
and the adoption of an unfamiliar technology; reduced mobility and possibly
the year-round occupation of certain favourable locations; increasing evidence
of social complexity suggested by the development of cemeteries and votive
deposits; and the introduction of small quantities of cereals and domesticated
animals to settlements whose economy was otherwise based on hunting, gathering
and fishing. These processes seem to have been vital to the adoption of
agriculture and to the changes that went with it.
As Peter Rowley-Conwy (2003; 2004) has observed, there are few signs of
these features in Mesolithic Britain and Ireland. If anything, the best evidence
for social complexity during the Mesolithic comes from the beginning of that
phase, and so it can have no bearing on the origins of farming in these countries.
The largest domestic buildings all belong to the early Mesolithic, as do the one
convincing cemetery at Aveline’s Hole and the possible indications of others.
The only place with much evidence of ritual activity is the early site at Star
Carr (Conneller 2004), and the sole indications of a complex material culture
are provided by a series of decorated artefacts which date from the first part of
this period. It even seems possible that the settlement pattern in Britain and
Ireland became more mobile during the later Mesolithic (Myers 1987; Costa,
Strenke, and Woodman 2005). At present there is a single site, at Ferriter’s
Cove in southwest Ireland, where domesticated animals have been identified
in a secure Mesolithic context dating from about 4300 BC (Woodman, Finlay,
and Anderson 1999), yet the first unambiguous evidence of cereal pollen on
either side of the English Channel dates from between 4050 and 3850 BC(Innes,
Rowley-Conwy, and Blackford 2003). The earliest traces of cultivation with
the primitive plough known as an ard occurs even later, between about 3500
and 3250 BC (Sherratt 1997: fig. 3.2). Still more important, the settlements
which do contain a mixture of Mesolithic and Neolithic material culture are
largely found along a restricted section of the Scottish coastline, but there is
nothing to prove that both assemblages were in use simultaneously (Wickham-
Jones 1990; Armit and Finlayson 1992). Although this could be compared with the evidence from Northern Europe, it seems to be a local phenomenon, and
one which may even have happened some time after the beginning of the
Neolithic period. It cannot support any model with a wider application.
Indeed, there are signs of discontinuity at the end of the fifth millennium BC
and the beginning of the fourth. One of the features that was once supposed
to indicate greater control over the environment during the later Mesolithic
period was deliberate burning of the natural vegetation (Simmons 1996). This
might have increased the food supply and could certainly have helped to concentrate
game animals in particularly favourable locations. It is now clear that
this was being done from the very beginning of the Mesolithic. If it happened
increasingly often, that would be expected as the landscape was invaded by
trees. In no sense can this be thought of as a ‘pre-adaptation’ to farming, and at
the beginning of the Neolithic this practice came to an end (Edwards 1998). In
the same way, there is new evidence for human diet from the analysis of human
bones found near to the sea. This work suggests that during the Mesolithic
period coastal communities had made significant use of marine resources, but
at the beginning of the Neolithic it was significantly reduced. From that time
much greater use was made of terrestrial plants and animals. The sample is
limited because few Mesolithic bones survive, but the results of this work can
be matched along the Atlantic coastline from Portugal to Brittany and also in
south Scandinavia (M. Richards 2003; 2004).
There is another revealing trend in the environmental record. To a large
extent the beginning of the Neolithic period corresponds with a dramatic
decline in the amount of elm pollen. It no longer seems likely that this is
the result of a specific economic practice, like ring-barking growing trees or
collecting leaf fodder. It is more likely that the phenomenon results from the
rapid spread of a disease which attacked this particular species (Girling and
Grieg 1985). It happened over such a restricted period of time that the ‘Elm
Decline’ has been used as a chronological marker, and yet it seems likely that
the spread of the disease itself was facilitated by the movement of people on an
increasing scale. It may also have been accelerated by the creation of more areas
of open ground. According to a recent study this phenomenon has a mean date
of 3940 BC (Parker et al. 2002).
The earliest radiocarbon dates for Neolithic artefacts and monuments begin
around 4000 BC, and again there seems to have been little or no chronological
difference from one end of these islands to the other. These dates raise certain
difficulties. This was a period where the calibration of radiocarbon dates
is not precise, meaning that the true ages of particular samples have a wide
distribution. As a result it is not clear whether the inception of the Neolithic
was more like an event than a gradual process. One possibility is that it took
place over three centuries or more, starting at about 4000 BC. Another is that
the main period of change was rather later and that events moved at a faster pace. It would certainly be implied by the direct dating of carbonised cereals,
which suggests that they appeared around 3800 BC (Alex Brown, pers.
comm.). Another difficulty is in deciding when Mesolithic material culture
went out of use. There are too few dates to shed much light on that process,
but they do show that in Northern England the use of microliths extended
into the very beginning of the fourth millennium BC (Spikins 2002: 43).
However this evidence is interpreted, the British and Irish Neolithic apparently
draws on two sources. There are a number of obvious links between
the insular record and the evidence from Continental Europe. Perhaps this has
been obscured because scholars have been looking for a single origin for insular
Neolithic artefacts. This has not been found, although there are general links
between the undecorated Neolithic pottery of northern France, Belgium, and
the southern Netherlands and the earliest pottery in Ireland and Britain (Louwe
Koojimans 1976), just as there may have been further connections with northwest
France (Sheridan 2003b). It is less clear how the insular flint industries are
related to those on the Continent, but in this case some of the variation may
result from the use of different raw materials. Far more important are two other
elements whose significance is not disputed. The main domesticated animals of
the British and Irish Neolithic were cattle, sheep, goats, and pigs. There were
no cattle in Ireland during the Mesolithic period, and, in Britain, the native
wild cattle were quite distinct from the animals that first appear in the Neolithic
period, which have much more in common with those in mainland Europe
(Tresset 2003). Pigs could have been domesticated locally, but there are no wild
progenitors for sheep or goat in either island. The same applies to wheat and
barley. These staples must have been introduced from the Continent, although
no traces of seagoing vessels survive.
The second component is monumental architecture, which seems to have
become an established feature of the landscape from an early stage of the
Neolithic, although this may not have happened immediately (Bradley 1998b).
Again it is probably unwise to seek a single source of inspiration for any of
these structures, but the tradition of long mounds or ‘long barrows’ that is so
widespread in Britain and Ireland was distributed along the edges of the Continent
as far east as Denmark and Poland and as far west as Brittany. Some of the
insular structures contained stone chambers of forms that are not unlike those
in Atlantic Europe (Sheridan 2003b). In the same way, distinctive earthwork
enclosures defined by an interrupted ditch had a very wide distribution on
the Continent, reaching from Bornholm to the east to Ireland to the west. It
also extends as far south as Spain. They are particularly common in Denmark,
northern Germany, Belgium, and the north of France. Many of these monuments
are remarkably similar to one another (Bradley 1998b: chapter 5). They
also bear a strong resemblance to those in southern Britain and may have been
used in the same ways, although in these islands they were first built some time
after the earliest long barrows. Since such connections are widely accepted, why has it been so difficult to
define the sources of the British and Irish Neolithic? Perhaps this is because
of the islands’ distinctive geography, which allowed them to form links with
regions of the European mainland that would not have been in regular contact
with one another. That may be why the insular Neolithic was so distinctive
even when it was obviously influenced by practices in Continental Europe. It
drew on many different sources of inspiration. The problem for future research
is not to discover the source of the insular Neolithic but to explain how so
many diverse influences were drawn together in a new synthesis over a short
period of time.
Why did Britain and Ireland ‘become Neolithic’? It is certainly true, as
Julian Thomas (1988) recognised, that this process happened simultaneously
across a considerable area. It affected southern Scandinavia at the same time as
these two islands. Since there is no evidence for contacts between Britain and
Northern Europe during the later Mesolithic period, it suggests that the initial
impulse must have begun on the Continent. Four different models have been
proposed, and, logically, they should apply not just to Britain and Ireland but
to Denmark, Sweden, and The Netherlands as well.
The first is that in each of these areas indigenous hunter-gatherers were
well aware of farming communities and their practices but were reluctant to
change their way of life. That reluctance might have been because of deeprooted
reservations about Neolithic systems of belief (Bradley 1998b: chapter
2), or it could have been because the local environment was so productive
that there was little need to experiment with new ways of producing food. In
the first case, their growing exposure to Neolithic customs and material culture
might have led to internal tensions and the growth of social inequality. If so, the
adoption of farming could have been a social strategy designed to build prestige.
The second possibility is that the adoption of domesticates was the result of
adverse circumstances as the economic prosperity of local hunter-gatherers
was threatened by failures of the food supply (Zvelebil and Rowley-Conwy
1986). Each argument is plausible in itself, but neither seems to have a direct
bearing on the prehistory of Britain and Ireland. There is very little evidence
for contacts between insular hunter-gatherers and Continental farmers, and
there are no signs of social complexity in the insular late Mesolithic. Rather,
Ireland became isolated from Britain and Britain became isolated from the
European mainland.
Another observation is that Continental farmers had little experience of the
sea (Dennell 1985: 174). That is logical since it appears that agriculture spread
across the Continent from southeast to northwest. This process was probably
accomplished through the movement of people whose only experience would
have been of travel overland. That might suggest that the colonisation of Britain
and Ireland would have been delayed until the settlers had familiarised themselves
with marine navigation. It is a plausible argument, but it fails to account for one important factor, as the insular Neolithic began simultaneously with
that in the Netherlands and much of Denmark, both of them regions that
could have been settled without the use of boats. Another element may have
been involved.
In Northern Europe the territory occupied by early farmers had remained
much the same over a long period of time. Why were they so reluctant to settle
larger areas? Perhaps the reason that the initial expansion of agriculture seemed
to falter was that it had reached the limits of the land where it could produce
a dependable supply of food. Beyond it there were regions with unfavourable
weather and poorer soils. Recently Clive Bonsall and his colleagues have
observed that the expansion of early farmers into just those areas seems to have
occurred as the climate changed, so that it would have been possible to practice
the same economy over a more extensive territory (Bonsall et al. 2002).
This may be a coincidence, but it does suggest an incentive for expansion
that had been lacking before. It also seems to be the only model which
explains developments in Britain and Ireland as well as those in Scandinavia.
Although it is expressed in very general terms, it provides a useful working
hypothesis.
Colonisation by sea is currently an unfashionable option, and it certainly does
not account for every aspect of the insular Neolithic. There are also a number
of features that could be more closely linked to a Mesolithic way of life. The
most important may also be the most intangible. Neolithic material culture was
adopted remarkably quickly over large parts of Britain and Ireland. Perhaps
this was influenced by existing knowledge of insular geography. Extended
journeys along the rivers and around the coastline might have depended on
the accumulated experience of the indigenous population. That could explain
why some of the Neolithic ceremonial monuments were built in places which
had already fulfilled an important role in an older pattern of communication.
It applies to the location of certain enclosures and tombs on sites that had been
used before, but, provided clearings were maintained by grazing animals, it
need not presuppose an unbroken sequence of human activity. The same may
apply to the use of other places in the landscape, and it is has been observed
that Mesolithic occupation sites at strategic places in the uplands often include
finds of Neolithic arrowheads (Spratt 1993: 77). These sites may have been
used by hunting parties, but it remains to be seen whether the two groups of
artefacts were contemporary with one another.
There was also a certain continuity in the use of woodland. The wellpreserved
timber trackways in the Somerset Levels show that coppicing was
important from an early stage (B. Coles and J. Coles 1986: chapter 3), and Tony
Brown (1997) has suggested that many of the earliest settlements were located
in natural clearings where the tree cover had been reduced by lightning strikes
and strong winds. Such places would have attracted animals and light-loving
plants on either side of the conventional distinction between the Mesolithic and Neolithic periods. The real change happened only when people made a
deliberate decision to increase the area of open ground.
That is less of a problem on those coastal sites where it is clear that shell middens
continued to accumulate after 4000 BC, but this was a local phenomenon,
confined to western Scotland and the Inner Hebrides (Armit and Finlayson
1992). Another intriguing convergence is represented by the use of raw materials.
It seems clear that in a number of regions, including parts of the southern
English chalk (Gardiner 1989) and the Scottish islands of Arran and Rhum,
stone sources that had first been used in the Mesolithic period were exploited
on a more intensive scale during the Neolithic (Wickham-Jones 1990). In the
same way, coastal hunter gatherers in northwest England had worked small
pieces of a distinctive kind of tuff which they probably collected from stream
deposits. During the Neolithic period the parent outcrop was discovered in
the Lake District mountains and became one of the largest sources of polished
axes (Bradley and Edmonds 1993). It is unlikely that this happened by chance.
