THE PREHISTORY OF BRITAIN AND IRELAND pt.3

TIME AND SPACE
If Chapter Two was really about time, Chapter Three is more concerned with
space. The earlier part of the Neolithic period was interpreted in terms of origins
and ideas about origins. If people introduced farming from the Continent,
they also seem to have acknowledged their past by the kinds of monuments
that they built. The construction of long mounds may have commemorated
the houses of their ancestors, and causewayed enclosures the settlements in
which earlier generations had lived. Chapter Two juxtaposed archaeological
arguments about the character and chronology of the Earlier Neolithic period
with an interpretation of the ways in which prehistoric people might have
thought about their own histories (Figs. 3.1 and 3.2).
Chapter Three takes a different turn. It considers an even longer sequence,
from the years preceding 3000 BC up to 1500 BC, but for part of that time
communities in these two islands seem to have closed in on themselves, and
links with Continental Europe may have lapsed. In their place there were
stronger connections between different parts of Britain and Ireland. Those
new alignments are summed up by the title ‘North and South’. The situation
did not change significantly until the adoption of metals in the late third
millennium BC.
In another way, the processes described in both chapters do have something
in common. Chapter Two was mainly concerned with the ways in which
unfamiliar practices were adopted in both islands. They included new ways
of living, new methods of food production, novel attitudes to the dead, and,
most of all, the creation of a distinctive range of monuments. These buildings
provided the clearest indication of a fresh way of thinking about the world.
There were certain changes during the crucial period between 4000 and 3300
BC, and yet these local differences seemed less important than more general
trends. But there is another way of interpreting some of the evidence. It is an
approach that can also be taken to the archaeology of the following period.

If the contents of long barrows and long cairns suggest that personal differences
were played down, this may not have gone unopposed. During the midfourth
millennium BC there were individual burials in Britain, most of them
in round barrows (Kinnes 1979). Such monuments were widely distributed,
and by 3000 BC they were associated with a distinctive range of grave goods
(Kinnes 2004). The idea that they were a special feature of northern England
has been discredited by the results of contract archaeology. The importance of
individual burials seems to have increased as the process of monument building
intensified and reached its limits. At the same time, these graves are often
found close to earthworks of long-established forms, including causewayed
enclosures and cursuses, or were even added to existing mortuary monuments
like those at Whitegrounds (Brewster 1984) or Biggar Common (D. Johnston
1997). One way of thinking about this sequence is to suppose that the very
process of organising and executing the building of these enormous structures
emphasised the roles played by particular individuals. The creation and use
of these monuments may have helped to highlight the importance of certain
members of society ( J. Barrett 1994: chapters 3 and 4). Perhaps it was these
people whose influence was signified by a new burial rite.
That sequence is widely accepted in British archaeology. Unfortunately,
these developments have also been misunderstood, for it has been claimed
that the adoption of individual burials in round barrows was something that
happened only once: it seemed as if it was established by 3500 BC and that
the tradition continued uninterrupted until the introduction of metalwork a
thousand years later (Kinnes 1979). To some extent the argument was based
on the chronology of PeterboroughWare which was associated with certain of
these sites, for it was thought to have been used throughout the Later Neolithic
period. The argument is no longer sustainable (Gibson and Kinnes 1997), and
in fact there is very little evidence of single graves between about 3000 and
2400 BC. Although a new burial rite may have been established during the
currency of causewayed enclosures, cursuses, and even long barrows, it seems
to have lapsed not long afterwards.
This has important implications. Individual burials were again deposited in
round barrows from the late third millennium BC, but they cannot have conformed
to an established insular tradition. Rather, they seem to have resulted
from contacts with mainland Europe which resumed at about this time. It may
be no accident that they occurred during and just after another peak in the
construction of monuments. This was a process that had happened once before.
When it took place in the middle of the Neolithic period, its effects had been
quite short lived, but when a similar development occurred between 2500 and
2000 BC it had a lasting impact (Barrett 1994: chapters 3 and 4).
The details of these changes are obscured by the terminology used in Britain
and Ireland. The major phase of monument building is normally treated as a
‘Neolithic’ phenomenon, whilst the development of mortuary monuments can be assigned to an Early or Earlier Bronze Age because they are associated with
the first metalwork (the term Earlier Bronze Age is preferred here because the
Later Bronze Age is discussed in Chapter Four). It is clear that these different
elements overlap in time, even within the same region. Thus in the small area
close to Stonehenge, large structures built out of timber and stone date from
the same period as the first graves with metal artefacts (Cleal, Walker, and
Montague 1995). It would be wrong to organise this account around a period
framework which distinguishes between phenomena that were contemporary
with one another. In any case the character of early metallurgy differed between
the two islands, although gold and copper may have been worked in both
areas before the production of bronze. For that reason Stuart Needham (1996)
refers to a ‘metal-using Neolithic’ which continued until the end of the third
millennium BC.
Such differences of terminology reveal a deeper unease. Some of the burial
practices that were to be important in the second millennium were clearly
established during a time of change which is difficult to characterise by a period
label, and yet these developments ran in parallel with others of great antiquity.
There is no obvious point at which to interrupt the narrative, and to do so
would obscure an important trend, for it was also between 2500 and 2000 BC
that a major phase of monument building was followed by the deposition of
burials with grave goods. Rather than seeking to keep those different elements
apart, this chapter considers them together, for both were vital components in
a more general sequence of change.
If this chapter cuts across conventional period boundaries, it also eschews
established regional divisions in order to identify the long distance networks that appeared and disappeared between 3300 and 1500 BC. Some of these concerned
different parts of Britain and Ireland and did not involve any contact with the
European mainland. Others linked some of these areas into an international
arena. The importance of such changing alignments is most apparent at a large
scale, and yet their impacts would have varied from one area to another. These
are some of the issues that have to be considered here.
HOUSES AND THE CHARACTER OF SETTLEMENT, 3300–2000 BC
Despite these regional alignments, there is one general trend that can be recognised
throughout Ireland and Britain. The character of prehistoric settlements
seems to have changed.
The clearest evidence of this transformation comes from Ireland, for it is here
that the nature of these settlements can be compared directly with the pollen
record. As noted in Chapter Two, in the first half of the fourth millennium BC
there was a peak of clearances associated with cereal pollen. The same phase
saw the building of substantial rectangular houses, whose remains are being discovered
increasingly often. After that time the evidence takes a different form.
Newclearings are certainly known but there is less direct evidence of crop cultivation
and in some areas human activity made a smaller impact. O’Connell and
Molloy (2001) suggest that there may have been more emphasis on stock raising.
The character of domestic dwellings changed too, for by the Later Neolithic
period sturdy rectangular buildings were often superseded by ephemeral and
generally smaller oval and circular structures, although slight rectangular houses
have also been identified which date from the earlier third millennium BC.
Buildings of both kinds are more difficult to recognise than their predecessors.
Excavation at Knowth in the Boyne Valley has shed some light on this problem.
It has defined a structural sequence which runs in parallel with important
changes in material culture (Fig. 3.3). The first phase is characterised by undecorated
pottery of the kind which is found throughout these islands during the
earlier fourth millennium BC. It is associated with a series of rectilinear buildings
and two discontinuous palisades which can be compared with causewayed
enclosures found on the Continent. Some of these structures had been disturbed
by later land use, but others were preserved beneath a series of tombs
(G. Eogan 1984; Eogan and Roche 1997).
The main passage tomb at Knowth also sealed the remains of stake-built
round houses associated with decorated pottery which was probably an Irish
counterpart of Peterborough Ware. The circular buildings overlapped one
another and not all of them survived intact, but they were so much slighter than
their predecessors that no trace of these structures remained beyond the limits
of the mound. Since that earthwork reached its final form about 3000 BC,
it follows that this kind of domestic architecture must have been established
before then. It seems possible that the structures identified at Knowth formed only part
of a more extensive settlement. That is certainly suggested by the remains
of circular buildings associated with two of the other concentrations of passage
tombs in Ireland. A series of rather ephemeral round houses has been
identified close to those on Knocknarea (Bergh 1995), and well over a hundred
round houses or stone enclosures have also been identified just below
the cemetery at Carrowkeel. It is not known how many of the structures
were used at the same time, nor is it certain how they were related to the
tombs, but these buildings seem to date from the late fourth and earliest third
millennia BC (Stefan Bergh, pers. comm.). They may have been used by
visitors to these cemeteries, but the best comparison could be with settlements
in Orkney where a similar change from rectangular to circular houses
took place in the late fourth millennium BC. Here there were small villages
of round or oval houses. The best known is Skara Brae (Clarke and
Sharples 1990), but the excavated settlement of Barnhouse is perhaps more
relevant to the argument. It was established by 3100 BC and was located
only a short distance away from the passage tomb of Maeshowe (C. Richards
2004b).
Neolithic houses are preserved in Orkney because they were built of stone
. Others are difficult to identify because they were constructed of wood, yet examples have now been recognised on the Scottish mainland and
also in England and Wales. The earliest circular dwellings are associated with
Peterborough Ware and others are found with Grooved Ware, the ceramic
tradition that replaced it during the earlier third millennium BC (Darvill 1996).
Although therewere still some oval and rectangular structures, circular buildings
are being discovered in increasing numbers by excavations in both islands.
Such houses were used throughout the period considered in this chapter, and
their forms did not change significantly with the introduction of metals. Some
were defined by rings of post sockets, but a few examples were so small that
they may have played a specialised role. The surviving evidence usually consists
of hearths and circles or arcs of stake holes. There are also pairs of isolated post
holes which face southeast. They may be all that remains of the porches of
insubstantial buildings which, like so many others, had been aligned on the
sunrise. All these structures are accompanied by pits. The clearest indication
of a wider settlement pattern comes from the extreme north. In Shetland oval
stone-built houses dating from the late fourth millennium BC are found in small
groups associated with irregular field plots and clearance cairns. Chambered
tombs were located in between these groups of dwellings but were sometimes
connected to longer land boundaries. Again that pattern continued unchanged
into the Earlier Bronze Age (Whittle 1986).
It is difficult to consider more than a small sample of well-preserved settlements
because the structural evidence is so easily destroyed, but in areas with
an abundant supply of raw material another source of information is provided
by the results of field walking. This has been a particular feature of southern
England where flint is widely available. This method can define occupation
areas only in the broadest terms, but again there is a striking contrast with the
period discussed in Chapter Two. The occupation sites of the Earlier Neolithic
period have not been easy to define by field walking because little material
remains on the surface. That is probably because the contents of individual
settlements were deliberately buried when a site was abandoned (Healy 1987).
Although later settlements do include some pits, their positions are generally
marked by extensive spreads of artefacts. These cover very large areas, and it
is sometimes difficult to decide where one concentration ends and another
begins. The inhabitants seem to have made profligate use of the available flint
and generated large quantities of debitage during the course of occupation.
They also employed a much wider range of separate types than their predecessors,
who had favoured a small number of lightweight multipurpose tools
(Bradley 1987). That may suggest that certain areas were used more intensively
and perhaps for longer periods of time. The largest lithic scatters may have
been the sites of settlements as extensive as any of those found in Ireland.
Such observations have not been made in isolation, for in Britain, pollen
analysis reveals a continuous expansion in the areas that were occupied (Bradley
1978: 107–9). The same pattern is indicated by the distribution of artefacts, and it suggests that a wider range of environments were being used than at the
beginning of the fourth millennium BC. Some of the clearings may have been
quite short lived, but it seems as if activity increased in the uplands, beside
natural wetlands, and along the coast. Indeed, by 2000 BC. the occupied
area shows a stronger resemblance to the Mesolithic pattern than it does to
that during the Earlier Neolithic period. Recent work around two of the
major ceremonial centres in England indicates that they had been built in open
grassland (Bush 1988; French and Lewis 2005).
Similar evidence is provided by a series of rock carvings made between about
3000 and 2000 BC. They were a feature of Northern Britain and Ireland, and
their distribution is revealing (Bradley 1997: chapters 5–10; Beckensall 1999).
Although they can cluster around the ceremonial monuments of the Later
Neolithic, they may be as closely related to the wider pattern of settlement.
They can be found in the vicinity of lithic scatters, as well as on higher ground
which overlooks the occupied area. They are associated with passes providing
routes across the uplands; they are around springs and waterholes; and they
are found by sheltered harbours on the coast. In northern England, western
Scotland, and southwest Ireland, rock art may be associated with valleys leading
through the wider landscape. Although such places must have been visited
during the Earlier Neolithic period, this evidence suggests that larger areas
were being used than before. It also raises the possibility that travel between
different regions became increasingly important.
The previous chapter drew attention to different interpretations of the Earlier
Neolithic settlement pattern. Archaeologists working in lowland England took a quite different view from their colleagues in Ireland. That was hardly
surprising since the remains of rectilinear houses were regularly discovered in
some regions and were virtually absent in others. There were various possible
explanations, extending from the differential destruction of field evidence to
a genuine contrast in the character of Neolithic activity, or even the nature
of the rituals involving the houses of the dead. After 3000 BC, many of these
differences disappeared, and in both countries settlements may have had rather
more in common.
It is with the expansion of contract archaeology that insubstantial circular
buildings have come to light. It is not clear whether these structures were
adopted because settlement was organised in a different way. It is obvious that
crop cultivation continued, but little is known about methods of food production,
as nearly all the excavated evidence comes from ceremonial sites. It seems
possible that the subsistence economy did become more diverse, but there is
little reason to suppose that the increase in pastoralism documented by pollen
analysis in Ireland extended to Britain as well. In any case it seems unlikely that
Later Neolithic houses were organised according to purely practical considerations.
Just as the shapes of dwellings changed, so did those of monuments.
HOUSES, TOMBS, AND ARENAS: IRELAND, ORKNEY,
AND NORTHERN BRITAIN
Again the initial development began beyond these islands. Chapter Two considered
one of the main styles of mortuary monuments and its origins. It seems
to have been based on the model of the long house and had obvious prototypes
extending along the coastline of the Continent from Denmark and Poland, at
one extreme, to Brittany and Normandy, at the other. Such monuments might
be formed of several different materials, and some were constructed of earth
and timber, whilst others contained stone chambers.