People may well have traced these pebbles to their point of origin, using the
same techniques as mineral prospectors.
There is more to say about the spread of portable artefacts, for here there
was not only continuity but change. It seems as if one of the first developments
of the Neolithic period was the establishment of flint mines close to
the English Channel coast in Sussex and Wessex (Barber, Field, and Topping
1999). It happened in the early fourth millennium BC and suggests that land
clearance and the working of timber were important from the start of the
Neolithic period. The mines themselves have close counterparts in many parts
of mainland Europe, extending from southern Sweden, through Denmark,
The Netherlands, and Belgium to northern France (Weisberger ed. 1981).
The English mines seem to have been devoted mainly to the production of
axes and appear to have been located in isolated positions beyond the limits of
the settled land. It is interesting that these mines originated at such an early
date since adequate raw material could have been found in surface deposits –
as it obviously was during the Mesolithic (Gardiner 1989). Not everyone may
have had access to mined flint, and it could be that certain objects took on a
special significance. It is not clear how far they travelled across country, but to
judge from those examples found in stratified contexts, the earliest products
of highland axe quarries like those in the Lake District were distributed over a
limited area (Bradley and Edmonds 1993: chapter 8). That contrasts with the
movement of jadeite axes across Europe. These had been made in the foothills
of the Alps, yet their distribution within Britain extends from the south coast
of England to northern Scotland (P´etrequin et al. 2002).
It is always easier to document the movement of objects than the spread
of ideas, but there were also certain social practices that may have retained
their importance from the Mesolithic to the Neolithic. In southern Ireland
one was the use of caves for the deposition of human bones (Marion Dowd, pers. comm.). This is especially interesting since disarticulated human remains
are also found in Neolithic chambered tombs. Vicki Cummings and Alasdair
Whittle have drawn attention to the close relationship between some of these
structures in Wales and features of the natural topography, suggesting that
places with a long established significance in the landscape may have been
monumentalised (2004: chapter 7). A number of these places are associated
with Mesolithic artefacts. Even if the Neolithic period began with settlement
from overseas, it is clear that the indigenous population had a part to play.
Almost certainly, it would have outnumbered the immigrants.
So far this account has employed two traditional terms, ‘Neolithic’ and
‘Mesolithic’, and has considered their connotations. Other terms have a more
straightforward chronological significance. For the purposes of this study, the
Neolithic period is divided between two largely independent sequences of
change, which will be referred to as ‘Earlier’ and ‘Later’ Neolithic, respectively.
It is the first of these that is studied in this chapter. The Earlier Neolithic
can be subdivided according to its pottery styles, and these show a certain
correlation with changing forms of monuments. The commonest division is
between an ‘Early’ Neolithic typified by fine undecorated vessels, when such
structures were relatively rare, and a ‘Middle’ Neolithic phase with a more
varied ceramic tradition and the development of a variety of mortuary monuments
and earthwork enclosures. The division is usually made at about 3600
BC (Gibson 2002: chapter 4). Unfortunately, few settlements or monuments
are dated with sufficient precision, and in the sections that follow the Early
and Middle phases are treated together unless otherwise stated.
HOUSES AND SETTLEMENT PATTERNS
The Neolithic took different forms in different areas. In England, the number
and size of extraordinary field monuments support Julian Thomas’s notion
that the period is best defined by a particular way of thinking about the world
(J. Thomas 1999). In Ireland, it also involves the rapid adoption of agriculture
(Cooney 2000).
One of the most striking results of field archaeology in recent years has been
the discovery of Neolithic houses in Ireland and to a smaller extent in Scotland
and Wales (Fig. 2.3; Grogan 1996; Grogan 2004a; G. Barclay 1996). This
contrasts with their remarkable rarity in England (Darvill 1996). This can hardly
be a result of differential survival. Many of the structures found in Ireland were
in areas that had experienced a long history of cultivation, just as some of the
projects carried out in England have examined well-preserved deposits. In fact
the expansion of commercial fieldwork has been a more recent development
in Irish archaeology, although the number of projects is rapidly increasing. In
England, on the other hand, there have been many more excavations on sites of every period, yet few Neolithic buildings have been found. The contrast is
a real one and needs to be explained.
Different kinds of Neolithic archaeology are represented by work in these
two countries. In Ireland, a long period of research concerned with megalithic
tombs has given way to investigations of occupation sites. In England, on the
other hand, attempts to locate substantial settlements have gone out of fashion.
Either the most promising candidates proved to be more specialised structures
like earthwork enclosures, or the remains of living places were so ephemeral
that they encouraged prehistorians to devote more attention to monuments.
In Scotland and Wales fieldworkers have been able to study settlements and
tombs, but only in Orkney is it possible to treat all these aspects of the Neolithic
together on equal terms.
Such variations do appear to reflect certain realities. In Ireland and in northern
and western Britain, the house and the individual settlement seem to have
provided a focus for domestic life, even when they were complemented by
specialised monuments. In England, and especially in the south and east, the
sheer scale of certain monuments draws attention to the very limited evidence
for everyday activities, and again this may reflect priorities in the past rather than the biases among fieldworkers in the present. On the other hand, such
differences have been made all the sharper by a process of ‘rethinking the
Neolithic’ which has drawn attention to the ways in which its archaeology
does not conform to established preconceptions.
Those expectations have been drawn from the archaeology of Continental
Europe and fail to convince for two reasons. First, they have been influenced
by the material remains of a much earlier period than the Neolithic of Britain
and Ireland. Such stereotypes grew out of work on the Linearbandkeramik
and its immediate successors. They relate to a time up to a thousand years
before the period in question. They refer to a remarkably stable pattern of
settlement typified by massive domestic buildings, but this was coming to an
end in the mid-fifth millennium BC, and by 4000 BC no trace of it remained
(Whittle 1996: chapter 6). Around the rim of Continental Europe the evidence
is rather like that in Britain and Ireland, and the remains of occupation sites
are less apparent. Often they are reduced to scatters of artefacts and pits, and
the presence of well preserved dwellings is altogether exceptional. A second
stereotype has been even more pervasive. Because the well-known settlements
of the Linearbandkeramik were characterised by massive longhouses, there has
been a tendency to interpret later timber buildings in the same terms. None of
the excavated structures in Britain and Ireland is of the same proportions, nor is
there any convincing evidence that people and animals had lived together under
the same roof. For that reason it is misleading to talk of Neolithic ‘longhouses’
in the study area. In fact many of these buildings can only have accommodated
a limited number of occupants.
In fact there is a further contrast that needs to be explored. The longhouses
of the sixth and earlier fifth millennia BC in Continental Europe had been
organised into hamlets or even small villages. At present it seems likely that
the first Neolithic dwellings in Britain and Ireland had a more dispersed distribution,
although an exception may occur at Stonehall in Orkney (Colin
Richards, pers. comm.). Otherwise it seems as if these buildings are quite isolated,
although they can be grouped into small clusters and distributed fairly
evenly across the landscape. Those in Ireland are often quite near to the coast.
To some extent the pattern may be changing as a result of recent fieldwork. For
example, the supposedly isolated Neolithic houses at Corbally in the east of
Ireland now appear to form part of a wider distribution, with more than one
nucleus where several houses occur together (Purcell 2002; Grogan 2004a).
There is also evidence that some Neolithic structures were replaced in the
same positions. Without a much finer chronology this provides virtually the
only evidence of sequence, and at present it is impossible to tell how many
other buildings were in use together. It is often suggested that this pattern
was established around 4000 BC, but a recent study has shown that much of
the dating evidence is based on the structural timbers which were usually of
oak. Radiocarbon dates on short-lived samples suggest that 3800 BC provides a better estimate (McSparron 2003). Again this has implications for the rate of
change.
There were regional differences in the forms taken by Earlier Neolithic
houses, although all the securely dated examples were rectangular, oval, or
square. Other distinctions emerged with the passage of time. Most were built
out of wood, but rather later stone houses with a similar ground plan have
been identified at Eilean Domhnuill in the Outer Hebrides (Armit 2003a) and
in the Orkney settlements of Knap of Howar (A. Ritchie 1983) and Stonehall
(Colin Richards, pers. comm.).
Despite these distinctions, a number of the structures shared certain features
in common. Their contents can be distinctly unusual, for they include significant
quantities of fine pottery but few other artefacts. In some cases, objects
seem to have been placed in their foundations, and a number of examples
have produced significant quantities of carbonised grain. These finds have led
to confusion. On the one hand, the exceptional character of such deposits
encouraged the view that the buildings were used in ritual and ceremonial
( J. Thomas 1996a; Topping 1996). On the other, it suggested that if large
quantities of grain are found in these houses, that should have been standard
practice at Neolithic settlements (Rowley-Conwy 2004). Both observations
may be misleading. The artefacts in the post holes and wall trenches might have
been offerings that were made when the building was erected. The exceptional
collections of pottery may have been placed there when those structures were
abandoned and destroyed, in which case neither group of objects need have
been associated with the occupation of the houses themselves. The discovery
of carbonised grain can also be confusing, for in common with other Neolithic
buildings, some of them were burnt down. A similar process happened on so
many sites, from cursus monuments to long barrows, that it seems to have
been intentional. These finds may shed light on the domestic rituals associated
with the occupation of houses, but they do not transform these dwellings into
some kind of specialised monument. It is possible that Neolithic longhouses
in parts of Continental Europe were abandoned when one of the occupants
died (Bradley 2001). That interpretation may also be relevant here, not least
in those cases where the sites of earlier settlements had been buried beneath
mounds or cairns.
The idea that certain rituals centred on the domestic buildings may also
explain the exceptional structures that were built in Scotland at this time.
At first sight these share the attributes of the Irish houses. They were rectilinear
with rounded ends, some were divided into small compartments, and
again they were associated with fine pottery and with burnt grain. On the
other hand, they have certain exceptional features. The artefact assemblage
contains significant quantities of ceramics, but very few lithic artefacts (Barclay,
Brophy, and MacGregor 2002). The buildings are apparently isolated
from any dwellings, yet they can be found quite near to specialised earthwork monuments of the kinds described later in this chapter. These structures had
been built on a massive scale, and yet the separate rooms were so small that circulation
about the interior would have been exceptionally difficult. Still more
revealing, they had been burnt down and replaced on at least one occasion. The
first to be excavated was selected on the basis of an aerial photograph which
suggested that it might have been an early medieval feasting hall (N. Reynolds
1978). That was obviously wrong, but the excavators of another example at
Claish have observed that it could have been a public building of a similar kind
(Barclay, Brophy, and MacGregor 2002). The features of an ordinary dwelling
were apparently reproduced on a monumental scale, but people would still
have understood the reference to domestic architecture.
Both the Scottish and Irish buildings have been used to counter the popular
idea that the Neolithic settlement pattern was based on mobility and perhaps
on the use of wild resources. Although domesticated animals provided food
and secondary products, it is the cereals from these sites that have attracted
most attention, for discoveries of this kind have been less abundant in southern
England, where Neolithic houses are rare (Robinson 2000). These new discoveries
certainly call into question some of the tenets of the new interpretation,
but it is important to see them in perspective, for buildings of these
kinds were quite short lived. They do not represent the normal pattern for the
Neolithic period. Many of them were erected during the first few centuries
of the fourth millennium BC, and after that time they largely disappear, to
be replaced by more ephemeral circular buildings. That is true of most of
the well-preserved structures in Ireland, but it does not apply so clearly to
the Scottish sites. A similar change is illustrated by the less extensive evidence
from England andWales. From approximately 3500 BC the settlement pattern
in different regions of Britain and Ireland seems to have been rather more
homogeneous, and almost everywhere houses and settlements are harder to
identify. The representation of substantial rectilinear dwellings had varied from
one region to another, but in most cases they were built over a limited period
of time.
That interpretation is strongly supported by recent work in Ireland. Here
it seems as if the first Neolithic settlements made a significant impact on the
natural environment. Pollen analysis has identified a series of major clearings
at the beginning of this period. These are associated with the appearance of
cereal pollen (O’Connell and Molloy 2001). That is not surprising in the
light of the excavated evidence, but it is revealing that, after a brief phase
of expansion during which substantial houses had been built, some of those
clearings regenerated. From then on, there are more indications of pastoral
farming than there are of crop cultivation. Rather similar claims have been
made at a number of sites in Britain where clearingswere either abandoned at an
early stage or reverted to less intensive land use (Whittle 1978), but in this case
there seems to have been greater regional and chronological variation, and the main impact on the landscape did not come until the Later Neolithic period.