Another tradition of mortuary monuments developed first along the western
coastline of Europe, and in this case practically all the structures were built of
stone (Scarre 1992). These are the megaliths known as passage tombs. They are
so called because they have two distinct elements. An entrance passage leads to a
chamber which is generally situated at the centre of the covering mound. Most
sites observe a clear distinction between these two components; the passage
restricts access to the interior and may be low and narrow; the chamber, on
the other hand, can be unexpectedly large and is sometimes spanned by a high
corbelled roof. A number of these sites were decorated by abstract pecked
designs (Shee Twohig 1981). Passage tombs in Britain and Ireland are usually
associated with circular cairns.
Although there are many variations among these monuments, two points
seem to be established. These structures have so many features in common that
they can be regarded as a single architectural tradition. It is no longer possible
to argue that different regional groups developed entirely independently of one another (Renfrew 1973b: 125–9). At the same time, it is clear that the oldest
of these monuments are found in Continental Europe. Among the areas in
which they developed was the Iberian peninsula. This is important for it lay
well outside the areas colonised by the first farmers (Arias 1999). Although it
may have experienced some settlement from other regions, there seems little
doubt that the native population remained largely unaffected. For that reason
it is likely that passage tombs were an indigenous invention, and throughout
Atlantic Europe there seem to be precedents for the building of small mortuary
monuments during the Mesolithic period (Scarre 1992).
There is some evidence for the chronology of passage tombs from the north
of Scotland. Those on the mainland are among the structures associated with
round cairns noted in Chapter Two, but they were considered together with
other forms of monument because some of them were incorporated into long
cairns during the earlier fourth millennium BC. The distribution of passage
tombs extends as far as the Northern Isles, and it was in Orkney that the
largest examples – those of the Maeshowe type – were built in the years around
3000 BC (C. Richards 2004b). That would suggest that such structures had an
extended history.
The Irish evidence is even more striking. It seems as if the first passage tombs
are those at Carrowmore in the west of the country which are associated with
radiocarbon dates extending back to the very beginning of the Neolithic period
(Burenhult 1980 and 1984; Sheridan 2003b). The monuments in question
consist of above-ground megalithic chambers, enclosed by a ring of boulders,
and may not have been covered by a cairn (Fig. 3.5). Comparable structures
are known from other areas around and near the coast of the northern half of Ireland and along the Atlantic fac¸ade of Britain, with structural echoes in
northwest France. In some cases there are signs of a rudimentary entrance
passage, and the tombs may have been open to the elements. The chambers at
Carrowmore were revisited at various times after their construction, especially
during the Iron Age.
The radiocarbon dates from Carrowmore extend between about 4000 and
3000 BC (Sheridan 2003b), but only one of the dated structures may have been covered by a cairn. This is Tomb 51, which is the focal point of the complex.
It is the largest monument and occupies the highest ground. It is also the only
one to have been decorated with pecked designs. It was not built until about
3550 BC. Most of the other dates for Irish passage tombs come from the Boyne
Valley, near to the east coast where much larger monuments were constructed
around 3000 BC (ApSimon 1986). The Mound of the Hostages at Tara was
built and used between 3350 and 2900 BC (O’Sullivan 2005). It is possible
that passage tombs were adopted through a gradual process of competitive
emulation: successive monuments increased in size and in the labour needed
to build them. That is the view of Alison Sheridan (1986), who interprets their
plans according to this hypothesis.
The sequence proposed by Sheridan would mean that during the fourth
millennium the earlier passage tombs were used in parallel with the court cairns
discussed in Chapter Two; the earliest examples were built as a result of contacts
along the Atlantic seaboard of Europe, whilst court tombs reflect a different
axis with its emphasis on areas further to the east. What seems important is
that by the later fourth millennium BC, passage tombs became the dominant
element in the Irish landscape, and that is why they are considered here.
They are mainly found in the north and centre of the country where their
distribution focuses on a zone extending between Drogheda to the east and
Sligo to the west. The main concentrations of these monuments form a series
of cemeteries, of which Carrowkeel, Loughcrew (Figs. 3.6 and 3.7), and the
concentration of monuments in the Boyne Valley (Fig. 3.8) are perhaps the
best known examples (G. Eogan 1986). In each case one massive monument is surrounded by a series of smaller tombs, but that configuration may be
repeated several times. Thus it applies to at least three different clusters of
chambered tombs at Loughcrew; there were distinct concentrations of monuments
at Carrowmore and on the nearby mountain of Knocknarea; and a
similar arrangement may once have existed with each of the large tombs in
the Boyne Valley: Newgrange, Knowth, and Dowth. Within any one cemetery
the tombs were carefully located in relation to one another: at Knowth
they surrounded the space that was eventually occupied by the largest monument,
and here the entrances of the satellite tombs were directed towards that
focal point (G. Eogan 1986). Such cemeteries could even incorporate tombs
of other kinds. On a still larger scale it seems as if groups of monuments in
different places might have been aligned on one another. Thus those on the
mountaintop at Carrowkeel acknowledge the position of a similar cemetery
on the summit of Knocknarea (Bergh 1995). Those sites are intervisible, yet
some of the passage tombs at Loughcrew seem to be orientated on similar sites
in the Boyne Valley, although they cannot be seen (Patrick 1975).
It has been difficult to work out a chronology for these monuments because
they were permeable structures. They could be entered and their contents
supplemented or removed over a considerable period of time. That raises a
problem, yet the contents of Irish passage graves were strikingly consistent,
and so were the practices associated with them. Most of the remains were
of adults whose bodies had been cremated. There were smaller numbers of
children, and a higher proportion was represented by unburnt bones. Some
of the cremations might be placed in stone basins inside the main chamber
or a series of side chambers, where other finds include a specialised group of
bone artefacts such as pendants and pins, stone balls, and a style of profusely
decorated pottery which takes its name from the cemetery at Carrowkeel. At
Fourknocks, the unburnt bones were mainly in the entrance passage (Cooney
2000: 103–12).
There is some evidence that Irish tombs were rebuilt or refurbished during
their period of use. Palle Eriksen (2004) suggests that the great mound at Newgrange
was modified on at least two occasions. Other monuments were treated
in a more drastic manner. Recent work on the main passage tomb at Knowth
has established that many of the orthostats were reused from another structure;
they had been taken down and moved, and much of the original decoration
was discovered on the backs of the stones (G. Eogan 1998). At Newgrange and
Knowth it is clear that the carved designs had been renewed and changed on
more than one occasion. George Eogan (1997) has identified several overlays
on some of the uprights at Knowth, leading him to suggest that the earliest
designs were incised linear motifs and that the pecked circles and spirals that
are often associated with the Boyne Valley may have been a later development.
In fact the decoration of such tombs could have had two distinct aspects.
Those people who were permitted to enter the chamber and passage would
have viewed a particular selection of carved motifs, but the very structure of the monument must have meant that they had to inspect them in sequence. At
the same time, those who were not allowed to go further than the perimeter
of the cairns would have been aware of a different selection of images. Once
again they would have seen them in a prescribed order as they moved round
the kerb.
These conventions may have changed over time, for most Irish passage tombs
do not have decorated kerbstones (Fig. 3.9). They are a special feature of
those in the Boyne Valley, and even here there may be evidence of a complex
sequence. It seems as if the decorated kerb of the principal mound at Knowth
was the last part of that structure to be built (G. Eogan 1997). The distribution
of quartz may follow a similar pattern, for Stefan Bergh has suggested that it
was originally deposited inside the chambers of Irish passage tombs. Only in
the latest of these sites did it embellish the exterior (Bergh 1995: 156). Again it
seems as if different features may have been directed towards different audiences.
That is especially important in the Boyne Valley where the raw materials used
around the exterior of the largest monuments seem to have been introduced to
the site from an enormous area extending from the Wicklow Mountains sixty
kilometres to the south, to the coastline about fifty kilometres to the north.
They seem to provide a model of the wider landscape, yet many were simply
pieces of distinctively coloured stone. This practice may show how far people
travelled to visit the tombs (Mitchell 1992).
Certain individuals may have been allowed to observe phenomena that were
denied to others. This is most apparent from the celestial alignments associated
with Irish tombs. Their full extent is uncertain because some of the monuments are poorly preserved or were restored without adequate record. It is accepted
that the entrance passage at Newgrange is aligned on the midwinter sunrise,
although the light actually travels down a specially constructed channel which
was built above the entrance of the tomb (O’Kelly 1982). The passages at
Knockroe were aligned on the midwinter sunrise and sunset, respectively (M.
O’Sullivan 2004), and at Knowth and Loughcrew they may have marked the
positions of the equinoxes (Brennan 1983). Two points are most important
here. All the convincing alignments are concerned with the movements of the
sun. There does not seem to have been the same interest in the moon. Second,
these effects could have been seen only from the interior of these monuments
and would have been lost on spectators who were excluded from the chamber.
Few people could have watched them, for if too many were present they would
have obscured the only source of light.
It seems possible that the use of passage tombs involved other experiences that
were restricted to a small number of participants. Jeremy Dronfield has argued
that some of the abstract designs associated with Irish megalithic art refer to the
visual images associated with altered states of consciousness (1995 and 1996).
A similar interpretation has been applied to Upper Paleolithic cave paintings
and is supported by research in neuropsychology (Lewis-Williams and Dowson
1993; Lewis-Williams and Pearce 2005). The problem is that such ideas explain
why certain designs assumed so much importance; but there was nothing to
prevent them being copied from one decorated surface to another until they lost
their original associations. There is also some experimental evidence that the
peculiar form of the passage tomb, with its narrow entrance passage and high
central chamber, creates unusual acoustic effects (Watson and Keating 1999).
They may even have helped to generate the mental and physical conditions
described by Dronfield.
It is quite possible that mounds covering Irish passage tombs were modelled
on the layout of the circular house, as they were in Atlantic Europe. That is
especially likely since the cemetery at Knowth overlay the remains of a series of
such dwellings. That is not the only link between them, for the circular ground
plan of the houses and the tombs is echoed in other media, in particular the
curvilinear decoration with which some structures were embellished. This
may have had an even wider significance as certain of those monuments were
laid out according to the arc traced by the sun across the sky (Bradley 1998b:
chapter 7).
The adoption of the largest passage tombs created restrictions of access that
may have been less apparent before. Space was very restricted within the chambers
and side chambers, and here there were rather different designs from those
on the exterior of the monuments. In the same way, only a small number of
people would have been able to appreciate the solar alignments that had been
built into the architecture of these buildings. More people would have been
excluded. Perhaps the monuments were associated with a smaller section of society than the court tombs that preceded them. It seems as if the knowledge
and ability necessary to plan these buildings may not have been generally available,
even though their actual construction would have required an enormous
workforce. Alison Sheridan (1986) is surely right to suggest that the construction
of passage tombswas a source of prestige. Different communities attempted
to outdo one another in mounting these ambitious projects. The same should
apply to the separate clusters of passage graves within the larger cemeteries.
There were few burials inside these monuments, and there is little to suggest
that particular individuals were provided with grave goods, although this
evidence may have been affected by later disturbance to the sites. Even so,
there are signs that social divisions were acknowledged in architectural form,
as the tombs within the individual cemeteries varied greatly in size. The passage
tombs in the Boyne Valley also had decorated kerbs that would have been
accessible to larger audiences than the chambers concealed behind them. That marks the beginning of a new development. The closed spaces of the chambers
eventually lost some of their importance, and the surrounding area was
occupied by new forms of architecture. At the same time the curvilinear decoration
which is such a feature of the exterior of these monuments was echoed
by rather similar designs on natural surfaces in the landscape (Bradley 1997:
62–5).
There are various clues to this development in Ireland. It seems possible
that the exterior of these tombs became more important with time. This is
most apparent on Knocknarea where Stefan Bergh (1995) has identified a series
of low platforms built around the flanks of several passage tombs (Fig. 3.10).
They could have accommodated a larger audience than the chambers. There
is another example at the decorated tomb of Knockroe (M. O’Sullivan 2004),
and something similar may have happened outside the kerbs at Newgrange and
Knowth where excavation has found many artefacts (O’Kelly 1982; Eogan and
Roche 1997). They have been interpreted as evidence of domestic activity, but
that seems unlikely. Much of this material accumulated after the construction
of these monuments and, in the case of Newgrange, after the quartz covering
of the mound had collapsed (O’Kelly, Cleary, and Lehane 1983).
There are important contrasts with the evidence from Scotland. Chapter
Two made the point that some of the long cairns found there might have been
based on a domestic prototype, albeit one in Continental Europe. That related
to the outward appearance of these structures. Perhaps more important was the
private aspect of the monuments. There is little relation between the appearance
and siting of the cairns and the layout of the chambers concealed within
them, and yet some of these evoked the internal organisation of the house.
That applied to the chambered tombs found on the mainland of northern
Scotland and to those in Orkney. Their characteristic ground plan resembled
the rectangular dwellings of the Earlier Neolithic period, which were arranged
as a series of ‘rooms’ on either side of a central passage.
Most of these monuments were fairly simple, and their chambers were generally
divided into three sections, but in Orkney this principle seems to have
assumed a life of its own. Here the most complex of these structures was Midhowe
which was separated into many more compartments, which is why they
have been described as ‘stalled cairns’ (Davidson and Henshall 1989). There
are some variations, but for the most part the chamber was entered at one
end of the building. This was not a major feature of these sites, but again its
position follows the organisation of the excavated dwellings. Both the houses
and tombs are associated with decorated Unstan bowls as well as plainware.
It seems as if the chambers contained the remains of a considerable number
of people, although individual body parts were later removed or rearranged
(C. Richards 1988).
It seems as if new kinds of monument emerged in the latter area towards the
end of the fourth millennium BC. These were passage tombs with structural marks the beginning of a new development. The closed spaces of the chambers
eventually lost some of their importance, and the surrounding area was
occupied by new forms of architecture. At the same time the curvilinear decoration
which is such a feature of the exterior of these monuments was echoed
by rather similar designs on natural surfaces in the landscape (Bradley 1997:
62–5).
There are various clues to this development in Ireland. It seems possible
that the exterior of these tombs became more important with time. This is
most apparent on Knocknarea where Stefan Bergh (1995) has identified a series
of low platforms built around the flanks of several passage tombs (Fig. 3.10).
They could have accommodated a larger audience than the chambers. There
is another example at the decorated tomb of Knockroe (M. O’Sullivan 2004),
and something similar may have happened outside the kerbs at Newgrange and
Knowth where excavation has found many artefacts (O’Kelly 1982; Eogan and
Roche 1997). They have been interpreted as evidence of domestic activity, but
that seems unlikely. Much of this material accumulated after the construction
of these monuments and, in the case of Newgrange, after the quartz covering
of the mound had collapsed (O’Kelly, Cleary, and Lehane 1983).
There are important contrasts with the evidence from Scotland. Chapter
Two made the point that some of the long cairns found there might have been
based on a domestic prototype, albeit one in Continental Europe. That related
to the outward appearance of these structures. Perhaps more important was the
private aspect of the monuments. There is little relation between the appearance
and siting of the cairns and the layout of the chambers concealed within
them, and yet some of these evoked the internal organisation of the house.
That applied to the chambered tombs found on the mainland of northern
Scotland and to those in Orkney. Their characteristic ground plan resembled
the rectangular dwellings of the Earlier Neolithic period, which were arranged
as a series of ‘rooms’ on either side of a central passage.
Most of these monuments were fairly simple, and their chambers were generally
divided into three sections, but in Orkney this principle seems to have
assumed a life of its own. Here the most complex of these structures was Midhowe
which was separated into many more compartments, which is why they
have been described as ‘stalled cairns’ (Davidson and Henshall 1989). There
are some variations, but for the most part the chamber was entered at one
end of the building. This was not a major feature of these sites, but again its
position follows the organisation of the excavated dwellings. Both the houses
and tombs are associated with decorated Unstan bowls as well as plainware.
It seems as if the chambers contained the remains of a considerable number
of people, although individual body parts were later removed or rearranged
(C. Richards 1988).
It seems as if new kinds of monument emerged in the latter area towards the
end of the fourth millennium BC. These were passage tombs with structural features in common with those in Ireland, yet they were also transformations
of the domestic dwellings of the same period. On one level the houses of
Later Neolithic Orkney form part of the general tradition of circular dwellings
considered at the beginning of this chapter. On another, they are quite idiosyncratic.
This may be because they are exceptionally well preserved, but they also
have a number of unusual features which are echoed in monumental architecture.
Among these are their entrance passages, the presence of intramural
compartments and the use of decorated stones (C. Richards 1991; A. Shepherd
2000). They changed their internal organisation over time, and it is the earlier
of these buildings that have the closest connection with passage tombs.
The first of these houses have a roughly cruciform interior and a circular
external wall, so that they combine elements of older and newer ways of
organising space. The circular structures have one entrance facing a central
hearth. There may be recesses set into the thickness of the walls, but the main
focal point was at the back of the dwelling (Fig. 3.11). Such buildings could
be grouped together into small villages. At Barnhouse, the separate dwellings
seem to have formed two concentric rings around an open area (C. Richards
2004b: chapter 3), whilst those at Skara Brae were connected by a series of low
passageways. In this case important thresholds were marked by incised motifs
(A. Shepherd 2000). Similar designs are also found on the portable artefacts
associated with these sites, including the decorated pottery known as Grooved
Ware (Saville 1994).
Many of these elements are represented in the passage tombs of the same
period. The main difference between them and their predecessors is that they
were increasingly subdivided. Whereas the earlier monuments had a simple
linear ground plan approached by a short entrance passage, now the chamber
had a more complex layout. It utilised some of the same elements as the
domestic dwellings, but it also exaggerated them (Figs. 3.11 and 3.12). Someone
entering even the most complex of the stalled cairns would still have moved
along the central axis between the successive compartments. In the Maeshowe
tombs, however, the main chamber was further from the exterior, and each
of the side chambers had to be accessed separately. Indeed, there are cases in
which the side chambers led to still further cells, each with a narrow passage
of its own. These subdivisions are similar to the recesses that were built into
the houses at Skara Brae. The importance of the thresholds inside these tombs
is emphasised by panels of incised decoration very like those in the dwellings
(Bradley et al. 2001). This comparison is important, for such images resemble
the earliest motifs found in the Boyne Valley (G. Eogan 1997), but they are
different from the pecked designs in the Irish tombs which may have referred
to altered states of consciousness.
How were these monuments used? Excavation at Quanterness suggests that
the chambers and side chambers contained large amounts of unburnt human
bone. The excavator, Colin Renfrew, argues that these monuments housed the bodies of entire communities (1979: chapter 12). Many of these corpses had lost
their flesh elsewhere. Colin Richards, however, suggests that the Maeshowe
tombs also included some of the body parts that had been removed from
earlier stalled cairns (Richards 1988). This is suggested by radiocarbon dates for individual bones, but it is difficult to interpret this evidence when so many
deposits had been disturbed by animals (A. Ritchie 2004).
Maeshowe tombs are not only elaborations of the basic plan of a dwelling,
they are located close to the living sites themselves. In some cases these tombs
seem to be paired with individual settlements. This applies to the relationship
between the passage grave on Cuween Hill and the excavated houses at Stonehall,
but the same pattern connects the passage grave on Wideford Hill to a
nearby living site, and the famous tomb at Quanterness to another settlement
at Crossiecrown (Colin Richards, pers. comm.). In other cases the monuments
may have been built over the sites of older dwellings. This certainly happened
at the Howe (Ballin-Smith 1994: chapter 2), and the same relationship has been
postulated at Maeshowe (C. Richards 2004b: chapter 9). There may also have
been a link between that tomb and the Later Neolithic village at Barnhouse.
The settlement included a stone decorated with a complex pattern. Fieldwork
in 1999 showed that a piece with exactly the same motif had been built into
the foundations of the chamber at Maeshowe (Bradley et al. 2001).
In another way the sequence in Orkney is similar to that in Ireland. Niall
Sharples (1985) has made the point that the earlier Orcadian tombswere entirely
focused on the interior space: that is where the human remains are found, and
there are few artefacts anywhere else. The later tombs have a different character.
Whilst they include even more bones, there are further deposits outside them.
In a few instances it seems as if special structures were created for the purpose.
Maeshowe is particularly relevant here, for the chambered tomb is built on
an earthwork platform enclosed by a ditch and possibly a wall. This seems to
have been contemporary with the tomb (C. Richards 2004b: chapter 9). At
two other Orkney monuments, Quoyness and Taversoe Tuick, a platform was
built around the perimeter of an existing passage tomb, restricting access to
the interior (Davidson and Henshall 1989: 154–8 and 160–3). The platform at
Quoyness included pottery and animal bones. At Isbister and Midhowe walled
enclosures may have been built on to the original construction (ibid., 125–30
and 146–8) and at Sharples’s own site at Pierowall a platform was constructed
against the collapsed perimeter of the cairn (Sharples 1984). In each case it
is as if the monument was turned inside out so that the exterior assumed
more importance than the interior. That is similar to the sequence that Audrey
Henshall (2004) describes on the Scottish mainland and in theWestern Isles. It
also recalls the evidence from Ireland where platforms were constructed outside
the perimeter of the tombs on Knocknarea and where the edge of the ruined
mound at Newgrange became the focus for deposits of cultural material.
Maeshowe contributes to this discussion in other ways. The tomb is most
unusual because the corners of the burial chamber are defined by four tall pillars,
whilst another four frame the entrance passage. Although they are undoubtedly
impressive, they do not contribute to its stability. Since they had been quarried
some distance away, why were they there at all? A possible answer is provided by a substantial socket cut into the earthwork platform. Is it possible that these
stones had been taken from another monument on the same site (C. Richards
2004b: 242–4)? If so, they could have been reused, like the decorated orthostats
incorporated in the passage at Knowth.
A possible model for such a setting exists not far away at the Stones of
Stenness, a ring of enormous uprights encircled by a ditch (Fig. 3.13; G. Ritchie
1976). Nearby there are two more monuments: an even larger stone circle, the
Ring of Brodgar, which is inside a similar earthwork (G. Ritchie 1998); and
the Ring of Bookan which may be the site of a second passage tomb set within
an enclosure (Henshall 1963: 232–3). Most of these sites are undated, but it is
generally agreed that Maeshowe must have assumed its present form around
3000 BC. The base of the ditch surrounding the Stones of Stenness is associated
with radiocarbon dates of 3100–3000 BC.
At the same time the plan of the chambered tomb at Maeshowe resembles
that of the latest structure in the nearby settlement of Barnhouse, which is
of similar age. This building was the largest on the site, but again its exterior
was supplemented by a platform (C. Richards 2004b: chapter 5). One of the
characteristics of this building is that it contained two massive stone-lined
hearths. There is a similar hearth in the centre of the Stones of Stenness, and
it is possible that there had been a structure of the same kind there. Both the
sites exhibit virtually the same organisation of space.
Colin Richards (1996b) argues that all three of these constructions – the
house, the tomb, and the stone circle – refer to one another at the same time as
this complex provides a microcosm of the surrounding landscape. There are also
contrasts between these sites. Whatever its origins, the preserved structure at
Maeshowe is a tomb. It may be augmented by an earthwork platform and
enclosed by a ditch, but it lacks the impressive facade of a monument like
Newgrange. The monument is closed to the wider world. Both are orientated
on the midwinter sun, but that could have been observed only from the chamber.
A stone circle, on the other hand, is a largely permeable structure, and
Richards himself emphasises the visual relationships between the monoliths
and more distant areas of high ground. It was essentially an open arena of the
kind that is known as a henge and could certainly have accommodated a large
number of people. The nearby stone circle of the Ring of Brodgar provides
some indication of the wider significance of such monuments, for recent work
suggests that, like Newgrange, it had been built out of raw material introduced
from several different sources (C. Richards 2004c).
Some of the same issues arise with the Irish monuments. Newgrange presents
problems of its own. The great passage tomb is enclosed by a series of monoliths.
They are not concentric with the kerb but emphasise the position of the
entrance and the section of the perimeter which had been enhanced by blocks
of quartz. The original excavators took the view that the chambered tomb
and the stone setting were contemporary with one another (O’Kelly 1982),
but a more recent investigation suggests that the monoliths were not erected until the mound had collapsed (Sweetman 1985). The evidence is ambiguous,
but the published section drawings suggest that the stone sockets might have
been dug from a ground surface associated with Grooved Ware and Beakers.
Stone circles are associated with chambered tombs at other sites in Ireland,
and the Mound of the Hostages is surrounded by a series of ‘fire pits’ dated
between 2030 and 1690 BC. These might have been the sockets for upright
stones that have since been removed (O’Sullivan 2005: 228–33). In northern
Scotland, monuments of superficially similar character were enclosed by stone
circles during the Earlier Bronze Age (Bradley 2005a: 100–6).
The stone circle is not the only structure to be built outside the principal
monument at Newgrange, for adjacent to the great mound there was a massive
oval enclosure made up of concentric rings of pits delimited by an outer
circuit of posts (Fig. 3.14). This occupied virtually the same amount of space
as the passage tomb and was associated with fragments of burnt animal bone
(Sweetman 1985). Its chronological relationship to the stone circle is unclear,
but this structure was built long after the tomb. It dates from the later third millennium
BC. Like the Stones of Stenness, this monument has been described
as a henge.
The termalso describes several circular enclosures in the Boyne Valley which
had been built by excavating soil from the interior and using it to construct
a bank (G. Stout 1991). These earthworks are usually found close to passage
tombs, and an example not far from Newgrange contained one of these
monuments. Another of these sites, at Monknewtown, included a cremation
associated with Carrowkeel Ware (Sweetman 1976). The most informative
example is the Giant’s Ring at Ballynahatty near Belfast (Fig. 3.14), which may have been laid out around the position of an existing passage tomb (Hartwell
1998). Such earthworks are usually compared with Later Neolithic monuments
in Britain, and the link with megaliths certainly suggests that their histories
might have overlapped, but a note of caution is necessary here as Helen Roche
(2004) has demonstrated that the only extensively excavated monument related
to this kind, The Grange ‘stone circle’ at Lough Gur, was built during the Later
Bronze Age.
It is usual to distinguish between ‘Irish’ henges, which are embanked enclosures
(Condit and Simpson 1998), and a ‘British’ formin which the material for
the bank came from an internal ditch (Wainwright 1969). In fact a few earthworks
of the latter type have been identified in Ireland, although there is room
for some uncertainty as very similar monuments date from the Later Bronze
Age and Iron Age; there may have been no connection between them. On the
other hand, one of these enclosures at Dun Ruadh certainly originated in the
Later Neolithic or Earlier Bronze Age periods (Simpson,Weir, andWilkinson
1992), whilst another on the west coast has a radiocarbon date in the late third
millennium BC (MacDonagh 2005: 12–13). The putative henge of Longstone
Rath has produced pottery of similar date although the excavation has never
been published (Sheridan 2004a: 29–30).
The Irish henge monuments are defined by their characteristic earthworks,
but enclosures of other kinds were built of wood. One was immediately outside
the principal passage tomb at Newgrange, but the details of its structure
have still to be defined. Another enclosure, apparently defined by two parallel
palisades, has been identified by geophysical survey on the Hill of Tara
(Fenwick and Newman 2002). Again it includes a passage tomb within its
area. A better preserved monument is found not far from the Giant’s Ring
and, like its counterparts at Newgrange and Tara, it seems to have formed
only one component of a more extensive monument complex, including
a series of ‘megalithic cists’ (Hartwell 1998). These were miniature passage
tombs which were built inside shallow pits, and, like their full-size prototypes,
they were associated with Carrowkeel Ware. Again the principal enclosure
was defined by two parallel circuits of posts (Fig. 3.14). They enclosed the
position of a timber circle with a massive square setting at its centre. The
wooden structures at Ballynahatty are associated with Later Neolithic Grooved
Ware and with a quantity of animal bones which are perhaps the remains of
feasts. The enclosure and the timber circle were both burnt to the ground.
Afterwards the positions of the posts in the circular building were marked by
cairns.
The evidence from Ballynahatty is particularly relevant to the argument, as
a similar circle was constructed outside the eastern entrance to the principal
tomb at Knowth (Eogan and Roche 1997). Again this was associated with
placed deposits of cultural material, including sherds of Grooved Ware. They
should be contemporary with at least some of the features in front of the entrance to Newgrange, and not far away from them were the sockets of
another ring of posts (Sweetman 1987). Further examples, again associated with
Grooved Ware, have been found in the course of commercial excavations in
Ireland.
ACROSS THE SEA
If the archaeological sequence in Ireland took a similar form to that in northern
Scotland, were there any direct connections between those processes?
Alison Sheridan (2004a) has recently discussed this question. She emphasises
two important points. The megalithic tombs on either side of the Irish Sea
seem to have developed through a process of competitive emulation among
different communities until these monuments grew to enormous proportions.