It is not clear why these changes happened, but Petra Dark and Henry Gent
(2001) have made the interesting observation that the first cereals introduced
to these islands would probably have been protected from crop pests. That
immunity would have broken down over time, meaning that the earliest crops
would have been unusually productive.
Another development in Irish archaeology has been the discovery and dating
of systems of stone-walled fields on the west coast (Fig. 2.4). These are
associated both with settlements and tombs, but until quite recently it appeared
that they had been used during the later part of this period. Now it is clear that
they were of a much earlier date (Molloy and O’Connell 1995). An extensive
programme of radiocarbon dating shows that C´eide Fields were established by
about 3700 BC and that they had already been invaded by peat by 3200 BC.
Although they were used mainly as pasture, it is clear that cereals were being
grown there too. It remains to be seen how many field systems existed at this
time, for it is clear that land boundaries of many different periods are buried
beneath the Irish blanket bog. There may have been similar developments in
other areas. It is possible that fields were established on the Scottish island of
Arran during this period (Barber 1997: chapter 11). A rectilinear system of
ditches at Billown on the Isle of Man may have formed part of another system
of land management during the Earlier Neolithic, but the excavator prefers
to regard this as a monument complex (Darvill 2000). Neolithic field systems
are found also in the Shetland Islands, but they do not appear to have been
established at such an early date (Whittle 1986). Perhaps it is misleading to
place too much emphasis on these fields, as soil micromorphology suggests
that domestic middens might have been cultivated as garden plots (Guttmann
2005). It would not have been necessary to transport their contents to fertilise
arable land.
The environmental record from England does not follow quite the same
trajectory as its Irish counterpart, and the evidence for Earlier Neolithic settlement
takes a very different form. Here there are few signs of Neolithic
buildings. There is no shortage of pit deposits containing the distinctive kinds
of material associated with well preserved settlements, and in the Thames Valley
there are even the remains of open air middens (Allen, Barclay, and Landin-
Whymark 2004). Nor is there much evidence, even from extensive excavations,
that several houses existed together on the same sites. There are fewer
signs of superimposition than occur with their Irish counterparts, and both
in England and Wales massive roofed structures on the scale of the Scottish
‘halls’ have not been excavated, although one candidate in eastern England
has been identified by aerial photography (Oswald, Dyer, and Barber 2001:
fig. 3.14). This raises two possibilities. The first is that settlement really was
more ephemeral. That might be suggested by the small size and limited contents
of most of the artefact scatters and by the restricted number of pits associated with many of these sites. This is reflected by the rarity of carbonised cereals
among their contents, compared with the remains of wild plants. That applies
even to the large houses recently excavated at Yarnton and Whitehorse Stone
(Gill Hey and Alistair Barclay, pers. comm.). On the other hand, it is not
inconceivable that large rectilinear houses had originally been more widely
distributed and that they were built in a way that did not leave subsoil features
behind. Waterlogged wood from the Somerset Levels shows that people were
using sophisticated carpentry techniques during this early phase (B. Coles and
J. Coles 1986: chapter 3). One indication that such houses may once have
existed in England is provided by extensive excavations at settlements along the
east coast which are characterised by a number of pits (Garrow, Beadsmoore,
and Knight 2005). Traces of structures might not have survived, but there
are open areas in between these features which are of roughly the shape and
dimensions of the domestic buildings of this date. Other arcs or clusters of pits
may have been inside the houses themselves (Fig. 2.5).
A persistent feature of these occupation sites is the presence of pits whose
contents were sometimes organised when they were placed in the ground.
They can include pottery, lithic artefacts, querns, and animal bones that seem to
have been taken from middens, but occasionally they also contain disarticulated
human remains (J. Thomas 1999: chapter 4). It was once argued such pits were
grain silos, but it seems more likely that some of them were dug specifically
to receive these contents. The same kinds of deposits were in the hollows
left by fallen trees (Evans, Pollard, and Knight 1999). This practice of burying
particular combinations of artefacts and other material is a distinctive feature of
the insular Neolithic, and at present it seems to be commoner in Britain than
it is in Ireland. It seems possible that this material was carefully buried when
occupation sites were abandoned (Healy 1987). That practice has implications
for field archaeology. In those parts of England where flint is readily available
it can be difficult to recognise concentrations of surface finds dating from this
period. In Ireland, the situation is rather different as similar deposits seem to
be associated with timber buildings rather than pits.
It is difficult to account for these contrasts. Julian Thomas (2004a) has suggested
that one reason why the Irish Neolithic was so different from its British
counterpart is that the indigenous population had no large animals apart from
pigs. The introduction of domesticated cattle, sheep, and goats might have had
a greater impact here than it did in other areas. That seems possible, but it does
not explain the emphasis thatwas evidently placed on cereals, nor can it account
for the presence of equally impressive houses and settlements in northern and
western Scotland where the native fauna was less impoverished. Perhaps his
suggestion accounts for a rather different development, as the Irish pollen evidence
suggests an initial emphasis on cereal growing and a greater emphasis on
stock raising during a subsequent phase. Maybe that might be explained by a
lack of familiarity with cattle before the beginning of the Neolithic period. It is clear that much has still to be learnt about the Neolithic economy of
either island. Studies of carbonised seeds and animal bones have not been particularly
informative because of the circumstances in which those materials were
deposited. The largest collections of faunal remains come from the excavation
of specialised monuments, whilst the main groups of cereals are associated with
structures that may have been burnt down deliberately. Perhaps more direct
evidence of human diet will be obtained by other methods of analysis, although
such work is still in its early stages. Just as the evidence of stable isotopes suggests
a drastic change from a marine to a terrestrial diet at the start of the Neolithic
(M. Richards 2003; 2004), there seems to be evidence for a major emphasis on
animal products during this period. That is not to imply that every community
practised the same economy. Human remains from single tombs provide very
similar signatures to one another, but there are important contrasts between
the samples taken from different sites, suggesting that there was a wide range
of variation. In particular, some people seem to have eaten much more meat
than others (M. Richards 2000). There is also a little evidence from the chemical
analysis of pottery. Some years ago Legge (1981) suggested that Earlier
Neolithic people practised a dairy economy. This was because older cows feature
so prominently among the animal bones. That suggestion was questioned
at the time, but now there is a growing body of evidence that ceramic vessels
had indeed held milk or milk products (Copley et al. 2005).
BEYOND THE HOUSE: LONG MOUNDS
AND MORTUARY MONUMENTS
What of the earliest monuments? Most accounts of the Neolithic period provide
a commentary on a series of separate types. Understandably, those associated
with human remains dominate the discussion, not least because so many
of these mounds and cairns can still be seen today. This procedure raises a
series of problems. First, it suggests that prehistoric people were building these
structures according to a series of prescribed templates when excavation has
shown that the final forms of such buildings provide little indication of their
internal construction. Superficially similar forms of mound or cairn may have
developed along radically different lines. Second, there is a tendency to treat
all the component parts of such monuments as a unitary conception whereas
recent fieldwork suggests that many of them could have had an independent
existence and could even be found on their own. A third problem is that closer
attention to the sequence at individual monuments suggests that the mound
or cairn that had provided the main focus for discussion was sometimes the last
element to be built and may even have been intended to bring the use of that
site to an end.
In fact it seems as if some of the earliest monuments may have taken a distinctive
form in their own right, although the chronological evidence leaves much to be desired. Two elements are particularly important here and in some
ways relate to rather similar ways of utilising components of the natural landscape.
For the most part they have different distributions from one another.
Along the coastline of the North Sea and the English Channel timber ‘mortuary
houses’ are quite common, although this term does not do justice either
to their distinctive form or to the material associated with them. Their distribution
hardly overlaps with that of ‘portal dolmens’ which are found in both
islands but usually occur in areas further to the west. Excavation of the ‘mortuary houses’ has identified settings of D-shaped post
holes of considerable proportions (Kinnes 1992). Usually these occur in pairs
(Fig. 2.6), although other configurations are found.Work at the well-preserved
site of Haddenham in eastern England has confirmed what was long suspected,
that these are the remains of tree trunks which had been split in half (Evans
and Hodder 2005). Many of these sites are associated with human remains.
There are many examples which are buried beneath barrows or cairns, but
they were usually built after these posts had been set in place, and there are
even sites at which such structures were never covered by a monument. They
are particularly interesting since there is evidence that at the beginning of the
Neolithic period the holes left by fallen trees could be marked by offerings of cultural material (Evans, Pollard, and Knight 1999). Perhaps these components
of the native woodland were thought to be especially significant.
In upland areas, especially in Ireland and western Britain, a comparable
process may be evidenced by portal dolmens. These are usually classified as
megalithic ‘tombs’, although it is not certain that theywere originally associated
with human remains. Nor is there any satisfactory evidence that they were
buried beneath cairns. In fact they seem to have been intended as closed stone
‘boxes’ covered by a disproportionately massive capstone which could be higher
at one end than the other. A number of these monuments resemble natural
rock formations (Bradley 1998c), but in certain cases it seems as if the massive
covering slab had been obtained on the site itself. Some of the chambers were
raised above shallow pits, which may have been the quarries from which the
capstone was taken (C. Richards 2004a). The ground surface was split open
by this process (Fig. 2.6), just as in other regions trees were divided in half and
transformed into wooden monuments. Alasdair Whittle (2004) refers to portal
dolmens as ‘stones that float to the sky’, and that description seems entirely
appropriate (Fig. 2.7).
Closed chambers of other kinds may belong to the beginning of the
Neolithic sequence in Britain and Ireland, just as they do in western France,
where a number of sites were originally associated with monumental cists
(Sheridan 2003b). The interior did not become accessible until a later phase.
There are a number of indications that a similar process took place in Britain
and Ireland. Among the more convincing examples are the ‘rotunda graves’ of
the Cotswolds which were incorporated into later long cairns (Darvill 2004: 60–2), and a remarkable site at Achnacreebeag in the west of Scotland which
was associated with pottery in a similar style to that found at monuments in
western France (Sheridan 2003a). The Carrowmore cemetery on thewest coast
of Ireland provides another example of this development, for here a number of
massive stone chambers seem to have formed a freestanding element set within a
circle of boulders. Such structures are associated with radiocarbon dates beginning
around 4000 BC (Burenhult 1980; Sheridan 2003b). The boulder circles
may also have parallels in Brittany, but in the Irish case the cemetery developed
into a group of passage tombs associated with a purely insular material culture.
That development is discussed in Chapter Three.
There are other ways in which the Irish Neolithic is distinctive, for here
the remains of houses seem to have been closely associated with mortuary
monuments, which most often take the form of elongated mounds or cairns.
At Ballyglass in the west of the country two of these overlie the remains
of wooden structures (Fig. 2.8; O´ ’Nualla´in 1972; 1998). One of the cairns
was superimposed on the site of an earlier house, but in such a way that
its layout acknowledged the position of the demolished building. The same
happened at the neighbouring monument, but in this case excavation exposed
the plans of three separate structures, two of which overlapped. It seems as if
only parts of them survived, but again it is clear that the layout of the tomb
was influenced by the organisation of the older settlement. Not far away at
C´eide it is possible to observe how similar monuments had been integrated
into a Neolithic field system (Cooney 2000: 25–9). In such cases it seems clear
that mortuary monuments and settlements were closely associated with one
another. These are not the only instances in which such a close connection
has been suggested, but they are perhaps the most convincing.
Similar links have been inferred in other areas. At Bharpa Carinish in the
Outer Hebrides the exiguous traces of another dwelling were found only thirty
metres from a chambered tomb and may have shared the same alignment as
that monument (Fig. 2.4; Crone 1993). In England andWales, there are further
cases in which a chambered tomb overlay the remains of earlier dwellings. This
was clearly demonstrated at Gwernvale (Britnell and Savory 1984: 42–150),
but slighter domestic structures also seem to have preceded the well-excavated
cairns at Ascott-under-Wychwood (Alistair Barclay, pers. comm.) and Hazleton
North (Saville 1990). In the latter case the relationship between these features
is especially striking. The building of the cairn proceeded gradually as separate
masses of rubblewere packed inside a series of bays, each of them defined by dry
stonewalls. It is clear that the position of an earlier middenwas left exposed until
the late stage of the project, then the burial chambers were built alongside it.