That is as true of the stalled cairns of Orkney as it is of Maeshowe, and the
same argument applies to the largest passage tombs in Ireland and, perhaps,
to the largest and most elaborate court tombs as well. Eventually that process
may have involved the emulation of exotic artefacts, practices, and beliefs. It is
illustrated by the latest monuments in Orkney, whose extraordinary architecture
drew on the Boyne passage tombs as one source of inspiration (G. Eogan
1992). It may also explain the adoption of the first henges in Britain and
Ireland.
At the same time, both these sequences shared another feature which could
have developed independently. This was a gradual change in the audiences
who used the monuments. Space would always have been restricted inside the
passage tomb, but it seems as if the areas outside these buildings may have
assumed a growing significance. The building of external platforms, timber
settings, or earthwork enclosures continued a process that was already well
advanced, and as these structures came to overshadow the original importance
of the tombs, it seems likely that social relations were transformed. There may
still have been the same emphasis on the supernatural, but the significance of
mortuary rites might have been reduced.
One connection is especially important. Grooved Ware apparently originated
in Orkney, where it may have developed from the kind of pottery
associated with Unstan bowls (Hunter and MacSween 1991). It could have
been introduced to Ireland some time before Carrowkeel Ware went out of
use (Sheridan 2004a). The earliest Grooved Ware in Orkney had incised decoration,
and similar material has now been identified on the other side of the
Irish Sea. The same kind of decoration is found in other media and can be
used to illustrate a wider network of contacts at the turn of the fourth and
third millennia BC. Some of the main divisions of space inside the Maeshowe
tombs were indicated by incised motifs (Bradley et al. 2001). They are entirely
absent from the earlier tombs in Orkney, but have an exact counterpart in the first phase of decoration in the passage graves of the Boyne Valley (Fig. 3.15; G.
Eogan 1997). Similar designs were added to stone artefacts in Later Neolithic
Orkney (Saville 1994), and again these have their counterparts in other regions.
The most obvious are several stone plaques found on the Isle of Man (Burrow
1997: fig. 6.3). There is a similar object from an axe quarry in north Wales
(Warren 1921) and a decorated antler from a small henge monument in East
Anglia (G. Simpson 1981: 44–7). In fact such connections extend to a recently
discovered panel of rock art on Fylingdales Moor in northeast England (Brown
and Chappell 2005: 64–9).
There are other connections between the monuments on either side of the
water. It has long been argued that the characteristic form of Maeshowe makes
a direct reference to the passage tombs of the Boyne Valley (G. Eogan 1992). This would certainly explain why the history of Orkney changed its course
so drastically. There are other monuments in Britain which are closely related
to those in Ireland. The most obvious are the two decorated passage tombs
on Anglesey (Hemp 1930; Lynch 1967), but further links have been suggested
with individual sites in the Outer Hebrides and the Isle of Man (Henley 2004;
Darvill 2000). Similar connections may extend to megalithic cists of the kind
found at Ballynahatty (Hartwell 1998), for they bear some resemblance to
rather larger structures found in Orkney (Dalland 1999).
It is easy to make too much of the evidence of megalithic art, for whilst it is a
common feature of the later tombs in Ireland, in Britain it is rare. The incised
decoration of the Maeshowe tombs may have close affinities with the early
designs found in the Boyne Valley, but the pecked motifs that are also known
in Orkney have their closest parallels along the west coast of Britain. They
extend from an early stone circle, Temple Wood, in western Scotland (Scott
1989), to the passage tomb of Barclodiad Y Gawres in north Wales (Lynch
1967). There are other links between distant areas. A number of chambered
tombs in Scotland and Ireland had platforms built outside them during, or even
after, their main phase of use. Similarly, Newgrange, Maeshowe, and Knockroe
are all aligned on the midwinter sun.
Henges and stone circles are also shared between regions on either side of the
sea, and sometimes their structures suggest a direct connection between British
and Irish monuments. Whatever the origin of the stone circle at Newgrange,
it seems possible that a monument of the same kind at Kiltierney in the north
of Ireland was used around 3000 BC (Sheridan 2004a: 30). That would be
consistent with the dates for the Stones of Stenness (G. Ritchie 1976), and
both these sites are associated with GroovedWare. At Llandegai on Anglesey, a
henge monument may be of similar age. This is associated with Peterborough
Ware and dates from about 3200–3100 BC. The enclosure is unusual because
its bank is placed inside the ditch; a second henge at Llandegai, built around
2700 BC, shares the same alignment, but in this case the earthwork adopts
the conventional form (Lynch and Musson 2001: 36–77). A number of other
monuments could have been built at the beginning of the third millennium
BC. A stone setting at Temple Wood in western Scotland (Scott 1989) is decorated
with the same design as two of the passage graves in Orkney. The large
stone circle at Long Meg and her Daughters in northwest England illustrates
the same relationship, for here an outlying monolith is embellished with the
characteristic repertoire of Irish megalithic art. It also establishes an alignment
on the midwinter sunset (Burl 1994).
A key site is Calanais on the island of Lewis in the Outer Hebrides (Henley
2005). Here a stone circle with unusually tall monoliths is approached by no
fewer than four avenues or alignments which are laid out at right angles to
one another and converge on the centre of the circle. It is not clear when
they were built, but inside that circle is a tiny passage tomb. Its entrance is orientated on one of the rows of uprights. These structures date from the early
third millennium BC and were associated with GroovedWare (Ashmore 1999).
They are of particular interest as the layout of the ring of monoliths resembles
that of a cruciform passage tomb, but in this case it has been opened out to
form a monumental enclosure. There are a few other sites where such tombs
are juxtaposed with stone circles, but again their chronology is not clear.
Whilst the stone circle at Calanais is unique, henges generally take standard
forms. There are some connections between the Later Neolithic monuments
on either side of the Irish Sea. The earthwork monuments in Ireland were generally
embanked circular enclosures. Two of these occur in northern England,
at either end of a land route crossing the Pennines (Topping 1992; Moloney
et al. 2003), and there may have been other examples in southwest Wales,
although this needs to be confirmed by excavation (G. Williams 1984; Darvill
and Wainwright 2003).
The timber circles come in a variety of sizes, from small post settings indistinguishable
from ordinary houses (Fig. 3.16) to more grandiose structures that were probably specialised buildings (Fig. 3.17; Gibson 2005). The smaller
could easily be roofed; this would have been difficult with the largest examples.
The important point is that these buildings form a continuum and that
structures of quite different sizes observe the same organisation of space. A
particularly striking pattern is exemplified by the circle at Ballynahatty which
encloses a central square of large posts (Hartwell 1998). That design is found
quite widely and extends from Ireland to southern England (Sheridan 2004a).
It occurs outside the passage grave at Knowth (Eogan and Roche 1997), whilst
the same arrangement was followed on the island of Arran off the Scottish coast
(Haggarty 1991) and at Durrington Walls on the Wessex chalk (Wainwright
and Longworth 1971: 41–4). All these sites were associated with Grooved
Ware. In some ways these structures raise the same problems as the halls discussed
in Chapter Two. Perhaps they were public buildings whose form was
modelled on that of the domestic dwelling. They come in a variety of sizes
just like enclosures and tombs, and it may be that, like those monuments,
the escalating scale of construction was one way in which different groups
of people competed in a process of conspicuous consumption. That term
seems especially appropriate since some of these places provide evidence of
feasts.
Finally, there is a more general relationship to observe between Later
Neolithic Ireland and the same period in Britain. For the most part local
communities drew on distant areas as a source of ideas which they employed in
idiosyncratic ways. In certain cases these geographical connections may have
focused on quite small regions at the expense of their wider hinterland. To
some extent that was true of Orkney, and it was certainly true in Cumbria
in northwest England and Kilmartin Glen on the west coast of Scotland. The
same idea might apply to Anglesey, with its decorated passage graves and the
early henge monument at Llandegai. All these places have a number of features
in common with Neolithic Ireland. Perhaps that is not surprising, for they were
on the coast and yet were generally cut off from inland areas. Cumbria and
Anglesey were isolated by mountains, Orkney and Kilmartin by water. The
Isle of Man, on the other hand, shares elements with Ireland and Scotland and
has distinctive features of its own, in particular a series of cremation cemeteries
(Burrow 1997: 191–2).
The evidence for such connections takes many forms: particular types of
field monuments, unusual portable artefacts, decorated tombs, and open-air
rock art. What matters is that these bear a stronger resemblance to features of
Irish archaeology than they do to the repertoire of the British Later Neolithic.
Perhaps the best example is Cumbria (Watson and Bradley in press). The
Langdale quarries provided some of the stone axes used in Ireland. One of the
major henges in northwest England, Mayburgh, lacks the customary ditch and
is an embanked enclosure of Irish type. As mentioned earlier, the large stone
circle, Long Meg and her Daughters, includes an isolated monolith decorated in the same style as the passage tombs of the Boyne Valley. The same is true
of a number of decorated outcrops and smaller monuments. It seems as if
communities living in this area developed stronger links with their neighbours
to the west than they did with communities across the Pennines.
Such connections may be acknowledged by two exceptional monuments:
the passage tombs of Maeshowe (Davidson and Henshall 1989: 142–6) and
Barclodiad Y Gawres (Lynch 1967). Both are particularly similar to monuments
in Ireland. That applies to their architecture and to their characteristic decoration. At the same time, they diverge from the predominantly eastern
orientation of such sites. Maeshowe faces southwest and is aligned on the midwinter
sunset. Barclodiad Y Gawres faces north-northwest and could not be
aligned on either the sun or the moon. What they do have in common is
that both are directed towards Ireland. It is impossible to be sure that this was
originally intended, but, given the evidence that Irish Neolithic cemeteries
were orientated on one another, the idea should be taken seriously.
THE LATER HISTORY OF HENGES
It may seem unusual to devote so much space of the archaeology of Ireland and
Scotland, when some of the most famous monuments of the Later Neolithic
are in southern Britain, but it seems impossible to understand them in terms
of local developments.
There is a problem which is not always acknowledged when studying these
monuments. In the absence of excavation, virtually all the attention has been
paid to their earthwork perimeters, with the result that the sites are classified
according to their size and the number of entrances. That procedure cannot
take account of any internal features. Moreover the timber circles and stone
circles that are sometimes inside them are also found in isolation, suggesting
that the earthwork perimeter was not an essential component of these sites.
For that reason it seems wise to begin this discussion with freestanding rings
of posts and stones. That has one advantage, for ditches can be cleaned or recut
and banks can be rebuilt. It may be difficult to find material associated with
their original construction. On the other hand, the wooden uprights were
sometimes charred when they were set in the ground. As a result, they provide
a reliable source of dating evidence, although individual posts could have been
replaced. Alex Gibson (2005) has studied their chronology in relation to the
scale on which these monuments were built. The earliest were of modest proportions;
they became much larger in the period around 2500 BC, when some
of them adopted a more complex layout, and during the second millennium
BC their size diminished. Much less is known about stone circles, but to a limited
extent they may have followed a similar pattern (Barnatt 1989; Burl 2000).
Some of the largest examples date from the mid to late third millennium BC,
and after that time they decreased in size. Since stone structures often replaced
those of timber, it is not surprising that these trends should overlap.
A growing number of post settings have been found in isolation. Others
have been located outside the earthworks of major monuments, as they have
at Ferrybridge in northeast England (Roberts 2005: 37–41 and 197–9). This
has important implications, for the distribution of Later Neolithic ceremonial
sites is essentially a distribution of earthworks, which sometimes occur together
in small groups. Certain regions have always stood out because such features
are rare. The results of developer-funded excavation raise the possibility that timber circles are more widely distributed. It was the practice of bounding
these structures by a bank and ditch that varied from one region to another.
Again a comparison with stone settings may be helpful. It is known that certain
of them replaced timber monuments, but many of the most impressive examples
are completely unenclosed. That cannot be explained by the problems of
constructing earthworks in places with an intractable bedrock, for the same
areas often contain cairns (Burl 2000: fig. 2).
Two features of these sites are especially revealing: their spatial organisation
and their development over time.
The timber circles were not necessarily isolated. More than one example
could be built on the same site, and some of them were approached by wooden
avenues, comprising two rows of posts; similar alignments have been found in
isolation in northeast England and are generally associated with Later Neolithic
ceramics (Tavener 1996). The Northern Circle inside the earthwork henge of
Durrington Walls illustrates this point. It takes the characteristic form of a
square within a circle, and is approached by two rows of close-set timbers
which provide access through a wooden screen just outside the monument
(Wainwright and Longworth 1971: fig. 17). A similar phenomenon is recorded
with stone circles. For example, two of those at Stanton Drew in southwest
England are approached by short alignments of paired stones. Geophysical
survey suggests that they were built on the sites of timber buildings (David
et al. 2004).
Although the wooden monuments are described as circles, that was not
always the case as some are ovals or ellipses. This is important as it creates a
long axis. That orientation could also be emphasised by a formal entrance. A
good example is the timber setting at Woodhenge where a combination of
these techniques meant that the site was orientated on the midsummer sunrise
(Cunnington 1929). It was possible to vary the lengths of the upright posts
so that the structure was graded in height from one side to the other. This is
difficult to demonstrate when nothing remains but post holes, yet there should
have been a consistent relationship between the depths to which the uprights
were buried and the extent to which they stood above the ground. On that basis
Roger Mercer has argued that a timber circle at Balfarg in eastern Scotland was
higher towards the west (1981b). That is not improbable as many stone circles
illustrate the same phenomenon. The well-preserved timber setting known
as Seahenge in eastern England provides evidence of a similar arrangement
around 2000 BC. Here the tallest uprights were towards the southwest, facing
the setting sun (Brennand and Taylor 2003). It is not clear how common this
alignment might have been, but it was first used at chambered tombs like
Maeshowe.
The internal organisation of timber circles varies from single rings of posts
to more complex patterns (Gibson 2005). The largest monuments involve
the erection of as many as six concentric rings, and the setting revealed by geophysical survey beneath a stone circle at Stanton Drew may have had nine
(David et al. 2004). A visitor to such monuments might have expected to pass
through the entrance and to proceed to the centre of the circle, but this was not
always possible as the posts may have been linked by screens. The distribution
of the artefacts deposited insideWoodhenge certainly suggests that people had
to move around the perimeter of the building, approaching the innermost
space along a curving path that has been likened to a maze ( J. Harding 2003:
74–81). Another monument, identified by air photography at Catholme in the
English midlands, was associated with radial lines of posts, extending out from
the centre like the spokes of a wheel (A. Harding 1987: 268–71).