The connection between houses and monuments perhaps goes further. The
stone houses at the island site of Eilean Domnhuill were approached through
an impressive stone setting which the excavator compares with the forecourt
of a megalithic tomb (Armit 2003a). In Orkney, the comparison even extends to the internal structure of the earlier chambered cairns which are divided into
a series of separate compartments, linked by a central corridor and separated
from one another by screens. Colin Richards has argued that the architecture of
these tombs refers to the characteristic form and organisation of the domestic dwellings at sites like Knap of Howar. It seems appropriate to consider such
monuments as the houses of the dead (Fig. 2.9; C. Richards 1992). That is
certainly suggested by the organisation of the fittings inside structures of both
kinds. The stone shelves on which human bodies were placed within the tombs
occupy the equivalent positions to the beds inside the domestic buildings.
These monuments can also be studied on a larger scale, but this work often
focuses on the evidence for regional variation. Minor architectural details can
easily assume a life of their own until any wider patterns are obscured. This
is particularly true when megalithic tombs are considered one class at a time,
for it presupposes that prehistoric people intended to build particular ‘styles’
of monuments in the first place. That is by no means obvious.
Two issues seem to be particularly important: the relationship between the
external appearance of different tombs and the organisation of the chambers
found within them; and the structural sequences that led to these monuments
assuming their present forms.
The first point needs careful consideration, for it is clear that there is no consistent
relationship between the shape of the mound or cairn and the layout of
any structures that were concealed inside it. In northern Scotland, for instance,
rectangular or oval cairnswere usually positioned so that their long axiswas visible
on the skyline from the settlement area. That did not apply to all the round
cairns built during the same period, but this simple distinction is obscured
when the classification of these monuments extends to the burial chambers (T.
Phillips 2002). The most likely reason is that the external appearance of these
cairns was apparent to everyone in the course of daily life. Access to their interiors
may have been restricted to a smaller group. By conflating the internal and
external features of such monuments, the main source of variation is obscured.
At the same time that example makes it clear that more than one form of
mortuary monument might have been constructed during this period. The
simplest distinction is that between long mounds or cairns and the round
barrows which are discussed in a later section of this chapter. Sometimes they
occupy different positions in the landscape, either in relation to the local topography,
as happens in the north of Scotland (T. Phillips 2002), or in relation to
different areas, for instance on the Yorkshire Wolds ( J. Harding 1996). On
the other hand, the distinction is not clear-cut, for excavation has shown that
smaller circular constructions were often encapsulated within more massive
long mounds. In other words, the outward appearance of these monuments
disguises the fact they might have achieved the same form by quite different
means. That is sometimes overlooked in categorising their outward appearance.
Apart from portal tombs and those at Carrowmore, the earliest style of
monument in Ireland was probably the ‘court cairn’ (de Valera 1960). They
have this name because the internal chambers are generally approached though
a curvilinear walled enclosure or ‘court’ which remained open to the elements
(Figs. 2.8 and 2.10). Although many monuments have been damaged, these features were normally associated with long mounds like the two examples at
Ballyglass. The most obvious characteristic of these sites is their sheer diversity,
and yet it seems that many of them had been constructed in a single operation
or developed over a restricted period of time. The mound or cairn normally
took the same form, but there is considerable variation in every other element.
This concerns the nature of the court itself – it can occur singly or in pairs,
and individual examples can be more or less enclosed; it may concern the
number of chambers or their relationship to one another; or it may be related
to differences of scale or accessibility between the separate parts of these tombs. These distinctions have led to many subdivisions among the plans of Irish court
cairns, but the reasons for their creation are very rarely discussed.
Andrew Powell makes the interesting suggestion that these different configurations
reflect the complex patterns of kinship and alliance between different
segments of the local population (Powell 2005). The tomb plans with their
bewildering variety of chambers acted like a genealogy, expressing the connections
between different groups of people and enshrining those arrangements
in a lasting architectural form. If so, then a single tomb might represent more
than the individual household. Rather, the combinations of different courts
and chambers might provide an idealised image of social organisation across a
wider landscape. Of course, this will be impossible to prove, but the very fact
that the remains of so many people are found at these sites would certainly
support the notion that the monuments stood for a wider community. Hardly
any grave goods were associated with individual bodies, so it seems as if the
differences that were apparent in life were not expressed in the structures of
these tombs.
In Britain, there are other problems in approaching megalithic tombs, or
their equivalents built out of less durable materials. It has been easy to suppose
that mounds and cairns were always built to contain burial chambers in the way
that seems to have happened in the Irish court tombs. That is not always the
case, for there is so much diversity that it is difficult to identify any overarching
scheme.
There is a paradox here, for it is true that many of these monuments look very
much alike, but in this case they reached their final forms by different means.
The classic long barrow or long cairn has certain clearly defined elements,
although not all of these are present on every site. In theory, each monument
includes most of the following (Kinnes 1992):

an elongated mound or cairn which should be rectangular, oval, or trapezoidal
and higher towards the eastern end;

a monumental forecourt or facade in front of the entrance, normally
facing to the east;

‘porches’, or even avenues, of spaced posts which approach the entrance
to the monument;

stone walls or wooden revetments to support the edges of the mound;

ditches or quarries which extend along the sides of those monuments and
occasionally around the ends;

a ‘burial chamber’ or chambers which are approached through the forecourt
or may be located in the flanks of the mound or cairn;

and a blocking of timber or stone introduced when the monument was
closed.
Stone chambers could assume many different configurations, but at nonmegalithic
long barrows the ‘mortuary house’ took the formconsidered earlier It usually consisted of two or more split tree trunks which could be bounded
by banks or low walls. These structures were sometimes contained within freestanding
ditched or fenced enclosures, and there are even sites that contained
more than one of them. The precise forms taken by all these features vary from
one region to another, and this has led to the recognition of a series of local
styles (Masters 1983; Kinnes 1992). Although they do have some validity, the
process of classification obscures a large amount of variation.
Put simply, nearly all the elements listed above can be found in isolation, and
they occur in various combinations which do not equate with any of the ‘classic’
forms of monument (Fig. 2.11). Thus mortuary houses are recorded as isolated
structures, or they may be recognised inside open enclosures. There arewooden
forecourts which have no barrows or burials at all, and there are mounds which
never seem to have contained any human remains. The classic form of the
monument may encapsulate the positions of smaller structures, including older
round barrows or round cairns (Fig. 2.12), and the entire scheme may have been
conceived on a modest scale, or it might have required a much greater amount
of labour. In the south these monuments were usually sealed or allowed to
decay, whilst those in northern England and eastern Scotland were sometimes
set on fire.
It seems as if these separate elements should be understood in terms of a
process or series of processes which was more or less extended at different sites.
Perhaps there never was an ideal conception of how that sequence should end.
Rather, there were many local variations which may have been influenced
by the status of particular individuals and the ways in which they were to be  remembered. There is another complication. The results of recent fieldwork
suggest that the massive monuments which have dominated the literature so far
were altogether exceptional and that it is the smaller structures which may really
represent the norm. On the chalk of Lincolnshire, for instance, eleven long
barrows can be identified as earthworks, but aerial photography has revealed
the sites of more than fifty other mounds and enclosures which may have
been their less monumental counterparts (D. Jones 1998). Some of these were
located in different positions in the landscape, and because they were so much
slighter they have been levelled by the plough. On many sites in Britain the
structural sequence did not end with the creation of a classic long barrow, and,
more often than not, it may never have been intended to do so. With these qualifications, there is a dominant pattern in Britain which contrasts
with what has been said about the monuments in Ireland. Whereas Irish
court tombs exhibit a wide variety of ground plans, their architecture makes
use of a few recurrent features that were combined in many different ways.
This process seems to have taken place over a limited time, or at least it did
not involve any drastic changes to the configuration of the monuments. The
British evidence suggests that individual examples developed from many different
starting points. In this case monuments were modified, extended, and even
changed their shapes, and parts of these structures may have been replaced.
The only consistent feature of the larger sites was that in the end many of them
assumed a similar outward form. When these structural sequences reached their
conclusion, the sites were sometimes covered by an elongated mound or cairn
(Fig. 2.12).
It follows that these constructions were not necessarily an integral part of
such sites until a very late phase, and in the case of non-megalithic long barrows
it seems unlikely that the ‘chambers’ were covered by mounds until those
sites went out of use. That mound sealed the features created during earlier
phases and cut them off from the living. It may be why the building of these
earthworks sometimes happened during a phase when such structures were
set on fire. This was not a cremation rite in the way that is often claimed.
Like many other Neolithic monuments, the use of these places was brought to
an end by burning them down and burying them under a mantle of rubble.
It happened particularly often in northeast England and eastern Scotland. In
western Britain the equivalent was the closing of megalithic tombs with a
deposit of rubble. There is little to suggest that most of these were used for
more than a few generations.
If that is true, then other possibilities come to mind. In both countries there
are circumstantial links between long mounds and the rectilinear houses of
the Earlier Neolithic period (Bradley 1998b: chapter 3), but the distinction
between the character of Irish court cairns and most of their British counterparts
may correspond to a broader division in the settlement pattern. In Ireland,
houses are widely distributed and sometimes occur in small groups on, or close
to, the sites of chambered tombs. These tombs have a variety of ground plans,
but they are built out of segments which could be arranged in many different
configurations from one monument to the next. Even when this happened,
the structure retained its distinctive form. There were few radical changes in
the evolution of individual sites, and it often seems as if the formation of these
particular cairns happened over a fairly short period.
In Britain, on the other hand, it seems possible that the settlement pattern
was much less uniform, and in the south and east it may have been characterised
by significantly greater fluidity. Here a bewildering variety of structures was
associated with the celebration of the dead; indeed, their number and diversity
are still increasing as a result of modern fieldwork. Some of these monuments remained unchanged, including a series of round barrows which will be discussed
later in this chapter, but where larger mounds and cairns were built over
them, they would have been among the few fixed elements in the Neolithic
landscape.
BEYOND THE HOUSE: SETTLEMENTS, BODIES, AND TOMBS
Why was it that domestic buildings suggested the appearance of the tombs? A
useful starting point is the work of L´evi-Strauss (1983) on what he calls ‘house
societies’. He is most concerned with kinship organisation and the emergence
of hierarchies, but his work is particularly important because it shows how the
idea of the ‘house’ extends to the occupants for the building and the members
of a wider community.
It is in this sense that the term is used in a recent paper by Mary Helms
(2004), who discusses the different worldviews of mobile hunter-gatherers and
the first farmers. One might almost say that it is by the construction of houses,
both real and metaphorical, that particular groups distinguish themselves from
others and define their membership. Their composition is less fluid than that
of hunter-gatherer communities, and it is maintained over a longer period of
time. Such concerns are particularly relevant when people are exploiting an
unfamiliar environment, and the new arrangement may also reflect the labour
requirements of early agriculture. Perhaps that is one reason why houses are
such a conspicuous feature at the beginning of the Neolithic period.
A pertinent observation comes from a paper by Janet Carsten and Stephen
Hugh-Jones introducing an edited volume devoted to L´evi-Strauss’s ideas. They
observe how common it is for houses to be regarded as living creatures and to
be thought of in the same terms as human beings, who are born, grow old,
and die. Like people, houses have biographies of their own.
Houses are far from being static material structures. They have animate
qualities; they are endowed with spirits or souls, and are imagined in terms
of the human body. . . . Given its living qualities . . . it comes as no surprise
that natural processes associated with people, animals or plants may also
apply to the house (Carsten and Hugh-Jones 1995: 37).
Perhaps the reason why the British and Irish evidence poses so many problems
is that the histories of the buildings in which people had lived were reflected
by the ways in which their bodies were treated when they died.
The most obvious contrast is between the regions with houses and those
in which they are rare. In lowland Britain, where pit deposits are particularly
common, the artefact assemblage occasionally contains human bones. It seems
possible that relics had circulated in the same manner as the objects with which
they were found. That is consistent with the evidence from mortuary monuments
which not only include the remains of complete corpses but can also feature certain body parts to the exclusion of others (Whittle and Wysocki
1998: 151–8). This suggests that the dead were reduced to disarticulated bones
and that their remains were distributed in the same manner as domestic artefacts
(Pollard 2004). Although deposits of pottery are associated with some of
these monuments, many excavated sites in the south provide little evidence of
portable material culture.
Compare this with the situation in Ireland where houses are much more
common and isolated pit deposits are rare. Considerable numbers of artefacts
are associated with court tombs which also contain charcoal-rich soil similar
to that in settlements. Humphrey Case (1973) has shown that this material was
usually placed on top of a deliberately laid floor, meaning that it must have
been introduced after the tombs had been built. Not surprisingly, such deposits
are associated with human remains. Other regions in which stone-built tombs
are associated with significant quantities of artefacts, especially pottery, include
the north and west of Scotland, both of them regions where the remains of
houses have been found.
That contrast is interesting, but it says little about the treatment of the dead.