Different parts of these buildings were associated with different kinds of
archaeological material. For example, at Durrington Walls the distributions of
pottery and flint artefacts in the Southern Circle avoided one another (Richards
and Thomas 1984). Sometimes it seems as if such deposits could have been
placed there in sequence. AtWoodhenge, the groups of wild animal bones were
closer to the entrance than those of domesticates. A similar contrast has been
recognised between deposits of pig bones and those of cattle. Such conventions
also extended to portable artefacts, so that stone axes were associated with the
outside of the monument. Fragments of these tools occurred in the outer part
of the building, but, apart from one example made of chalk, none was found
towards its centre (Bradley 2000b: 124–7).
It is not clear how these deposits were related to the individual posts. Had
they been placed around them whilst they were still intact, and were they
preserved in the hollows left behind when the timbers decayed? Were they
placed in pits when those uprights had already rotted, like the cairns built over
the postholes in the circle at Ballynahatty? Recent excavation at the Sanctuary
at Avebury even suggests that on some sites the individual timberswere replaced
soon after they were erected, implying that the act of setting them up was more
significant than the overall design (Pitts 2001). The same is true at Durrington
Walls (Parker Pearson et al 2006: 235).
It is commonly supposed that timber circles were replaced by settings of
stones, but there may have been composite monuments in which both materials
were used together (Fig. 3.17). For example, the post setting at Mount Pleasant
was structured around four corridors laid out at right angles to one another. To
the north and south, the passages were of even width; the other two became
narrower as they approached the centre of the circle. Although the excavator
suggested that this building was replaced by a series of standing stones, it
seems more likely that these features were contemporary with one another
and that the monoliths functioned as barriers screening the centre from view
(Pollard 1992). Something similar may have happened in the Sanctuary, close
to Avebury, where the innermost part of a timber circle was hidden by placing
upright stones in between adjacent pairs of posts. This was executed so precisely
that it seems likely that those posts were erected at the same time. The stone
circle closed off the innermost part of the building. At other sites the replacement of timber monuments in stone might have
fossilised their characteristic form and protected them from decay. This seems
to have happened throughout most parts of England and Scotland (Fig. 3.18;
Wainwright 1969); it is suggested by the results of geophysical survey in southwestWales
(Darvill andWainwright 2003); and it is one way of understanding
the sequence of structures outside Newgrange (Sweetman 1985). The successive
monuments occupy the same area of ground, and sometimes they may be
exactly the same size as one another. The stone circles also take up details that
were present in the wooden buildings. They respect their orientations, and
even the timber avenues may have been replaced by monoliths. Quite often
the circles are graded in height towards the south or west. The connections
with wooden monuments may be emphasised in other ways, as several of the
uprights at Avebury had been used for grinding stone axes (Gillings and Pollard
1999).
Just as timber circles might be replaced in a more durable medium, the rarer
wooden avenues had their equivalents in stone (Fig. 3.19). In this case there
is no evidence that post settings were removed, but the routes leading to a
number of major henges are certainly lined by monoliths. The best known
example is the West Kennet Avenue at Avebury which links the henge to
the Sanctuary, but recent fieldwork shows that another avenue extended from
the western entrance of the monument and cut across the remains of an older
enclosure (Pollard and Reynolds 2002: 100–5). The same could have happened
at Broomend of Crichie in northeast Scotland ( J. Ritchie 1920), and there may have been an alignment of standing stones linking the Stones of Stenness and
the Ring of Brodgar on Orkney (C. Richards 2004b: chapter 8). Another large
stone circle which once possessed a long stone avenue was at Shap in Cumbria
(Clare 1978). These are not the only examples, yet little is known about how
these structures were used. If one avenue at Avebury communicated between
two different stone circles, its counterpart seems to have been decommissioned,
and at one stage it was apparently closed by a row of standing stones (Pollard
and Reynolds 2002: fig. 39).
Some of the monuments are enormous and their construction would have
involved much more labour than the timber settings that preceded them. That
is particularly true when the source or sources of the stone were some distance
away. It was obviously important to use materials taken from particular places,
and the Ring of Brodgar may have been built out of materials from different
parts of Orkney (C. Richards 2004c). That may have been the case at Stanton
Drew (Lloyd Morgan 1887), and it even seems possible that stones were taken
from older buildings in the way that had happened at Knowth (G. Eogan 1998).
Parker Pearson and Ramilsonina (1998) draw attention to the physical properties
of the raw materials used on these sites. Wood is an organic substance
and comes from living trees. Stone, on the other hand, stays the same for ever.
They compare this evidence with contemporary practice in Madagascar, where
wood is associated with the living, and stone monuments with the dead. Commentators
have taken exception to this use of ethnographic analogy (Barrett
and Fewster 1998), but that interpretation is consistent with the archaeological
evidence from Britain. The timber structures found in henges are often associated
with a rich material culture, including deposits of pig bones which may
result from feasting (Albarella and Serjeantson 2002). The stone monuments,
on the other hand, have few associated artefacts. They are found with human
remains and with little else. It seems as though places which had once been
connected with the living were now devoted to the dead.
A similar development can be recognised at another group of monuments.
Not all the sites described as henges were above-ground structures. Another
way of defining a circular space was by a ring of pits. The smallest were the
same size as the simpler timber circles, whilst the largest was at Maumbury
Rings in southern England. This was an enclosure with an external bank, but
instead of the usual post settings a ring of deep shafts was excavated. There were
approximately forty-five of them, each dug about ten metres into the chalk.
It is clear that the shafts had been filled in deliberately. It is known that the
process took place in stages, and that each episode had been accompanied by
the deposition of artefacts and bones (Bradley 1975: 34–6). A similar sequence
was identified at a smaller monument forty kilometres away on Wyke Down,
where the earliest deposits included carved chalk objects and antler. There
followed a series of flint implements and animal bones, whilst the latest material
also included sherds of Grooved Ware, human cremations, and part of an
unburnt skull (Barrett, Bradley, and Green 1991: 92–106). A similar pattern was identified at Maumbury Rings, where carved chalk objects and deer skulls
were in the middle filling of the shafts and finds of human bones were towards
the surface (Bradley 1975: 18–22).
Human bone is so common in pit circles that their presence has been misunderstood.
This was particularly true during excavations at Dorchester on
Thames in southern England, where these monuments were interpreted as
cremation cemeteries (Kinnes 1979). This approach was based on a misunderstanding.
It is clear that the deposits of human remains were secondary to the
monuments at which they were found, so those structures cannot have been
built for the disposal of the dead. The point is particularly obvious at Dorchester
on Thames, where a series of cremations had been placed in the hollows left
after posts had rotted or had been removed from the ground. Rather than associating
the remains of the dead with the creation of these structures, they may have registered a change in how such places were used. They may have become
more closely associated with the dead. Similar evidence has been recognised
on other sites, including the upper levels of the Neolithic round barrow at
Duggleby Howe, but in this case the monument was associated with burials
from the outset (Kinnes et al. 1983).
It has never been easy to account for the earthworks of henges, especially
since it would have needed so many people to build them. To construct the
perimeter of Avebury would have taken about a million worker hours (Startin
and Bradley 1981). They might have acted like amphitheatres, allowing a large
number of people to watch the events taking place inside them. The banks
would provide an excellent view, and yet on the British sites theywere separated
from the interior by a ditch. The basic format was the same whatever the
extent of the enclosure and whether or not there were any structures within
it. In principle, this would suggest that large numbers of people participated
in the use of these places, but the architecture of the monuments might also
have emphasised the distinction between those who were allowed into the
enclosures and the audience on the bank.
Another way of looking at this evidence is to suggest that the banks of henge
monuments were intended as barriers or screens. They actually concealed
events from the greater part of the population. That may be why they were
sometimes higher at the entrance. Until recently that idea would not have
found much support, but it is becoming apparent that similar structures were
enclosed by massive palisades (Gibson 1998). They are relevant to this discussion
because they would have hidden the interior of these monuments from view.
They could not have accommodated any spectators, and the people inside these
enclosures would have been cut off from the wider community.
There are a few sites at which it is possible to compare the timber enclosures
with earthwork monuments (Fig. 3.20). At Mount Pleasant a large construction
was built inside an already existing earthwork (Wainwright 1979). Not far away,
at Greyhound Yard in Dorchester, there was a freestanding palisaded enclosure
of similar proportions (Woodward, Davies, and Graham 1993: chapter 2). It
seems unlikely that the banks were intended to provide vantage points and that
timber structures acted as screens, for their roles seem to be interchangeable.
Durrington Walls and Greyhound Yard enclose dry valleys which might be
regarded as natural amphitheatres, but one was ringed by an earthwork and the
other by a wooden stockade (ibid; Wainwright and Longworth 1971). Mount
Pleasant occupies a domed hilltop which would not have allowed as much visibility
over the interior, yet here an earthwork was replaced by a timber enclosure.
All those sites are in central southern England, but at Blackshouse Burn
in southern Scotland a bank was constructed on the line of the two palisades.
Here there was no evidence of an internal ditch (Lelong and Pollard 1998).
Other palisades were equally extensive, but in this case they were not accompanied
by any earthworks. Some of them contained the sites of pit or post circles. Such settings have been identified by excavation at West Kennet in
Wessex (Whittle 1997) and at Dunragit ( J. Thomas 2004b) and Meldon Bridge
(Speak and Burgess 1999) in southern Scotland. They have also been identified
by geophysical survey and air photography at other sites in northern and
western Britain. At Dunragit, a ring of enormous uprights was surrounded by
a palisade associated with Grooved Ware.
It is difficult to reconstruct these monuments from subsoil features, but two
characteristics stand out. The palisades were very high and their entrances were
exceptionally narrow. These monuments were generally between five and ten
hectares in extent, but the largest, at Hindwell in Wales, enclosed thirty-six
hectares (Gibson 1999). They seem to have been constructed at various times
between about 3000 and 2000 BC and may be arranged in sequence according
to the few radiocarbon dates that are available (Gibson 1998). The earlier
monuments were defined by wide-spaced posts with a less substantial fence in
between them. In time, the intervals between the major uprights were reduced
until, finally, contiguous posts were bedded in a deep foundation trench. It is
clear from this that increasing amounts of wood were employed in constructing
these barriers. There are also hints that the scale of these projects changed in the
way that happened with timber circles. The first enclosures were the smallest.
They increased in size in the mid-third millennium BC, and the latest examples
were conceived on a less ambitious scale.
The individual posts seem to have been between five and nine metres high.
Thus they would have hidden whatever was happening behind them. The
entrances add to this effect, for they are very narrow, and in certain cases
people would have passed through them in single file. At Dunragit, Forteviot,
Walton, and Meldon Bridge they are approached by short avenues of paired
uprights not unlike that at Durrington Walls. They do not share a common
orientation (Speak and Burgess 1999: 24–6 and 110–14).
Having considered the different components of a series of Later Neolithic
monuments, how were they related to one another? It is clear that these sites
saw a complex sequence of activity, and Alex Gibson (2004a) has argued that in
certain cases the earthwork perimeter was the last feature to be built. It might
have been impossible to erect the circles inside it because of the presence of the
ditch; sometimes it is poorly aligned with the internal structures; and at Milfield
North a post circle must have been buried by the bank. He suggests that this
sequence was followed at a number of monuments, including Woodhenge,
Arminghall, Arbor Low, Balfarg, and Cairnpapple. His argument is supported
by work on the Scottish henge at North Mains, for here a timber circle was
associated with radiocarbon dates in the Later Neolithic, but a cremation burial
sealed beneath its bank dates from the Earlier Bronze Age (Sheridan 2003c:
167); the interval may have been as much as five hundred years. Gibson suggests
that far from providing a monumental setting for the buildings inside them,
the construction of the earthwork brought their use to an end. It is still more difficult to relate the palisaded enclosures to the structures
that were built within them. At present the only satisfactory evidence comes
from Mount Pleasant, where a substantial earthwork seems to have been supplemented
by a circuit of posts. There is no way of relating the bank and ditch
to the timber circle on the site, but it is clear that the palisade was a later
development. This is shown by the radiocarbon dates and also by their ceramic
associations. The palisade was constructed at a late stage in the development
of the site, and in time it was burnt down (Wainwright 1979).
A few earthwork or palisaded enclosures were also associated with mounds.
Whilst these are often thought of as barrows, that may not have been their
purpose, and excavation at the largest of these, Silbury Hill inWessex, showed
that it had not covered a burial (Fig. 3.21; Whittle 1997). Although they are
poorly dated, they might be considered as raised platforms whichwould provide
a vantage point from which a small number of people could observe activities
that were otherwise concealed. It is certainly true that Silbury Hill commands a
view into theWest Kennet palisaded enclosures, just as the building of Conquer
Barrow on the edge of the henge at Mount Pleasant might have allowed people
to see over the palisade into the interior of the monument (Wainwright 1979).
The same could have happened at other sites, although it is not clear whether
the principal function of these platforms was to raise certain people above the
level of those inside the monuments or to allow them privileged access to
events that were hidden from view. It is usually supposed that these practices
were peculiar to southern England, but that need not be true. There seems to
have been a mound on the axis of the entrance at Dunragit ( J. Thomas 2004b), and viewers could have seen into the Later Neolithic enclosure at Newgrange
from the raised area on top of the largest chambered tomb.
MAKING CONNECTIONS
Although this account has dealt separately with timber circles and stone circles,
earthworks and palisaded enclosures, it is clear that these elements overlap and
were to some extent interchangeable. There seems no point in regarding them
as separate ‘types’, for not only do the categories overlap, structures of ostensibly
different forms are often found together within the same complexes. These
might juxtapose open sites and enclosures, large monuments and much smaller
ones. In the same way, certain sites are associated with considerable numbers of
artefacts, whereas others have very few. This cross-cuts any classification based
on their architecture and implies that some places were used in very different
ways from others. Certain structures may have been visited infrequently;
others may even have been inhabited, whatever their more specialised roles
(Wainwright 1978).