There is an important difference in the principal mortuary rite. In some regions
bodies were cremated and in others they were allowed to decay. Certain of the
burnt bones might result from the combustion of an entire body; others might
have been placed in a fire after they had lost their flesh. The important point
is that the unburnt bones could circulate as relics. That would be less likely
in the case of cremated bone, and there is little evidence for its deposition in
other places.
The distributions of these two ways of treating the body are rather revealing.
The burial of corpses is most often found in Britain, where it is commonest
in the southern half of the country, although isolated cremations associated
with decorated pots are occasionally found in eastern England. Unburnt bones
occur on domestic sites over a much larger area. There is evidence for the
burning of wooden mortuary structures along the North Sea coast, and in the
megalithic tombs of northern Scotland human bones were exposed to fire but
were not reduced to ashes. Along the west coast of Scotland, on the other
hand, cremation was often employed. That is not surprising as it was one of
the main ways of treating the body in Irish court tombs.
Thus the preservation and circulation of human bones was a particular
feature of southern England, but seems to have extended across a wider area.
The use of cremation characterises many of the Irish sites, but it also extended
to parts of northern Britain, where it is more difficult to identify one prevailing
rite. That is not to deny that some monuments are associated with cremation
and interment. There are a few cremations on southern English sites, but these
are generally secondary to deposits of unburnt bone (Darvill 2004: 145). Similarly,
Irish court tombs show evidence for both these practices, but in this case cremations are more common than inhumations (Cooney and Grogan 1994:
fig. 4.14).
The most striking contrast in ways of dealing with the dead is between
the areas that show the greatest divergence in the evidence for settlement
sites. In southern England houses are rare and pit deposits are more common,
including those which combine a selection of domestic artefacts with fragments
of human bone. It is here that there is most evidence for the deployment of
unburnt corpses in long barrows and megalithic tombs where they might be
arranged according to age, gender, or different parts of the body (Whittle and
Wysocki 1998). Artefacts are not common at most of these monuments, and it
seems possible that the residues of older settlements were allowed to decay and
were dispersed in the same manner as human remains. In Ireland, on the other
hand, houses are commonly found and isolated pit deposits are unusual. The
residues of domestic occupation might have been deposited in tombs together
with the remains of the dead (Pollard 2004). In this case the bodies were often
burnt and there is little to suggest that bone fragments circulated in the same
manner as artefacts.
One last contrast is important. Dermot Moore (2004) has shown that a high
proportion of the Irish houses had been destroyed by fire. Although this might
have been a result of warfare, the evidence is rather ambiguous, and it seems
much more than a coincidence that human corpses should have been treated
in exactly the same way as these buildings. Perhaps that is because the careers
of particular people and the histories of their houses were in one sense the
same. The house was a living creature and its life had to be extinguished in
a similar manner to the human body. That may be why, in Ireland, what are
apparently domestic assemblages accompanied the dead person to the tomb –
they might even have been the contents of the settlement. By contrast, in
southern England, the remains of settlement sites were dispersed along with
the human bones, some of which were eventually deposited in tombs at which
finds of artefacts are uncommon.
The Irish houses could be more substantial than their English counterparts.
That may be because they were to play a spectacular role at the end of their lives
and those of their occupants. By contrast, the dwellings inhabited in England
did not need to do this, and so they might have been quite ephemeral structures.
That could be why they have been difficult to find by excavation. These
buildings were more than shelters from the elements: they were animated
by their involvement in human lives, and, when their inhabitants died, their
treatment followed the same principles as that of human bodies. In England,
they decayed and their contents were dispersed. In Ireland, they were burnt
down and their contents were concentrated in tombs (Fig. 2.13). That may
be why there was a close relationship between such monuments and domestic
dwellings.


BEYOND THE HOUSE: CURSUS MONUMENTS
AND BANK BARROWS
Some of the largest structures to employ the idea of the house were neither
mounds nor cairns. These were the wooden halls that have recently been
found in northern Britain. It is difficult to assess their claims as residential
buildings, but their characteristic architecture was obviously based on that of
the domestic dwelling (Fig. 2.14). Their siting is important, too. Recently
excavated examples at Claish, Littleour, and Carsie Mains are located within
a short distance of enormous earthwork mounds (Barclay and Maxwell 1998;
Barclay, Brophy, and MacGregor 2002; Brophy and Barclay 2004).
These wider connections raise problems. The rectangular buildings are of
several forms, and only the earlier of them could have been roofed (K. Brophy,
pers. comm.). Certain connections are very important here. Such structures
were erected over the same period of time as one another; they were frequently
burnt to the ground and replaced in exactly the same positions; and yet they
are usually assigned to different categories. The smaller examples are thought
of as halls or monumental houses, and the larger ones are sometimes described
as ‘long enclosures’; they have even been called ‘mortuary’ enclosures because
they have a similar ground plan to long barrows. The problem does not end
there, as many of those enclosures were associated with still more elaborate timber
alignments of the kind known as ‘cursuses’ (Barclay and Harding eds. 1999).
There is no consistent method of distinguishing between these three types, and
the halls, the long enclosures, and the cursus monuments actually form a continuum.
That suggests one way of thinking about these features, for it is possible
that individual examples might have grown incrementally, although that process
would have been interrupted by episodes of burning and rebuilding.
The excavated structures at Claish and Littleour suggest another dimension
to the problem, for each is located close to an exceptionally long mound of
a kind which is sometimes described as a ‘bank barrow’. In each case that
earthwork seems to have developed out of one or more burial mounds of
conventional proportions. The Auchenlaich long mound, which is close to
the hall at Claish, certainly contained a megalithic chamber (Barclay, Brophy,
and MacGregor 2002: 114–19), whilst detailed survey at Cleaven Dyke near
to Littleour strongly suggests that it began as a conventional burial mound that was gradually extended until it ran for two kilometres. Excavation suggests
that a developed stage in the sequence can be dated to the early or mid-fourth
millennium BC (Barclay and Maxwell 1998).
The surviving earthwork at Cleaven Dyke poses another problem. For most
of its length it consists of a central mound in between two widely spaced
ditches, but towards one end those ditches may have defined the limits of an
open avenue or enclosure. That raises the question of definition, for elongated mounds of this kind are usually classified as bank barrows, whilst extended
rectilinear enclosures are known as cursuses. Cleaven Dyke is not the only
composite monument of this kind, and field survey in other regions of Britain
has shown that these two classes of field monument are closely associated with
one another (Bradley 1983). They are based on the same principle of extending
a rectangular or oval monument on a significant scale.
The relationship between the conventional categories can be expressed in
this following way (Table 2.1):
table 2.1. Possible relationships among linear timber and earthwork monuments.
Short Long
Timber monuments Hall Long enclosure /
‘Mortuary’ enclosure
Cursus
Earthwork monuments Barrow Long enclosure/
‘Mortuary’ enclosure
Cursus Bank
barrow
This scheme seems to suggest that in Scotland Earlier Neolithic halls, cursuses,
and bank barrows were closely related to one another (Fig. 2.14). Cleaven
Dyke is particularly important because of the proximity of the timber building
at Littleour, but it has other links with the tradition of long barrows and
long cairns. The western end of the monument seems to have originated as a
barrow, and a cairn of massive proportions is aligned on its eastern terminal. In
the Later Neolithic period the chain of connections extended to a further Scottish
site at Balfarg where a structure which has been compared with a timber
hall was buried beneath a mound. In time its position was emphasised by the
construction of a ditched enclosure. This building dates from about 3000 BC
(Barclay and Russell-White 1993).
Are there similar indications from other areas? With the possible exception
of the bank barrow at Stanwell (Barrett, Lewis, and Welsh 2000), none of the
monuments in England seems to have had a timber precursor, nor have any
post-defined monuments been identified there. Other bank barrows are found
along the east coast and extend into the southern Pennines, the Thames Valley,
and Wessex, but none attains the remarkable length of the principal Scottish
examples. No timber hall has been excavated in England or Wales, and, like
the cursus monuments, the features that have been described as ‘mortuary
enclosures’ are usually defined by a bank and ditch. The Irish evidence is
still more limited. Although a few candidates have now been recognised, the
cursuses in that country have not been firmly dated (Condit 1995; Newman
1999). This is significant as their characteristics overlap with those of the ditched
roadways leading to high-status sites of the Iron Age.
Chronology is very important, as it seems as if the idea of elongated timber
and earthwork monuments originated in the north of Britain during the
currency of the Earlier Neolithic timber halls with which they are sometimes
associated. They experienced a similar cycle of deliberate burning and rebuilding. This is a feature which they share with some of the domestic
dwellings in Ireland. Major earthwork cursuses seem to be later in date. A
recent review concluded that they were probably constructed between about
3650 and 3350 BC (Barclay and Bayliss 1999).
Although the English cursuses may have originated after those in the north,
they eventually grew to enormous proportions. In their final form they consisted
of considerable elongated enclosures with an internal bank and an external
ditch. These earthworks were occasionally breached by causeways, but they
were rarely provided with formal entrances at the terminals. Individual examples
included existing structures in their paths – long barrows, round barrows,
or ditched enclosures of various sizes – and sometimes they were also aligned
on an earlier earthwork. It was relatively common for a cursus to lead towards
(or extend from) an older rectilinear enclosure or even a long barrow, but other
arrangements have also been recognised. At Springfield there was a timber circle
in this position (Buckley, Hedges, and Brown 2001). The Scottish hall at
Carsie Mains was aligned on a similar feature which dates to between 3350 and
3000 BC (Brophy and Barclay 2004).
Some cursuseswere built in comparative isolation, but there are also instances
in which they seem to have been constructed incrementally, as one of these
enclosures was added on to the end of another. Alternatively, they can run
side by side, they may converge on certain focal points, or they may overlap.
The two largest cursus complexes illustrate some of these patterns (Fig. 2.15).
At Rudston in northeast England no fewer than five of these converge on
the tallest monolith in Britain (Stoertz 1997: 25–30). Two of the earthworks
meet at right angles, and two of them intersect. Long barrows can be seen on
the horizon from these sites (H. Chapman 2005), and the terminal of one of
the monuments was enlarged to resemble another burial mound of this type.
Taken together, the cursuses run for about ten kilometres.
The other major complex is known as the Dorset Cursus, although, properly
speaking, it consists of two, and possibly three, such monuments built end to
end (Barrett, Bradley and Green 1991: chapters 2 and 3). Again their combined
length is ten kilometres. It would have taken half a million worker hours to
complete the entire structure. They incorporate burial mounds in their path,
and further monuments of the same kind were aligned on the ends of these
earthworks. As at Rudston, one of the terminals was built on a massive scale, as
if to echo the characteristic form of two adjacent long barrows. At its northern
limit, the Dorset Cursus runs up to another mound which was subsequently
lengthened to form a bank barrow. The earliest section of this composite
monument was orientated on the midwinter sunset. According to radiocarbon
dates, it was one of the last examples to be built, most probably between about
3360 and 3030 BC.
Such monuments are generally interpreted as avenues or processional
ways, although it is not known whether the construction of the earthworks
monumentalised existing paths or represented a new development in the landscape. In the same way, because certain examples led between mortuary
mounds, it may be that such routes were for the use of the dead rather than
the living. Human bones have been found in the excavation of cursuses, and
the idea that they were intended primarily for the deceased might also account
for the interplay between cursuses, which are entirely open structures, and
bank barrows, which are solid mounds. It would certainly help to explain why
there were so few points of access to the interior of these monuments. In fact
there has been little discussion of the fact that these putative paths were blocked
at both ends. That would surely have impeded access to the interior, and yet
it was only from there that anyone could have seen the alignments created by
these monuments. One possibility is that these structures were originally open
and that, like many of the long barrows and long cairns described in this chapter,
they were closed during a subsequent phase (I must thank Roy Entwistle
for this idea). This has not been established by excavation, although there are
occasional sites where no terminals have ever been found.
Recent fieldwork has shown that most cursus monuments and bank barrows
are closely integrated with the features of the local topography (Fig. 2.16). Two
particular arrangements are widely recognised. The first is where they extend
across valleys so that their terminals are visible from one another and command an extensive view over the lower ground. Alternatively, they cut across the
contours so that any people travelling along them would have experienced a
sequence of changing vistas at different points in their journey (Tilley 1994:
chapter 5; Barclay and Hey 1999). It is not always obvious how much of
the surrounding landscape would have been clear of trees, but pollen analysis
undertaken close to the two largest monuments, the Dorset Cursus and the
Rudston complex, has produced unexpected evidence of an open landscape
(French et al. 2003; French and Lewis 2005; Bush 1988).