Nonetheless certain tendencies can be identified amidst so much variety. In
some cases these monuments increased in size until they reached a peak in the
middle and late third millennium BC, after which theywere usually constructed
on a smaller scale.Wooden structures thatwere subject to decaywere sometimes
replaced in stone, and when this happened the activities associated with them
seem to have changed their character so that a site which had originally been
used for feasting became a place for the dead. Even when this did not happen
it seems that at a late stage in their history, some of these monuments were
screened off – and possibly closed – by an earthwork or a palisade.
That did not happen in every area. Many timber and stone circles were never
conceived on an elaborate scale. Nor were their positions emphasised by the
construction of enclosures. Some of the regions in which henge monuments
are apparently rare include Neolithic post circles which are not associated with
any earthworks. That was particularly true along the North Sea coast where
their place was sometimes taken by small palisaded enclosures. That contrast
between larger and smaller monuments is found in many parts of Britain and
may reflect important differences in the ways in which societies were organised
or in their capacity to mobilise workforces for the building of monuments ( J.
Harding 1995: 127–32). In Wessex, the Thames Valley, northeast England,
central Scotland, Orkney, and the Boyne Valley large monuments were the
norm, but In East Anglia, the English midlands, and northern Scotland smaller
structures may have played the same roles.
A striking feature of the larger sites is the way in which individual monuments
or groups of monuments increased in scale in parallel with one another.
Within the limits of radiocarbon dating this happened over a restricted period
of time, and some of these developments could have taken place simultaneously in different regions. The consumption of human labour was associated with
consumption of other kinds, as these sites were used for lavish feasts and for
deposits of special artefacts (Richards and Thomas 1984; Albarella and Serjeantson
2002). Neither process was confined to the interior of these monuments,
and similar material was buried in pits in the areas around them. Not all
these pit deposits were distinct from those on settlements, but they were created
with greater formality and had a wider range of contents. The position of one
of those outside Durrington Walls was marked by a cairn (Stone and Young
1948), and at Lawford and Crouch Hill middens with similar contents seem
to have been enclosed by a ditch (Shennan, Healy, and Smith 1985; Gardiner
1987). Artefacts were deposited also in water. In Ireland ard shares were placed
in rivers together with stone axes (D. Simpson 1993).
Sometimes the closest connections seem to be between regions that were
long distances apart. For example, the oval stone settings described as ‘coves’
are found inside a small number of stone circles, but these are distributed
from Scotland to Wessex (Burl 2000: 31–3). Similarly, the double-ditched
henges of northeast England are closely paralleled in the upper Thames Valley
(Bradley and Holgate 1984). If communities were seeking to emulate each others’
achievements, that process was geographically extensive. It seems to have
extended from Orkney to southern Britain and drew in people on either side
of the Irish Sea. What is quite remarkable is that after the demise of passage
tombs, it does not seem to have involved any reference to Continental Europe.
Monuments are the embodiment of particular ideas and there will always be
room for doubt over the strength of such connections. Fortunately, they are
evidenced in other media, among them artefacts which had moved a long way
from their sources. This is most obvious in the case of ground stone axes, which
were made in increasing numbers at quarries in Britain and Ireland. Some of
them had been established during the Earlier Neolithic period, but petrological
analysis suggests that their distribution was most extensive during the currency
of henges (I. Smith 1979). That is very striking, for much of the stone had been
obtained in remote and inaccessible locations, on mountains like Tievebulliagh
and Pike O’ Stickle. They were produced also on offshore islands like Lambay
Island and Rathlin Island which seem to have been inhabited at the time. A
high proportion of their products are found in distant areas which had adequate
raw materials of their own (Bradley and Edmonds 1993: chapter 9; Cooney
1998). Moreover, the contexts of these finds suggest that when their use-life
was at an end, some of them were deposited with some formality. Similar considerations
apply to the use of high quality flint. Some of the most elaborate
artefacts – polished knives and finely flaked arrowheads – were made in workshops
near to the northeast coast of England (Durden 1995), whilst similar
raw material was extracted by mining at Grimes Graves close to the Fen Edge
(Mercer 1981c). It seems likely that the finest of their products were distributed
along the North Sea littoral from northeast Scotland to the Thames Estuary.  Such connections are equally apparent from the distribution of Grooved
Ware. This ceramic tradition seems to have developed in Orkney, where it
has features in common with the designs on houses, passage graves, and stone
artefacts, but it was soon adopted in other parts of Britain and Ireland. The
Grooved Ware tradition is peculiar to these two islands, and yet it is widely
distributed within them. In that respect it contrasts with the kinds of decorated
ceramics that developed from the mid-fourth millennium BC. British and Irish
Grooved Ware can be divided into a series of substyles, but the main source
of variation seems to be chronological rather than geographical (Garwood
1999). Their styles vary according to the contexts in which these vessels are
found. Grooved Ware is a particular feature of Later Neolithic ceremonial
centres, and it may be no accident that occasional vessels found at these sites
are decorated with motifs that occur in megalithic art. These vessels might have
been made locally, but such designs link them to widely distributed traditions
of monumental architecture.
One way of explaining such connections is by studying where the main
monuments are found. A number of writers have expressed surprise that they
should be situated so near to Roman roads or forts, or even close to the
royal centres of Early Medieval Scotland (Loveday 1998; Driscoll 1998). The
probable explanation is that such places were especially accessible and that they
were located on obvious routes across the landscape. Some of these areas may
have included Later Neolithic settlements, but that is not always apparent,
and many of the places with most surface finds of the same period lack such
monuments entirely. Henges and related structures can be found along the
valleys leading through the uplands, near to navigable rivers and the places
where they were easiest to cross, and also near to the sea. These would have
been the ideal routes for later roads to follow and would have been well placed
as power bases from which to oversee the local population.
Sometimes it is possible to connect this evidence with the movement of
portable artefacts. The stone axes mentioned earlier provide the best example.
Some of these were made in the Cumbrian mountains throughout the
Neolithic period and were brought down to the surrounding lowlands, where
they were ground and polished (Bradley and Edmonds 1993: 144–53). There
are a number of major monuments in this area, each of them located on one of
the routes leading towards the quarries. There was one group of major monuments
near to Carlisle on the route leading north into Scotland (McCarthy
2000). To the north east of the high ground there is also the unusual henge
monument of Mayburgh, an embanked enclosure of Irish type (Topping 1992).
It is located beside one of the principal routes across the Pennine mountains.
On the opposite side of the high ground there was another henge at Catterick
which was built in exactly the same way (Moloney et al. 2003): the two sites
seem to be paired (Fig. 3.22). A ground stone axe was deposited in the entrance
of Mayburgh, whilst the lowlands between the Pennines and the North Sea contain an unusual concentration of artefacts from Cumbria (Manby 1988).
Other monument complexes are found to the south of Catterick, each of them
at the entrance to a similar route leading through the high ground towards the
west coast of Britain.
The strategic siting of such monuments is apparent at a more general level.
Deposits of unusual or exotic artefacts are commonly found near to them.
Good examples are the cluster of such finds around the monument at Arbor Low (Bradley and Hart 1983) or the concentration of broken mace heads close
to the Ring of Brodgar (C. Richards 2004b: 223–6). The same happens in
Wessex, where special artefactswere deposited in pits andwere accompanied by
a selection of animal bones and decorated pottery ( J. Thomas 1999: chapter 4).
A similar pattern is found in northeast England, although there some of the
imported items may have been deposited in water ( J. Harding 2003: 97).
Other artefacts from the vicinity of these sites provide a different kind of
information. These are the less elaborate flint tools that were used and discarded
in the surrounding area. Around the Thornborough complex they included raw
materials which could not have been obtained in the vicinity. It seems likely that
they provide one indication of the distances that people had travelled to these
places. Jan Harding (2000) has compared their visits to a pilgrimage, and Colin
Renfrew has taken the same approach to the henge monuments of Orkney
(2000: 16–17). Often such structures were built in groups, and people may
have needed to move between them, performing particular rituals in particular
places. Alternatively, individual monuments may have been the prerogative of
different sections of the population, in which case their construction and use
could have provided a focus for competition.
Such monuments were located not only for their accessibility. It was important
that they should also acknowledge what had been inherited from the
past. Although some henges are located close to causewayed enclosures, Later
Neolithic monuments are more often associated with cursuses. Occasionally
there is a direct connection between the new construction and an older earthwork.
The bank of the henge at Catterick incorporated a round cairn associated
with Peterborough Ware (Moloney et al. 2003), and the same may have happened
at Arbor Low, although the sequence is usually thought to be the other
way round (Barnatt and Collis 1996: 133–4). The stone circle of Long Meg and
her Daughters was built beside an existing earthwork which may be related to
a causewayed enclosure. Some of the monoliths had slumped into its infilled
ditch (Burl 1994). Similar earthworks have been identified in the Thornborough
complex, and here it seems as if the henges had been superimposed on
these older structures: an interpretation which is supported by recent fieldwork
( J. Harding 2000).
Itwas still more important to create these monuments in places with the right
natural setting, for many of the henges used the surrounding topography in the
same ways. From Avebury to Orkney, circular monuments were purposefully
built at the centre of circular landscapes: places that were surrounded by a ring
of hills (Bradley 1998b: chapter 8). Sometimes the horizon merged with the
enclosure bank, but in other cases it was concealed from view. Aaron Watson’s
fieldwork has demonstrated that such effects had been carefully contrived so
that a particular monument might appear to occupy the centre of the world.
Had it been sited anywhere else, those visual effects would have been lost (Watson 2001 and 2004). If such a site were to provide a microcosm of the
surrounding country, it was vital that it should have an unimpeded view of
the horizon. A number of enclosures were also aligned on prominent hills
or valleys, and others on the position of the sun (Parker Pearson et al 2006:
238–40).
A good way of summarising these points is to consider the best known of all
the prehistoric monuments in Britain, for it epitomises many of the processes
that have been described so far.
In some respects Stonehenge is unique. It has an unusual structural sequence;
its architecture makes use of techniques that are not found anywhere else;
some of its raw materials were introduced from a great distance; and its scale is unprecedented. At the same time, each of these characteristics is
related to more general trends during the Later Neolithic period and helps
to define the issues that need discussing here (Cleal, Walker, and Montague
1995).
The first enclosure at Stonehenge is one of a distinctive group discussed
at the end of Chapter Two. It dates from 3000 BC or a little after, and its
form seems to be related to that of the last causewayed enclosures. Its ditch
was interrupted at many points and had been deliberately refilled to cover a
number of placed deposits, mainly of animal bones (ibid., chapter 5). This
earthwork was probably contemporary with a nearby cursus, although only
small parts of those sites are visible from one another ( J. Richards 1990: 93–6).
The segmented enclosure at Stonehenge was precisely circular and was located
in the centre of a circular landscape formed by a horizon of higher ground,
but that effect was limited to the area around the monument. The Stonehenge
Cursus was aligned on a long barrow and on the equinoctial sunrise. It also
pointed towards the valley of the River Avon.
It is ironic that the enclosure at Stonehenge should have given its name to
an entire class of sites, for it is a class to which it does not belong. In contrast
to the earthwork of a henge, its bank is inside the ditch, and those elements
that do have affinities with monuments of that kind belong to a later period.
Although there are no absolute dates for the second phase at Stonehenge, the
existing enclosure was probably associated with two structures, as well as an
enigmatic setting of timber uprights in its northern entrance. At the centre of
the monument there may have been a timber building, although its plan can
no longer be recovered (Cleal, Walker, and Montague 1995: chapter 6). That
is because its position is still occupied by standing stones, but also because it
formed the principal focus for antiquarian excavations. Even so, it is known
that post sockets existed in this area and that some of them were earlier than the
monoliths. The distribution of these post holes is confined to the zone occupied
by the later stone circles, suggesting that one kind of structure replaced another,
as so often happened at these sites.
The second structure is undated. It consists of two parallel rows of posts
extending between another entrance to the monument and the central area
just described. These seem to have formed a narrow avenue. Close to the centre
of the site it was interrupted by a wooden screen (ibid., chapter 6). Exactly the
same arrangement is found at Durrington Walls, where its prehistoric context
is well established. At Durrington, it led to the Northern Circle, a ring of
uprights associated with Grooved Ware (Wainwright and Longworth 1971:
fig. 17).
The third component of this phase at Stonehenge is a series of pits concentric
with the inner edge of the bank. They contain deposits of artefacts and cremated
bone in their secondary filling. One reason for suggesting that they are later in
date than the earthwork enclosure is that more cremation burials were found in features dug into the top of the bank and the upper levels of the ditch.
Richard Atkinson (1956 and 1979) interpreted this as evidence of a cremation
cemetery, but that idea is untenable, for these features had obviously been
refilled before such deposits were made. It is better to follow the ideas of the
original excavator, Colonel Hawley, who regarded the Aubrey Holes as post
sockets. He observed signs of friction on the edges of these features where
uprights had been manoeuvred into place (Cleal,Walker, and Montague 1995,
102). They are likely to be the remains of a palisaded enclosure of the early
form defined by Alex Gibson (1998).
By this stage an earthwork which was probably related to a causewayed
enclosure had been reconstructed as some kind of henge. This was not a unique
occurrence, as a similar sequence has been postulated at a series of monuments
in northeast England, the best known of them at Thornborough ( J. Harding
2000). The interior of Stonehenge was screened off from the surrounding area
not by one palisade but by two, and the positions of older structures were
marked by deposits of human bone. This may have marked the closure of the
original monument. It also prefigured an increasing concern with the dead.
The next development at Stonehenge typifies a wider development in the
late third millennium BC. This was the translation of the timber monument
into stone. It is not quite clear when this happened, and it may have been a
complex process. There seem to have been two main elements which were
not necessarily built simultaneously. There are the four Station Stones (Cleal,
Walker, and Montague 1995: 26). These are located just inside the enclosure
bank where they form a precise rectangle. This has never been adequately explained, and yet the basic configuration seems familiar. A circular earthwork
perimeter enclosed a rectilinear arrangement of uprights. This resembles the
layout of several timber monuments, including the North Circle at Durrington
Walls (Wainwright and Longworth 1971: 41–4). It may be that the first setting
of monoliths at Stonehenge conformed to the same organisation of space. It
is more difficult to reconstruct the arcs of monoliths in the middle of the site.
There appear to have been two circles of uprights, arranged in pairs, although
they may never have been completed (Cleal, Walker, and Montague 1995:
chapter 7).