The second situation is perhaps more common. This is where the configuration
of cursuses or related monuments seems to follow the course of nearby
rivers (Brophy 2000). Sometimes the same principle extends to the siting of
several monuments within the same small area. Since people may have followed
watercourses as they travelled across country, it is tempting to suggest
that these earthworks monumentalised sections of their route, but this hardly
accounts for the distinctive forms and associations of these structures. Perhaps they really commemorated the paths that had been taken by the dead or by
previous generations as they first settled the land. The flow of the river might
even have provided a metaphor for life itself. It is an association that is found
in many different societies (C. Richards 1996a), but the only evidence for a
connection of this kind comes from the material, primarily ceramics, axeheads,
and human bones, that seems to have been deposited in the River Thames
(A. Barclay 2002: 88–93). Along its banks there are a number of cursuses.
Sometimes they run up to other monuments, usually enclosures or mounds
located towards one of the terminals. In relation to the course of the Thames,
anyone approaching those structures would have been moving ‘downstream’.
The same relationship can be recognised in other regions of England, such as
the Great Ouse valley (Malim 1999).
Until the fieldwork of the last few years, cursuses had been dated to the
Later Neolithic period, and even now it is difficult to appreciate the implications
of the new chronology. Three points are of fundamental importance.
The first is that linear monuments of the kind considered here may have originated
in northern Britain, where the earliest examples were probably wooden
structures, rather than the earthwork monuments that are more familiar in the
south. The sites which attracted most attention until recently were among the
last ones to be built. It is small wonder that they have been so difficult to
interpret. In lowland England they are associated with long barrows and allied
monuments, but their northern prototypes seem to form part of a still broader
pattern which had extended to a variety of large timber buildings.
The second point is that the main distribution of these structures is in exactly
the regions where little evidence of domestic buildings can be found. This may
be because they had been constructed in a different technique from Earlier
Neolithic dwellings in Ireland, but the character of excavated occupation sites
in Britain suggests that there were real contrasts in the pattern of settlement
between the different parts of the study area. In that respect it may be revealing
that major linear monuments are rarely found on the west coast of Scotland and
do not occur in the Northern Isles where the remains of houses survive. The
situation in Ireland is different again, for there may be some monuments of this
type, but they were never substantial features and it is uncertain whether they
are of the same date as their British counterparts (Newman 1999). Although
the situation could change as a result of fieldwork, it seems as if linear monuments
were uncommon in those regions where rectangular houses played a
prominent role. Conversely, such monuments are widely distributed in areas
where Neolithic settlements have left little trace. Here the domestic buildings
could have been more ephemeral, yet it was in such areas that the idea of the
house gave rise to an extraordinary public architecture.
The third point is even more important. Cursus monuments and bank barrows
may be much more common in Britain than they are in Ireland, but they
are unlike other Earlier Neolithic monuments in being an insular development Long mounds were built in Continental Europe, but they did not take this specific
form, nor did it happen during this particular phase. They are well known
in Poland and the north of France, but that phenomenon had no connection
with developments in these islands. In fact the closest structural parallels to the
British cursuses are found in the Netherlands, north Germany, Belgium, and
northern France, but they date from the first millennium BC (Roymans and
Kortelag 1999). In this respect it seems as if the insular Neolithic diverged from
developments in Continental Europe. The fact that this happened first in northern
Britain prefigures the development of henges considered in Chapter Three.
In fact it is necessary to consider the monuments of this period in relation to
at least two different axes. This section has been concerned with the importance
of specifically local developments, which probably began in Scotland. The
other axis connected these islands with the Continent and is represented by
the emergence of yet another class of earthwork, the ‘causewayed’ enclosure.
The contrast is all the more striking because this particular kind of monument
is so widely distributed. Not only does it take a remarkably stereotyped form,
but individual examples seem to have been used in similar ways.
CAUSEWAYED ENCLOSURES
It is an accident of history that the characteristics of the British Neolithic
were first defined by the excavation of a causewayed enclosure, so that the
earlier part of this period took its name from Windmill Hill, an earthwork
on the Wessex chalk (Whittle, Pollard, and Grigson 1999). Now it is apparent
that monuments of this form have a limited distribution across the two islands
and that they were not among the first monuments to be built. Rather, their
use commenced between about 3700 and 3600 BC. They are often found in
southern and midland England, but they are rare in the north, in Scotland,
and in Wales (Oswald, Dyer, and Barber 2001). Similar enclosures have only
recently been identified in Ireland, but some may have taken rather different
forms from the other sites. It also seems likely that they were widely adopted
in these islands after the first cursus monuments had been constructed in the
north, so that for a while the distributions of the two kinds of monuments may
have complemented one another.
Causewayed enclosures are so called because they are defined by interrupted
ditches (Fig. 2.17). It is not always clear whether the positions of the causeways
correspond to gaps in the bank, but it is obvious that this was more than a
constructional technique. On the Continent, palisaded enclosures are known
whose perimeters were broken in a similar manner. The same may have happened
at Knowth (G. Eogan 1984) and perhaps at other sites in Ireland. Some of
the enclosures are approximately circular or oval, but this is not always the case.
A recent study draws together much of the research on earthworks of this
kind (Oswald, Dyer, and Barber 2001). Two points need making at the outset These monuments were first identified from surface remains and were subsequently
recognised on air photographs, but there are other structures which
may have been closely related to them. These are found in western Britain
and are typified by the excavated enclosures at Carn Brea and Helman Tor in
Cornwall (Mercer 1981a; 1997). They were built on granite and were defined
by drystone walls as well as earthworks. They are of the same date as the
more conventional sites, but took a rather different form. Often the perimeter
extended between a series of prominent natural outcrops, and at Carn Brea the
enclosure banks were interrupted at intervals. Similar monuments have been
claimed in northwest England (Fig. 2.18; Pearson and Topping 2002) and the
west of Ireland (C. Jones 2004: 42–3) but remain unexcavated. The second observation concerns the roles played by these monuments.
It has always been supposed that they had a single function which could be
established by careful fieldwork. There is some reason to doubt this view. Their
Continental counterparts had a lengthy history extending from the late Linearbandkeramik
to the Copper Age. Whilst the earthwork perimeter retained its
characteristic form, the sites seem to have been used in quite differentways from
one period to another. There is also evidence for considerable regional variation.
Individual examples extend along a continuum from enclosed settlements
or even hillforts to ceremonial centres. Some of these sites were associated with
buildings, pits, and specialised deposits of artefacts; others were almost empty
(Bradley 1998: Chapter 5). It comes as no surprise that the same should be true
in Britain. Thus a walled enclosure like Carn Brea is associated with wooden
houses, defences, and areas of cultivated ground (Mercer 1981a), whilst the
earthwork monument at Etton in eastern England contained placed deposits
of artefacts, animal bones, and human remains. Although it was especially well
preserved, it did not include any buildings (Pryor 1998a). The site was liable
to flood and would probably have been inaccessible for part of the year.
This does not imply that individual enclosures in Britain and Ireland were
either ceremonial centres or settlements, although opinion has certainly oscillated
between these two extremes. There are several grounds for caution. Even
at carefully excavated monuments from which houses seem to have been absent,
the material that was deposited has a similar composition to the finds from
domestic sites. It seems as if the pottery is in the same styles, and a recent
study of the worked flint from the extensively excavated enclosure complex at Hambledon Hill failed to identify any features which distinguished it from the
assemblages found in settlements (Saville 2002). In the same way, both settlements
and enclosures can be associated with pits containing carefully selected
groups of cultural material. Both can provide abundant evidence of burning,
and human bones are recorded from enclosures as well as the places where
people lived. There are some exotic items from the earthwork monuments,
but they can be found elsewhere. The main distinction between causewayed
enclosures and open settlements is simply the greater formality with which
artefacts and other items were deployed (Edmonds 1999).
This has several aspects. There was the formality that was inherent in the
very design of these earthworks. They comprised one or more circuits of bank
and ditch interrupted at regular intervals by causeways and sometimes by more
formal entrances. The perimeter was entirely permeable. There are individual
monuments with several concentric earthworks where different activities may
have taken place in different areas, but this has only been demonstrated at
Windmill Hill where the character of the deposits seems to have changed as
one moved towards the innermost enclosure. Here it appears that the outer
part of the monument was associated with deposits of cultural material and
human remains, mainly those of infants. Craft production, butchery, and the
deposition of meat joints were associated with the next sector of the enclosure,
whilst it was the innermost space that was most obviously linked with the
domestic world (Whittle, Pollard, and Grigson 1999). At Etton, which had
only one ditch, the internal space was divided in half (Pryor 1998a). One part
included a series of deposits associated with the activities of the living, whilst
the other contained imported artefacts and burnt human and animal bones.
Again there is no uniformity beyond the basic point that such material had
been deposited according to certain conventions.
The extensively excavated enclosure at Etton is revealing in another
way, for the excavator has questioned the conventional assumption that this
monument was ever conceived as an enclosure in the way that archaeologists
had supposed. He suggests that it is better thought of as an arrangement
of separate pits, not all of which were dug or maintained simultaneously.
Rather than forming a continuous (or discontinuous) bank, the excavated
soil was spread around the rims of these features to create a platform from
which people could view their contents. The separate pits or ditch segments
contained a wide variety of deposits which had been carefully placed there,
including human skulls, meat joints, axeheads, and complete pots. These were
often laid out in formal patterns then covered over. The individual sections
of the perimeter were repeatedly recut, allowing the original offerings to
be inspected and further material to be added. A similar process has been
identified at most extensively excavated sites. Although there is little evidence,
there is no reason to suppose that material culture was deployed with any less
formality at the walled enclosures of western Britain. Having stressed the ways in which these deposits had been treated, it is
important to add that certain monuments changed their form and character
over time. They may have been rebuilt with continuous earthworks at a late
stage in their development, and it was then that most of the causeways in
their ditches were removed, creating a continuous barrier. It seems possible
that the banks were rebuilt with vertical outer walls or timber revetments, and
formal entrances may also have been established. This has not been observed
at many sites, but it does raise the possibility that some of the earthworks
were gradually changed into defensible enclosures. It may be no accident that
among them were Hambledon Hill and Crickley Hill, both of which occupied
conspicuous positions that were reused as Iron Age hillforts (Oswald, Dyer, and
Barber 2001: 127–30). Other examples may include Carn Brea and Hembury,
again in western England. It is particularly interesting that several of these sites
include unusual concentrations of arrowheads, suggesting that they had been
attacked. The bodies of two men who had clearly been killed by arrows were
found in the ditch of the Stepleton enclosure at Hambledon Hill (Healy 2004:
32). There may have been a high level of violence during this period, but the
evidence has often been overlooked. It is certainly consistent with the injuries
shown by human bones found in other contexts in Britain (Schulting and
Wysocki 2005).
Just as causewayed ditches could be rebuilt as defensive earthworks, entire
enclosures could be duplicated. Thus there are a number of sites where at least
two of these monuments were built close to one another or even side by side.
Examples include Hambledon Hill and Rybury, where they are located some
distance apart, and Hembury, Court Hill, Fornham All Saints, and Etton, where
the two structures were adjacent to one another (Oswald, Dyer, and Barber
2001: 112–13). Only at Hambledon Hill is there evidence that they were built
at different times and had different contents (Healy 2004). Enclosures could
also extend their area, and Oswald, Dyer, and Barber have suggested that a
number of examples with several concentric circuits may have developed over
a significant period of time (2001: 75–7). In most cases they seem to have been
enlarged, and, as this happened, they might assume a more regular ground
plan. These authors comment on a number of examples which would have
appeared to be approximately circular to anyone inside them or encountering
them from a distance.
Early fieldwork at causewayed enclosures was concerned with investigating
their credentials as the elusive Earlier Neolithic settlements of lowland England.
Although they produced what seemed to be ‘domestic’ assemblages, from an
early stage it became apparent that therewere too many anomalies for this interpretation
to be warranted. Instead of the general scatter of animal bones that
might have been found at a living site, substantial meat joints had been committed
to the ground and sometimes entire animals had been buried there. As
well as the remains of animals there were human bones, sometimes burials but more often disarticulated fragments similar to those associated with long barrows.
All this material had been placed in the ground with some deliberation,
and normally it had been carefully covered over. It seemed more likely that these
deposits resulted from specialised events which may have included episodes of
feasting, the sacrifice of animals, offerings of food, and the celebration of the
dead (Edmonds 1999). There is even evidence of opium poppy from the enclosure
at Etton (Pryor 1998a: chapter 10). These ideas received powerful support
from excavation at Hambledon Hill which showed that in the enclosure ditches
there had been a large quantity of human skulls (Healy 2004).