The form of the innermost stone setting raises problems, for the uprights
were later removed and erected in other positions. Nevertheless some of the
reused monoliths had been shaped so that they could support lintels. It is a moot
point whether that plan was abandoned before the project was complete, or
whether these stones had been brought from somewhere else. At all events it
suggests that a deliberate attempt was being made to copy the characteristics
of a timber building in a more durable medium (Gibson 2005: 143–51). Such
a scheme was executed during a later phase in the history of Stonehenge, but
using a different raw material.
The original project had made use of the distinctive types of rock which are
generally called ‘bluestones’. They originated in southwest Wales, in another
landscape with stone and earthwork monuments. They also came from a small
area where stone axes were made (Thorpe et al. 1991). It is not clear how they
were transported, but enough is known to reject the idea that they had been
carried to southern England by melting ice. It would have taken a huge effort
to obtain them and to bring them to the site. Thismust be the ultimate example
of the links between such monuments and materials from distant locations.
By 2400 BC the abandoned scheme seems to have been reinstated using a
different raw material. Two massive structures were built, using local sarsens
instead of imported bluestones. Although some of its elementswere rearranged,
that really represents the final form of the monument. It consisted of a massive
circle of sarsens joined together by lintels enclosing a second horseshoe-shaped
setting in which pairs of uprights were linked in a similar fashion. Inside each
of these circuits there were low pillars ofWelsh rock which seem to have been
reused from an earlier structure (Cleal,Walker, and Montague 1995: chapter 7).
Each stone must have had its own history, for nearly all those making up the
inner setting of bluestones came from the same source. By contrast, the outer
settings made use of material from a variety of different places in the Presceli
Mountains. The finished structure was not only a striking monument in itself;
it referenced a landscape over two hundred kilometres away (Bradley 2000b:
92–6).
During the same phase the perimeter of Stonehenge was modified, and now
an earthwork avenue extended from the northern entrance of the enclosure to
the Avon where there were timber circles close to the river atWoodhenge and Durrington Walls (Cleal, Walker, and Montague 1995: 291–329). There must
have been a link between these different sites, for Stonehenge was very much a
copy of awooden monument. The sarsens had been shaped, and the lintelswere
secured in place using joints that would have been more appropriate in timber
carpentry. The surfaces of some stones were treated so that they resembled
carved wood, and one of them had probably been used to polish axes. Much
later, in the sixteenth century BC, bronze axeheads were depicted by carvings
on three of the monoliths (ibid., 30–4). In many ways this structure epitomises
tendencies that have already been identified at other monuments. For example,
the sarsen horseshoe is tallest on the southwest side of the circle, and the
bluestone pillars are paired with the sarsen uprights in a similar fashion to the
composite structures at Mount Pleasant and the Sanctuary. Still more striking,
the finished structure is obviously aligned on the sun. Looking northeastwards,
it rises at midsummer over the course of the avenue. To the southwest, it
sets at midwinter behind the tallest pair of stones. The rising ground along
the final section of the Avenue enhances this effect. From this position the
horizon is hidden from view, and behind the circle there is nothing but the
sky. The whole layout of the monument can also be understood as a series of
barriers screening an area in its centre which few people could have ever have
seen.
That hypothesis applies to many of the larger monuments constructed during
the later third millennium BC, but the building and use of Stonehenge
introduce some new elements. Although many of these finds are poorly documented,
it seems clear that the monument was associated with an unusually
large quantity of human bones. Few of them have any contexts and none has
been accurately dated, but if this observation is correct it adds weight to Parker
Pearson’s and Ramilsonina’s interpretation (Parker Pearson and Ramilsonina
1998). Their case is all the stronger since radiocarbon dates suggests that the
sarsen structure at Stonehenge would have been contemporary with the timber
buildings at DurringtonWalls andWoodhenge (Parker Pearson et al 2006:
233). Not only were the stone structures a more durable version of those timber
monuments, their associations were completely different from one another.
The timber circles were associated with many portable artefacts and a considerable
quantity of animal bones. Stonehenge contained very few objects. In
their place were finds of human remains.
As the associations of such monuments changed, so did the scale on which
they were built. Stonehenge must have made greater demands on human
labour than any other building of the time, and yet it is comparatively small.
Even the final sarsen circle is not much over thirty metres in diameter. This
chapter has considered the ways in which access to particular monuments was
restricted. This structure took that process to its limits. A building that would
have required more labour than ever before was impossible for large numbers
of people to visit. The crucial transition from a timber monument, less impressive than its
neighbours at Durrington and Woodhenge, to a uniquely complex piece of
architecture came at a special time in the archaeology of southern England,
for this was when the first metalwork appears and when a new ceramic style,
the Bell Beaker, is found in graves. Both these developments would have been
impossible without renewed contacts with Continental Europe, and yet these
novel media surely expressed some of the same social divisions as more traditional
monuments like Stonehenge. This chapter began by suggesting that
monument building was one of the processes by which an elite might distinguish
itself from other members of society, and the growing scale of these structures
during the later third millennium BC suggests that competition within
and between different communities must have intensified at this time (Barrett
1994: chapters 1, 2, and 4). How appropriate that Beaker pottery should be
associated with the transformation of Stonehenge and that it should be at
Amesbury, only a short distance away, that one of the earliest burials associated
with these new forms of material culture should be found (Fitzpatrick 2003).
This was not only one of the first of a new tradition of individual graves, it was
one of the richest. A Beaker mass grave was found nearby, and close to it there
was another enclosure containing sherds of Grooved Ware (Fitzpatrick 2005).
The growth of Later Neolithic monuments was sometimes associated with the
promotion of long distance relationships. Now that process extended beyond
Britain and Ireland, and both these regions were drawn into an international
network.
A WORLD ELSEWHERE
The first appearance of Beaker pottery at Stonehenge forms a logical part of a
sequence in which the successive buildings on the site became more elaborate
but less accessible. The stone structures make sense only in terms of insular
developments. That applies even to the deposition of human remains, as that
had started when the timber monument went out of use. On the other hand,
it no longer seems possible to argue that individual burials or the building of
round barrows formed part of a current tradition in southern England. The
first Beakers were associated with a new development and one which must
have resulted from contacts with mainland Europe. Exactly the same applies
to the earliest metalwork.
If Grooved Ware was an insular ceramic tradition, originating in the north
but ultimately adopted in most parts of these islands, Bell Beakers represent
a very different phenomenon. Their distribution is truly international and
extends from Denmark to North Africa. They are found as far east as Hungary
and as farwest as Portugal (Salanova 2002). That has raised problems of interpretation,
for they are often associated with early metalwork as well as new burial
practices. It would be easy to argue that they represent a class of prestigious artefacts employed in social transactions, but that cannot supply the entire
answer. Like GroovedWare, Bell Beakers can be found in special contexts, but
they occur also in settlements.
One important influence was the practice of individual burial with grave
goods associated with Corded Ware in Northern and Central Europe (Case
2004a). Scholars have discussed its relationship with Bell Beakers, and for a
long time it seemed most likely that these two ceramic styles were used in
succession, so that the earliest Bell Beakers might have been in the north
where they could have developed out of the Corded Ware tradition (Lanting,
Mook, and Van derWaals 1976). Now that is uncertain. The oldest radiocarbon
dates for Bell Beakers come from Portugal (Kunst 2001), and it seems as if the
earliest vessels have a distribution extending up the Atlantic coastline from the
Iberian Peninsula (Salanova 2002). These finds are commonly associated with
human remains, some of them in megalithic tombs, and with the earliest use
of metal. The first Bell Beakers in Northern Europe are contemporary with
the vessels that were once taken as their prototypes.
That realignment poses a further problem. If Bell Beakers and CordedWare
were independent developments, how and why were elements of both these
two traditions combined? Bell Beakers are closely associated with metalwork,
including weapons and personal ornaments, and with archery equipment.
Corded Ware was also associated with burials, but these contained a rather
different assemblage in which stone battle axes and flint knives played a more
prominent part (Harrison 1980: chapters 2 and 3). Both groupswere dominated
by fine decorated pots which could have contained liquids, and some of those
found in the Iberian Peninsula certainly seem to have held fermented drinks
(E. Guerra Doce, pers. comm.). Humphrey Case (2004a, 2004b) suggests that
these sets of objects were employed to display social position.
Those two groups were first linked together through a well-established
exchange network based on the long distance movement of Grand Pressigny
flint from western France (Fig. 3.25; Salanova 2002). This distinctive honeycoloured
material bore some resemblance to metal and was distributed over
an enormous area extending along the Atlantic and Channel coasts as far as
The Netherlands and reaching inland until it connected the distribution of
early Bell Beakers to that of CordedWare. In time the two traditions lost their
separate identities, and it was during this second stage that further artefact types
were introduced to Britain and Ireland.
Despite this complex process, the new forms of material culture adopted
in these islands were part of an international phenomenon. That development
presents some problems of its own.
Chapter One discussed the ‘invasion hypothesis’ which had played such an
important role in prehistoric archaeology during the mid-twentieth century,
and the reaction that followed Grahame Clark’s review of this concept in 1966.
He took a minimal view of migrations during prehistory and attacked what he artefacts employed in social transactions, but that cannot supply the entire
answer. Like GroovedWare, Bell Beakers can be found in special contexts, but
they occur also in settlements.
One important influence was the practice of individual burial with grave
goods associated with Corded Ware in Northern and Central Europe (Case
2004a). Scholars have discussed its relationship with Bell Beakers, and for a
long time it seemed most likely that these two ceramic styles were used in
succession, so that the earliest Bell Beakers might have been in the north
where they could have developed out of the Corded Ware tradition (Lanting,
Mook, and Van derWaals 1976). Now that is uncertain. The oldest radiocarbon
dates for Bell Beakers come from Portugal (Kunst 2001), and it seems as if the
earliest vessels have a distribution extending up the Atlantic coastline from the
Iberian Peninsula (Salanova 2002). These finds are commonly associated with
human remains, some of them in megalithic tombs, and with the earliest use
of metal. The first Bell Beakers in Northern Europe are contemporary with
the vessels that were once taken as their prototypes.
That realignment poses a further problem. If Bell Beakers and CordedWare
were independent developments, how and why were elements of both these
two traditions combined? Bell Beakers are closely associated with metalwork,
including weapons and personal ornaments, and with archery equipment.
Corded Ware was also associated with burials, but these contained a rather
different assemblage in which stone battle axes and flint knives played a more
prominent part (Harrison 1980: chapters 2 and 3). Both groupswere dominated
by fine decorated pots which could have contained liquids, and some of those
found in the Iberian Peninsula certainly seem to have held fermented drinks
(E. Guerra Doce, pers. comm.). Humphrey Case (2004a, 2004b) suggests that
these sets of objects were employed to display social position.
Those two groups were first linked together through a well-established
exchange network based on the long distance movement of Grand Pressigny
flint from western France (Fig. 3.25; Salanova 2002). This distinctive honeycoloured
material bore some resemblance to metal and was distributed over
an enormous area extending along the Atlantic and Channel coasts as far as
The Netherlands and reaching inland until it connected the distribution of
early Bell Beakers to that of CordedWare. In time the two traditions lost their
separate identities, and it was during this second stage that further artefact types
were introduced to Britain and Ireland.
Despite this complex process, the new forms of material culture adopted
in these islands were part of an international phenomenon. That development
presents some problems of its own.
Chapter One discussed the ‘invasion hypothesis’ which had played such an
important role in prehistoric archaeology during the mid-twentieth century,
and the reaction that followed Grahame Clark’s review of this concept in 1966.
He took a minimal view of migrations during prehistory and attacked what he artefacts employed in social transactions, but that cannot supply the entire
answer. Like GroovedWare, Bell Beakers can be found in special contexts, but
they occur also in settlements.
One important influence was the practice of individual burial with grave
goods associated with Corded Ware in Northern and Central Europe (Case
2004a). Scholars have discussed its relationship with Bell Beakers, and for a
long time it seemed most likely that these two ceramic styles were used in
succession, so that the earliest Bell Beakers might have been in the north
where they could have developed out of the Corded Ware tradition (Lanting,
Mook, and Van derWaals 1976). Now that is uncertain. The oldest radiocarbon
dates for Bell Beakers come from Portugal (Kunst 2001), and it seems as if the
earliest vessels have a distribution extending up the Atlantic coastline from the
Iberian Peninsula (Salanova 2002). These finds are commonly associated with
human remains, some of them in megalithic tombs, and with the earliest use
of metal. The first Bell Beakers in Northern Europe are contemporary with
the vessels that were once taken as their prototypes.
That realignment poses a further problem. If Bell Beakers and CordedWare
were independent developments, how and why were elements of both these
two traditions combined? Bell Beakers are closely associated with metalwork,
including weapons and personal ornaments, and with archery equipment.
Corded Ware was also associated with burials, but these contained a rather
different assemblage in which stone battle axes and flint knives played a more
prominent part (Harrison 1980: chapters 2 and 3). Both groupswere dominated
by fine decorated pots which could have contained liquids, and some of those
found in the Iberian Peninsula certainly seem to have held fermented drinks
(E. Guerra Doce, pers. comm.). Humphrey Case (2004a, 2004b) suggests that
these sets of objects were employed to display social position.
Those two groups were first linked together through a well-established
exchange network based on the long distance movement of Grand Pressigny
flint from western France (Fig. 3.25; Salanova 2002). This distinctive honeycoloured
material bore some resemblance to metal and was distributed over
an enormous area extending along the Atlantic and Channel coasts as far as
The Netherlands and reaching inland until it connected the distribution of
early Bell Beakers to that of CordedWare. In time the two traditions lost their
separate identities, and it was during this second stage that further artefact types
were introduced to Britain and Ireland.
Despite this complex process, the new forms of material culture adopted
in these islands were part of an international phenomenon. That development
presents some problems of its own.
Chapter One discussed the ‘invasion hypothesis’ which had played such an
important role in prehistoric archaeology during the mid-twentieth century,
and the reaction that followed Grahame Clark’s review of this concept in 1966.
He took a minimal view of migrations during prehistory and attacked what he saw as an uncritical approach to the subject. He believed that there were only
two cases in which significant numbers of people had settled distant areas. One
was the introduction of agriculture discussed in Chapter Two, and the other
was the invasion, or invasions, of the ‘Beaker Folk’. This was so much part of
the framework accepted at the time that his pupil David Clarke (1970) actually
suggested a greater number of episodes of Beaker settlement from Continent
than previous scholars in this field. Only recently has it been possible to devise
a more scientific method for studying migration. This is based on the analysis
of human bone. The first results certainly suggest a certain amount of mobility,
but it is too soon to discuss its extent or significance (Fitzpatrick 2003).