Similar anomalies soon became apparent among the artefacts from these
monuments. On a site like Etton there was a strong representation of decorated
pottery, but perhaps more important was the observation that atWindmill Hill
there were few large vessels suitable for storage and many small bowls that
could be used for serving food and drink (Howard 1981: 11–20). Similarly, the
presence of nonlocal artefacts raised an interesting problem. From the early
years of the excavation of Windmill Hill it had been apparent that stone axes
from distant regions of Britain were distributed on and around the site, and for
that reason the enclosure, and others like it, were identified as nodal points in
the Neolithic ‘axe trade’ (Bradley and Edmonds 1993: 50–2). This was illogical
since these objects has clearly been brought to these sites but had never left
them again. Closer attention to their contexts showed that many of them had
been deposited there together with the other material discussed in this section.
At Etton, it even seems as if they were destroyed by working them down and
placing the resulting fragments in pits together with burnt human bone (Pryor
1998a: 260–8). The walled enclosure at Carn Brea may have been linked with
the production of pottery and axeheads in the surrounding area (Mercer 1981a:
189), and what may be a similar site on Carrock Fell in northwest England was
located on a mountain top close to a Neolithic axe quarry (Fell and Davis 1988;
Pearson and Topping 2002). Such connections made it even less likely that these
monuments had been ordinary settlements. Perhaps they were aggregation sites
where public events took place.
There are good reasons for accepting this view. The development of causewayed
enclosures happened during a period which saw significant changes in
the nature of material culture. After an initial phase in which pots had taken
a similar form throughout most parts of Britain and Ireland, there are the
first signs of regional diversity, reflected by the first adoption of more local
styles of decoration (Gibson 2002a: chapter 4). At the same time, there is
evidence for a marked increase in the production and exchange of stone
axes. The products of individual sources seem to have been distributed over
larger areas, and fieldwork at the Cumbrian quarries even suggests that now
they were being made in larger numbers and with greater skill (Bradley and
Edmonds 1993: chapters 6 and 7). Both these new developments are evidenced at causewayed enclosures, although there is a danger of circularity as these sites
provide so many stratified contexts. If the artefact assemblage seems to have
been more diverse, so was the human population associated with these monuments.
Michael Richards has shown that the bones from different tombs show
quite separate dietary signatures from one site to another (2000: 125–35). Those
found in the enclosure complex at Hambledon Hill suggest that the people
whose remains were deposited there had lived in a whole range of different
environments or had engaged in different methods of food production. Further
work on a sample of skeletal remains from England suggests that Neolithic
people also had a wide variety of lifestyles (Wysocki and Whittle 2000).
In recent years more attention has been paid to the setting of these monuments,
and that tendency has only increased with the expansion of field
archaeology. It soon became clear that whilst some of the enclosures included
enormous collections of Earlier Neolithic artefacts, they were actually set apart
from the ordinary settlements of the same period. They might be located in
isolated positions towards the margins of the settled landscape and could often
be seen from a distance. Some of the sites commanded an extensive view, and
others were intervisible with the stone sources where flint axes were made,
yet environmental evidence suggests that a number of them had been located
within woodland clearings (K. Thomas 1982). Far from being the ‘central
places’ of this period, certain sites were located in peripheral positions and
may have been built in neutral locations in between the main concentrations
of population. That even applies to an enormous monument complex like
Hambledon Hill (Healy 2004).
It has long been accepted that some causewayed enclosures are closely linked
to the positions of long barrows, but this statement needs some qualification.
They were first built after the earliest mortuary monuments, and while a few
examples are located close to burial mounds of the orthodox type, they are
also paired with more specialised earthworks. These are usually smaller oval
or round barrows of a type that is often associated with a distinctive burial
rite. Instead of containing the disarticulated bones of a number of different
individuals, they cover the graves of a small number of people and sometimes
only one person. They also depart from normal practice because they are
associated with grave goods, usually decorated pottery, an arrowhead, or a
flint knife. This was demonstrated at Abingdon in the Thames Valley where
the ditch enclosing a burial mound of this type contained the same series of
deposits as the adjacent causewayed enclosure (Bradley 1992).
One result of the recent expansion of developer-funded archaeology is that
more is known about the areas outside causewayed enclosures. Close to the
oval barrow at Abingdon there seems to have been a flat cemetery of similar
age to the principal monument. The graves were not covered by any kind
of mound, and their chronology was only established by radiocarbon dating In the same area was what is best described as a ‘mortuary house’ of the
kind associated with long barrows (Barclay and Halpin 1999: 27–31)). This
contained the disarticulated bones of a number of individuals which had been
placed there during the period in which the enclosure was in use. Again there
was no evidence of any barrow. What may have been a similar feature has been
observed outside a causewayed enclosure at Husbands Bosworth in the English
midlands (Buckley and George 2003: 138–40). At Thornhill in the north of
Ireland a group of houses enclosed by a number of discontinuous palisades was
associated with further structures which may have been of this type (Logue
2003: 151–3).
Another result of fieldwork in the vicinity of causewayed enclosures has
been the recognition that the distribution of pits and artefact scatters may
extend beyond the limits of these earthworks. This is most obvious at Etton
in eastern England where a large number of pits have been found (French
and Pryor 2005: 167). They have the same range of contents as the deposits
associated with the enclosures. There is similar evidence from Robin Hood’s
Ball in Wessex ( J. Richards 1990: 61) and Husbands Bosworth in the East
Midlands (Buckley and George 2003: 138–40). At Eton in the Thames Valley
extensive and well-preserved midden deposits have been discovered in between
the sites of two causewayed enclosures (Allen, Barclay, and Landin-Whymark
2004).
The relationship between causewayed enclosures and cursus monuments
raises further problems (Fig. 2.19). The dates of the early timber cursuses
in northern Britain should overlap with those from causewayed enclosures
during the second quarter of the fourth millennium BC. To a large extent their
distributions did not coincide. The later cursuses, however, were considerable
structures and are found across virtually the entire distribution of causewayed
enclosures in the south. Here there seem to be two principal relationships
between these very different traditions of earthwork building. In some areas,
for example the Thames Valley, the cursus monuments seem to avoid the places
where causewayed enclosures had been built, so that it is perfectly possible for
both to have been used simultaneously. In other cases it is clear that a cursus or
related monument cuts across the position of an existing causewayed enclosure,
suggesting that it had fallen out of use.
The clearest examples of this relationship are at Ramsgate in Kent (Dyson,
Shand, and Stevens 2000), Etton in eastern England (Pryor 1998a), and Maiden
Castle in the south (Sharples 1991: 255–7). The excavation at Ramsgate is not
yet published, but at Etton, two cursuses were constructed running more or less
parallel to one another following the former course of the RiverWelland. The
Maxey cursus ended alongside the enclosure, whilst the Etton Cursus, which
may have continued the axis established by its neighbour, crossed the position
of that enclosure, cutting it in half. In principle, that should have involved a drastic transformation, but it is intriguing that the older earthwork had already
been divided into two parts, one associated with deposits connected with daily
life and the other containing human and animal remains (Pryor 1998a: chapter
16). The new division followed a different course but may have maintained
the same concept. This is less likely to have happened at Maiden Castle, where
it seems likely that a long barrow had been built just outside the perimeter of
another causewayed enclosure. That is not an uncommon pattern, but in this
case the mound was extended at both ends so that in one direction it ran straight
across the existing monument and, in the other, continued for a considerable
distance outside it (Sharples 1991: fig. 33). Something rather similar may have
happened at Fornham All Saints in eastern England. Here two causewayed
enclosures seem to have been built side by side, one of them as an annexe to
the original monument. The same axis was followed by a cursus which seems
to have cut across both earthworks following a path which runs parallel to the
River Lark. There were two other monuments of the same kind nearby, one
of them running past an older enclosure and perhaps extending to the river
bank (Oswald, Dyer, and Barber 2001: 75–6 and 134–5). It seems clear that an
important transformation of the landscape was under way.


THE FUTURE: CURSUSES, ROUND BARROWS,
AND CIRCULAR ENCLOSURES
Although cursuses and causewayed enclosures may have coexisted for a time, it
was the linear monuments that retained their significance for a longer period.
That is not to say that the sites of older enclosures were necessarily deserted,
for many of them do produce finds of artefacts of Later Neolithic and Earlier
Bronze Age date. Nor were these places entirely without later monuments,
generally burial mounds, but it was rare for the original structure to be refurbished
during these episodes of reuse. It seems more likely that attention shifted
towards the positions of the newly established cursuses.
In fact one reason why such structures were originally dated to the Later
Neolithic periodwas simply the fact that they remained the focus for deposits of
distinctive artefacts. These might be placed in and around their earthworks, and
there even seem to be cases in which their ditches were recut. The problem was
made even worse by the character of some of the pottery associated with these
monuments, or with the smaller structures that seem to have been built around
them. This belongs to a loosely defined tradition of lavishly decorated ceramics
collectively known as Peterborough Ware, although it can be subdivided into
a number of regional styles. Although decorated pottery came into use during
the currency of long mounds and causewayed enclosures, the greater part of the
Peterborough tradition had been assigned to the Later Neolithic period, and
it was usually supposed that it did not go out of fashion completely until the
Earlier Bronze Age. Radiocarbon dating suggests that this schemewas incorrect
and that pottery of this type may have been current at a significantly earlier date
than was once supposed; it may have originated about 3400 BC and have gone
out of use during the Later Neolithic period (Gibson and Kinnes 1997). The
implications of this work have not been properly assimilated, for this means
that a number of traditions of monument building could have originated by
the later fourth millennium BC and may have ended much sooner than was
originally believed.
These rather technical arguments affect a series of earthworks which are
commonly associated with cursus monuments, although it cannot be established
that they were of exactly the same dates as one another. They also raise
questions of terminology and interpretation.
The problem of terminology needs to be considered first. Here it is necessary
to consider two problematical categories that appear in the current literature:
round barrows and ‘hengiform’ enclosures (Fig. 2.20). Round barrows are
simply circular mounds, which are generally delimited by a ditch. Although it
is often thought that they are associated with individual burials, that is not always
the case, and particularly along the North Sea coast individual examples may
cover the remains of ‘mortuary houses’ formed by split tree trunks exactly like
those associated with long barrows. They may also include the disarticulated remains of a number of people (Kinnes 1979). The same applies to some of the
small circular cairns in western Britain, which can be associated with simple
stone-built chambers (Darvill 2004: 60–2). Both types can also be found buried
beneath more impressive long mounds. There are two reasons for the confusion. The first results from the chronological
problems mentioned earlier. So long as PeterboroughWare was considered
to be a Later Neolithic tradition, it was very tempting to compare these monuments
with the round barrows associated with Bell Beaker pottery and early
metalwork in the later third millennium BC. Rather than viewing them as a
new tradition that had been introduced from the Continent, it seemed likely
that these monuments were built in a long-established local style, in which
case only the artefacts selected for burial would have been a novel development
(Kinnes 1979). That idea is no longer tenable, and there is little or no
evidence for a continuous tradition of round barrows extending through the
Later Neolithic, although it is uncertain quite when their construction lapsed.
The second source of confusion was that a number of circular mounds
really did contain individual burials with grave goods, although some of these
monuments may actually have included a succession of such deposits. The
mounds were of various sizes, from the ‘great barrows’ of northern England to
the smaller examples recorded in East Anglia, Wessex, or the Thames Valley.
They are sometimes claimed to be a particular feature of northeast England,
but the results of commercial fieldwork have shown that this is not the case.
Moreover, some of the Neolithic ‘round barrows’ on the Yorkshire Wolds
excavated in the nineteenth century have proved to be only the most prominent
parts of denuded long barrows. This has been established by recent fieldwork
and is also apparent from the results of air photography (Stoertz 1997: 23–4).
That is not to deny that individual burials were important in this area, but one
reason why they have played such a prominent role is that northeast England
saw a major campaign of fieldwork during the Victorian era. The excavations
of the last twenty years have shown that similar graves can be found in most
parts of Britain.
In Ireland, there have been similar problems, but in this case they have been
resolved by radiocarbon dating. These concern what are called Linkardstown
cists (Raftery 1974). They are individual inhumation burials placed within
massive stone coffins and buried beneath a substantial round cairn. These burials
are quite distinct from the main classes of megalithic tomb, andwere inaccessible
once the monument had been built. Again there was the same temptation to
interpret these single burials in the light of superficially similar practices during
the Earlier Bronze Age, but radiocarbon dating has shown that in fact they
belong to the period between about 3600 and 3300 BC (Brindley and Lanting
1990). They occur in a restricted zone crossing the middle part of the island
from east to west, but may be of similar date to their British counterparts.
‘Hengiform enclosures’ raise a similar problem, for the term implies that
they are related to the enclosures known as ‘henges’ which were built in the
Later Neolithic period. They are supposed to be miniature versions of these
monuments (Wainwright 1969), but the morphological argument is far from
convincing, for in some cases they are simply circular earthworks with one or more entrances; others were defined by causewayed ditches. The chronological
argument is on even weaker ground, for the link with henges was postulated at
a time when it was believed that the currency of PeterboroughWare extended
well into the third millennium BC. Now that seems unlikely, meaning that in
lowland England at any rate the currency of these two forms of monument
need not have overlapped. Again the conventional terminology conceals some
dangerous assumptions about the chronology of different earthworks.
In this case there are even more problems to consider. Most of the hengiform
enclosures have been identified from crop marks or by excavations on sites
which had been levelled by cultivation. For that reason it is often uncertain
whether these were the remains of circular enclosures that were open at the
centre or were ploughed-out round barrows of the kind described already.
Moreover, one of the defining characteristics of a henge monument is the
presence of an external bank and an internal ditch. In many cases it is impossible
to reconstruct the formof the original earthwork with any confidence (Bradley
and Chambers 1988). All that is clear is that the ditch was repeatedly recut.
Sometimes the monument also increased in area as the perimeter was rebuilt. If
broader comparisons are needed, itwould be more consistent with the evidence
to suggest that these were miniature versions of causewayed enclosures. That
would account for the interrupted ditches which define a number of these
monuments and for the structured deposits of artefacts, human bones, and
food remains that are occasionally found there. It may be that these represent
the end of a long tradition. It is certainly true that they are commonly associated
with cursuses rather than large curvilinear enclosures.
There is another problem, too. In their desire to fit excavated monuments
into well established ‘types’, fieldworkers have exercised some latitude in their
identification of round barrows and ring ditches, for a number of these monuments
are not actually circular. That provides an important clue, for it is already
apparent that Earlier Neolithic mortuary monuments included a number of
oval mounds which are unhelpfully described as ‘short long barrows’. Until
recently these seemed to represent the end of the tradition of building elongated
mounds, but there is reason to doubt this now. Although certain examples are
associated with individual burials with grave goods, others include more complex
arrangements of human bones. Earthworks of this kind can be associated
with settings of split tree trunks exactly like those associated with more orthodox
monuments, and the radiocarbon dates from the two groups of earthwork
overlap. A few of these mounds were even rebuilt to a circular ground plan
during a secondary phase. As so often, an unnecessarily rigid typology conceals
what was probably a continuum. The point is illustrated by the excavation of
a valley-bottom site at Raunds in the east midlands which revealed a whole
series of ‘non-standard’ Neolithic monuments (Healy and Harding in press).
Oval mounds were built on various scales, from the substantial structures
that are found at the southern terminal of the Dorset Cursus (Barrett, Bradley and Green 1991: chapter 2) to much smaller and lower mounds like the example
associated with the Abingdon causewayed enclosure (Fig. 2.20; Bradley
1992). Even so, they share one important characteristic, for the earthwork was
commonly flanked by a ditch on three sides, leaving one end of the structure
open. That space was usually filled during a subsequent phase, so that the completed
structure had a continuous perimeter, although it was often broken by
narrow causeways. Such monuments are commonly found close to cursuses
and may even share the same axis as those monuments. Cursuses can extend up
to earthworks of this kind and may even incorporate them in their terminals.
A good example of these relationships may be found at Eynesbury in eastern
England where a U-ditched long barrow seems to have been built on the same
alignment as a nearby cursus (Ellis 2004). An even shorter mound of similar
type was located beside that earthwork with its long axis at right angles to
it. At some stage (the dating evidence is not altogether satisfactory) the larger
barrow seems to have been lengthened and contained within a second cursus
monument. The smaller mound was closed by excavating a narrow ditch across
its open end (Fig. 2.21)
Again the evidence is rather ambiguous. On most sites the excavated remains
are reduced to features cut into the bedrock, and it is not always clear whether
they enclosed a mound or whether some monuments were open at the centre.
Oval mounds certainly survive as earthworks, but low platforms and open
enclosures have also been recorded by field survey. This may reflect the amount
of human labour that could be invested in the creation of individual structures,
but it is just as likely that any mounds might have been composed largely of
topsoil and turf. In that case nothing would survive a long period of ploughing.
The argument is especially troubling as it is known that some of the surviving
long mounds, among them examples of considerable size, were built in exactly
this way. Despite all these uncertainties, it seems likely that oval mounds or
enclosures were used during the same period as Neolithic round barrows and
long barrows and were associated with similar mortuary rituals. What is not so
clear is when they went out of use.
Lastly, it seems as if a few large circular enclosures were built during the
currency of cursus monuments, and sometimes in their vicinity. Again there
are problems of terminology, for they have been described as early henges.
At present two groups of enclosures pose particular problems. In Wessex, the
earliest enclosure at Stonehenge is precisely circular and is defined by a causewayed
ditch with three main entrances and an internal bank (Cleal, Walker,
and Montague 1995: chapter 5). Another enclosure at Flagstones, closer to
the south coast, takes the same form, and both are associated with a series of
placed deposits of human and animal bone very similar to those from conventional
causewayed enclosures (R. Smith et al. 1997: 27–47). The only problem
is that they are unusually late in date and may belong to the very end of
that tradition. In the same way, several larger enclosures in northeast England are characterised by a circular ground plan and an interrupted ditch. One of
these monuments encloses the enormous Neolithic round barrow of Duggleby
Howe (Fig. 2.22; Kinnes et al. 1983), and the same relationship can be suggested
between monuments at Maxey in East Anglia (Pryor et al. 1985) and perhaps at
West Cotton (Healy and Harding in press) and Dallington in the East Midlands
(Oswald, Dyer, and Barber 2001: fig. 3.4). Yet another circular enclosure of
this kind lies astride the Thornborough cursus. These earthworks are especially
confusing, for in northeast England they seem to have been replaced by
henge monuments of conventional type ( J. Harding 2000: 90–5), in the same
way as may have happened at Stonehenge. This evidence is discussed in
Chapter Three.
Those developments were still in the future, and it seems unlikely that they
had a significant impact on the course of events during the Later Neolithic
period. They are better thought of as representing the last expressions of a
whole range of ideas that had been present in Britain, and sometimes Ireland, from the first introduction of agriculture. A different kind of sequence will
be considered in Chapter Three, but first it is necessary to sum up the main
themes so far.
THE PAST: ANCESTORS AND ORIGINS
Gabriel Cooney called one chapter in his influential study of Neolithic Ireland
‘The dead are everywhere’ (2000: 86–126). It could not be more appropriate.
Many of the settlements and monuments considered in this chapter have been
associated with human remains, often incomplete bodies or isolated bones.
It seems as if what have been described as burial mounds were really more
specialised monuments to which the dead could be brought and from which
relics might be taken away. Although there are intact bodies at long mounds
and megalithic tombs just as there are a small number of flat graves during this
period, it seems as if individual bones could also be treated like artefacts and
passed from one location to another before their final deposition.
Why should people have shown somuch concern with the dead, particularly
when finds of Mesolithic burials in Britain and Ireland are very rare? On one
level it is easy to argue that this is a general characteristic of agricultural societies.
The adoption of farming required forward planning, and necessitated a new
sense of time quite different from the annual cycles that dominate the lives
of foragers. Land clearance was a lengthy process which extended from one
generation to another, so that the success of food production was partly due to
the efforts of earlier generations. Strategic decisions had to be made which also
influenced activities in the future. A certain proportion of the domesticated
animals had to be maintained over the winter months if people were to have
sufficient breeding stock. Some of the harvest needed to be stored so that it
could be sown in the following year, and areas of land might have to remain
fallow in order to regain their fertility. In that sense one could argue that early
agriculturists place more emphasis on ancestry than mobile hunter gatherers
(Meillassoux 1972). Moreover, they feel a sense of obligation to their ancestors,
to whom they owe their prosperity.
The problem with this model is its wide application. James Whitley (2002)
has complained that British prehistorians invoke the importance of ancestry
in virtually everything they write, so that the significance of these ideas is
asserted rather than argued. There is some justification for this view: ancestors
are certainly fashionable in contemporary archaeology. Are there other ways
in which to explain the distinctive character of the Earlier Neolithic period in
these islands?
This chapter began by discussing the origins of the insular Neolithic, contrasting
the arguments for acculturation and continuity with those which favour
a more radical disruption. To what extentwere changes brought about by settlement
from overseas? The conventional approach to this question is remarkably unhelpful. In principle, it should be possible to define the source, or sources,
of any migrants through a close comparison between the material culture of
Neolithic Britain and Ireland and its counterparts on the European mainland.
Although there are a number of general similarities, particularly among the
undecorated ceramics which were the earliest in these islands, attempts to
define more exact links have so far failed. That is probably because the study
area would have been accessible from so many different areas of Continental
Europe. These included parts of the mainland whose inhabitants may have had
few contacts with one another. Indeed, it is ironic that it has been easier to
suggest the sources of the cattle introduced from overseas than it has been to
trace closely related artefacts. Not surprisingly, they connect northern France
and southern England (Tresset 2003).
That leaves the question of monumental architecture. This was an entirely
new phenomenon at the beginning of the British and Irish Neolithic. There
were no precedents apart from a setting of massive timbers on the site which
was later to become Stonehenge and a few post holes of similar date, and even
these were thousands of years earlier than the time considered here (Allen and
Gardiner 2002). Again it is important to emphasise that the only signs of social
complexity in the insular Mesolithic come from the first part of that period.
The new developments were of three main kinds: long mounds, causewayed
enclosures, and cursuses.
The first two of these traditions originated in Continental Europe, whilst
the third was possibly inspired by the building of monumental ‘houses’ in Scotland
and was considered in the previous section. There has been considerable
discussion of the mounds and enclosures and it is not necessary to repeat all
the arguments here. Nonetheless there is a strong case that long barrows first
developed as representations of domestic buildings. A common objection is
that these types are not contemporary with one another; long barrows originated
when long houses were going out of use. That is probably true, but this
very observation helps to account for the relationship between them. On a site
like Balloy in north-central France where such barrows were constructed over
the sites of older domestic buildings, it seems clear that people had intended to
commemorate the characteristic form of a kind of dwelling that was no longer
being built (Mordant 1998). Rather than confining their attention to classifying
these physical structures, prehistorians should consider the wider importance
of the house among early farmers. It was a metaphor for a certain way of
living, and it may have been this concept that was celebrated as the pattern
of settlement changed. It seems possible that the houses of the dead referred
back to a past when buildings of similar proportions had been inhabited by the
living. If so, their construction was also a statement of origins.
The same could apply to Continental causewayed enclosures, for they seem
to have had a similar history. The first were simply the earthwork perimeters
around settlements containing long houses, but soon the relationship between these features became more complicated. Empty enclosures might be built
at some distance from groups of domestic buildings, or they might enclose
the positions of settlements that had recently been abandoned. In a parallel
process to the development of long barrows, these earthworks assumed a life
of their own as the tradition of building massive dwellings lapsed. By this time
the earthworks had assumed a greater formality and their characteristic ground
plan was well established, but most of them were no longer used as settlements.
Rather, they seem to have been aggregation sites for a wider population: places
where feasts took place, dead bodies were exposed, artefacts were made and
exchanged, and where increasing numbers of objects, animal bones, and human
remains were deposited in pits and ditches. Whilst these were very different
from the practices that took place in older settlements, the way in which the
perimeter maintained its traditional form suggests that some links were still
maintained. Perhaps these enclosures evoked a tradition of communal living
that had existed in the past but which was no longer considered feasible or even
appropriate (Bradley 1998b: chapter 5). Again it seems as if the architecture of
these monuments made a direct reference to the past.
Why is this so relevant to Britain and Ireland, where there had never been
a tradition of long houses or enclosed settlements? Perhaps it is because these
monuments were still considered to refer to a distant past. Maybe their construction
and use evoked an origin myth, as it may have done in Continental
Europe. That could account for the continuous recycling of the remains of
the dead and would suggest that the Earlier Neolithic population showed an
almost obsessive concern with its own history. If so, it seems difficult to argue
for complete continuity with the Mesolithic period. If the inhabitants of both
islands celebrated a distant origin in other times and places, who is to say that
they were wrong?