Clarke took an increasingly flexible approach in a later publication (Clarke
1976). Thatwas because studies of the Bell Beaker phenomenon in Continental
Europe were meeting with a problem. Although this style of pottery might
be associated with early metalwork and with a restricted range of funeral gifts,
it was hard to identify the other elements that were necessary to define an
archaeological ‘culture’. There was no Beaker economy, these artefacts could
be found with a variety of monument types and, most important of all, Beaker
settlements and houses assumed many different forms. More often than not, this
distinctive material assemblage was associated with kinds of buildings that were
already well established (Besse and Desideri 2005). Faced with such anomalies,
Clarke (1976) talked of the Beaker ‘network’. Other researchers referred to a
Beaker ‘package’ (Burgess and Shennan 1976).
Again there are both empirical and theoretical problems to address. A practical
difficulty concerns Beaker chronology in Britain and Ireland. David Clarke
(1970) had postulated a series of migrations linking specific parts of Continental
Europe to particular parts of these islands, identifying such links on the basis of
pottery types and their associations. Once the colonists had settled here, they
made their pottery in more local styles which gradually changed their character
over time. Continental scholars disputed some of the long-distance connections
on which his interpretation was based and proposed a simpler scheme.
The British and Irish sequence had apparently moved in parallel with welldocumented
developments in The Netherlands (Lanting and Van der Waals
1972).
It remained to test these ideas by a programme of radiocarbon dating. This
was undertaken by the British Museum in the late 1980s (Kinnes et al. 1991).
The results were unexpected, for they did not support any of the existing
schemes. They suggested that there was little evidence for a succession of different
types, even when it was indicated by artefact associations. Some of the dates
were exceptionally late. This left students of the period in a quandary. Most executed
the difficult manoeuvre of rejecting the validity of any Beaker chronology
yet ascribing their material to styles which presuppose such a sequence.
Since the Beaker dating programme was published in 1991, more radiocarbon
dates have been obtained and a clearer pattern has begun to emerge
(Needham 2005). The overall date range has been narrowed to between 2,400 and 1800 BC, and now it does seem as if the earliest Beakers are especially close
to those found on the Continent; most later types represent insular developments.
It is only with the earlier styles that human migration can be postulated.
Whatwere the contexts of Beaker material in Britain and Ireland? They differ
so fundamentally that the contrasts between them help to identify the issues
that need discussion. It is commonly supposed that fine Beakers are associated with burials, where they form part of a stereotyped assemblage associated with
specific individuals. That was not what happened in Ireland. As this is the area
with the earliest copper mines, it is the logical place in which to begin this
account. Although it includes pottery with stylistic links to both western and
central Europe, Ireland was directly connected to the Atlantic seaways and in
an excellent position to be drawn into a network that had originated in Iberia
(Case 1995).
The introduction to this chapter touched on the problems of using a period
framework based on metal technology. That is important as there is a contrast
between the first use of copper and that of bronze in these islands. In some
parts of Europe metalworking was adopted through a growing familiarity with
the raw materials. Occasional items of metalwork might have been imported
into what was essentially a Neolithic setting. This did not happen in Britain
and Ireland where the first adoption of metals is best described as an event.
It marked the culmination of a process by which copperworkingwas adopted
in one society after another as the requisite knowledge became available. By
3000 BC, it had been taken up in north-central Europe and across the Iberian
peninsula, and in the mid- to late third millennium its distribution expanded
to Britain, Ireland, France, and south Scandinavia. The production of bronze
artefacts involved the addition of tin, and across the Continent the full adoption
of this process took longer. Moreover, the idea of doing so seems to have
travelled in the opposite direction, so that it had happened by 2200 BC in
Britain where local tin was available, but did not reach Central and Northern
Europe for at least another two centuries. In Spain and Portugal the change
came even later (Pare 2000).
The earliest Irish metalwork has always presented problems. It consists mainly
of axeheads, although there are also halberds and knives or dagger blades. All
were made of a distinctive type of copper whose most likely origin was in the
southwest of the country. Products from this source were distributed throughout
the island and also occur in Britain, particularly towards the west; in the
south, metalwork was imported across the English Channel. On chronological
grounds all these artefacts should date from the same period as early Beakers,
but the different kinds of material were not found in association with one
another. The problem was summed up by Humphrey Case in an article entitled
‘Were the Beaker-people the first metallurgists in Ireland?’ (1966). Whether
or not metalworkers had used Beaker pottery, where had they acquired such
a specialised technology? Despite the reaction against prehistoric migrations,
it was clear that such complex technological processes as metallurgy had not
developed spontaneously. They had to be taught and learnt.
Both these problems are closer to resolution with excavations at Ross Island,
near Killarney in southwest Ireland (O’Brien 2004a). This site had always been
suggested as the source of the earliest copper extraction, but until fieldwork
took place there in the 1990s it was assumed that any archaeological evidence had been destroyed. Fortunately that was not the case and it was possible to
investigate not only some of the mines where the rawmaterialwas extracted but
also a work camp in which the ore was processed. That specialised settlement
included a number of circular buildings and was associated almost exclusively
with Beaker pottery. At last the beginnings of Irish metallurgy could be set in
a wider context.
The project shed some light on the processes followed at the mines. This
permitted a tentative comparison between the technology employed at Ross
Island and early metallurgy in other parts of Europe. O’Brien suggests that the
closest links were with the procedures followed along the Atlantic coastline
between Spain and Normandy. The dates from Ross Island show that mining
there commenced around 2400 BC and continued until about 1900 BC.
Metalwork of similar composition and character is found in western France.
Apart from the settlement of Lough Gur, also in southwest Ireland, there
are no other sites where early metal is associated with Beaker pottery. That
is probably because they were deposited according to different conventions.
As O’Brien observes, the earliest metalwork is dominated by axeheads. They
may have replaced their stone equivalents, and, like them, they are sometimes
found in hoards or votive deposits. In contrast, Beaker pottery may have taken
on the existing roles of GroovedWare. In some cases it has been quite difficult
to tell the difference between these two styles.
Irish Beakers are associated mainly with settlements and monuments,
although there can be problems in distinguishing between these kinds of site.
The best known settlements are those at Newgrange and Knowth (O’Kelly,
Cleary, and Lehane 1983; Eogan and Roche 1997). In both cases there seems
to be the same link with an important structure from the past, and a small
tomb at Knowth actually includes one of the few Beaker burials in Ireland.
The supposed settlement at Newgrange was located in front of the entrance
to the tomb and overlay a deposit of rubble which had fallen from its mound.
Although it has been claimed that the settlement includes the remains of houses,
its main features were stone-lined hearths. The Beakers were apparently found
together with GroovedWare, and it seems possible that both were deposited in
the course of feasts (Mount 1994). At the neighbouring monument of Knowth,
Beaker ceramics were stratified above the levels containing GroovedWare, but
even here the excavated assemblage presents some problems. It was associated
with concentrations of ceramics, flint artefacts, and hearths. The Beaker pottery
is unusual, for some of these groups include a high proportion of finely
decorated vessels (Eogan and Roche 1997: chapter 5).
Beaker pottery is associated with another distinctive class of monument.
This is the wedge tomb which has a wider distribution than any other type
of megalith in Ireland (O’Brien 1999). These monuments will be considered
later in this chapter, but at this point it is worth saying that similar associations
are found in Atlantic Europe, where many of the Beaker burials were associated with the reuse of chambered tombs. In Ireland the decorated vessels
are generally found with cremations and are not accompanied by other grave
goods. Radiocarbon dates suggest that wedge tombs were built between about
2400 and 2100 BC (Brindley and Lanting 1992). After their construction they
remained in use for a considerable time. During their later history the wedge
tombs in the west of Ireland provided an alternative to the cist cemeteries in
the eastern part of the country.
One reason for stressing this contrast is that the very artefacts that were used
as grave goods in Britain are found in different contexts in Ireland. That obviously
applies to the Beakers themselves, but it is also true of the objects that
one might expect to find with them (Waddell 1998: 199–223). The characteristic
arrowheads of this period do occur in wedge tombs, but they were also
deposited in a hoard. The stone wrist guards that are associated with Beakers
in British burials have sometimes been found in bogs. The tanged copper knife
is another type that can occur in graves, but this did not happen in Ireland,
where again some of these artefacts may be associated with hoards and votive
deposits. When a tradition of individual burial first developed in Ireland, as
it did about 2200 BC, it was associated not with Beakers but with a different
style of pottery, the Food Vessel (Sheridan 2004b).
Other kinds of metalwork are rarely associated with the dead either in Britain
or Ireland. This is particularly true of axes, which seem to have been the main
products of the Irish smiths. These are generally found in isolation or with
other metal artefacts. Some were deposited in dry land hoards and occasionally
marked by a stone, but often they were associated with water and placed in
bogs, rivers, and lakes. In some cases it is clear that they had been laid down
with a certain formality. Even the single finds carried a special significance.
Stuart Needham (1988) has shown that they changed their associations during
the Earlier Bronze Age. The first examples were placed in bogs and the later
ones in rivers.
The early beginning of copper working in Ireland should not overshadow
the use of native gold, for this also began in the later third millennium BC.
Three kinds of artefact are important (G. Eogan 1994: chapter 2). There are
gold discs which were produced from about 2300 BC, as well as the distinctive
artefacts which are often described as ear rings, although they were probably
hair ornaments. Beaker-associated gold trinkets are known from various parts
of Europe, but these ones are relevant because they were probably produced
in Ireland. They are associated with the earliest Beaker burials in Britain and
date from approximately 2350 to 2150 BC. The other form is the lunula,
which is often interpreted as a decorated collar. A few come from Britain and
northwest France, but again the design appears to be Irish (Fig. 3.26). They
are usually single finds, and none is clearly associated with a burial. They were
made during the currency of Beakers and share their characteristic decoration
( J. Taylor 1970). It shows the special significance of this style of pottery. That contrasts with the situation in Britain. To some extent it may be due to
its geography, for the closest connectionswere probably those across the English
Channel and between the east coast and The Netherlands. They may account
for the differences between the burials in Britain. Some of them include the
classic components of the Bell Beaker repertoire. Others contain battle axes
whose forms refer back to the CordedWare tradition, and yet their distributions
overlap completely. The Irish practice of placing bodies in chambered tombs is represented only in northern and western Scotland, where Earlier Neolithic
monuments were selectively reused (Bradley 2000a: 221–4). By contrast, the
Central and Northern European practice of burying Bell Beakers in graves is
found in most parts of Britain (Clarke 1970).
Many of the same kinds of artefacts are found on both sides of the Irish
Sea, but in Britain they are normally associated with burials in small round
barrows. During an initial phase between about 2500 and 2150 BC such mounds
were usually small and associated with one inhumation burial, generally a
male, located in a central position beneath the monument (Garwood in press).
There were also flat graves. They include archery equipment, knives, daggers,
awls, and a variety of personal ornaments. Gold is found occasionally and
appears in the earliest burials. Although they were contemporary with large
henge monuments, they could be located a short distance away from them. For
instance, the rich Beaker burials recently excavated at Amesbury were over a
kilometre from Durrington Walls, although they were near to a circle of pits
associated with Grooved Ware (Fitzpatrick 2003, 2005). A similar pattern has
been identified around Stonehenge (Exon et al. 2000) and also occurs around
the Devil’s Quoits in the Upper Thames Valley (Barclay, Gray, and Lambrick
1995).
Beakers are found in settlements where they form the predominant, and
sometimes the only, ceramic style. These sites are more common than they
are in Ireland and seem to be a particular feature of the Western Isles and the
North Sea coast. Alex Gibson (1982) has studied the excavated material. His
work shows that the vessels occur in a wide range of sizes and fabrics and could
represent a complete assemblage rather than a set of specialised equipment. For
the most part the domestic ceramics are more robust than their equivalents in
graves, suggesting that certain vessels were meant to withstand normal wear
and tear whilst others might be finely finished but were used only once (Boast
1995). The occupation sites are often quite ephemeral but include exactly the
same mixture of pits, middens, and circular or oval houses as those associated
with Peterborough Ware and Grooved Ware. There does not appear to have
been a separate type of ‘Beaker’ house.
In southern England, some of the best preserved evidence is sealed by
deposits of hill wash in valley bottoms (M. Allen 2005). Here there are traces
of ard cultivation. Similar evidence is found on sand banks and gravel bars
beside the River Thames in London (Sidell et al. 2002: 30–5). On the edge of
the Fenland near Peterborough, Beaker pottery has been found together with
groups of stake holes (F. Pryor, pers. comm.). Some may mark the positions
of settlement sites, and there seem to be fence lines suggesting the existence
of small plots or fields. This is especially interesting in view of Pryor’s suggestion
that important land divisions were marked by the deposition of Beaker
artefacts in pits (2001: 70–2). These could well have been integrated into a
network of lightly built fences that would leave little trace behind. Perhaps they can be compared with the shallow gullies associated with Beaker and later
pottery found beneath the early medieval cemetery at Sutton Hoo in eastern
England (Fig. 3.27; Carver 2005: chapter 11). There may be similar evidence
from Snail Down in Wessex, where the earthworks of two field boundaries
seem to underlie a group of round barrows (N. Thomas 2005: 73–6). A circular
house was associated with the settlement at Sutton Hoo, but at Snail Down
arcs of stake holes provide the only evidence of domestic buildings. Something
similar may have happened at Belle Tout on the coast of southeast England where an ephemeral earthwork enclosed an occupation site associated with
Beakers and Food Vessels (Bradley 1970).
An important link with traditional practice is found at some of the henges.
For the most part, Beaker pottery is associated with the stone phases of these
monuments, although this association is not always clear-cut. One way of
viewing the connection is to suggest that the adoption of Beakers happened
once those structures became more directly associated with the dead. The same
idea might be expressed by graves in the vicinity. At the same time, it seems
likely that the distribution of cultural material at these sites was governed by
strict conventions: certain items had to be deposited in certain places. That was
obvious during the Grooved Ware phase, but now it seems as if some of the
same protocols determined the ways in which Beakers were to be deployed.
That is especially obvious with the excavated material from Mount Pleasant,
where there were continuities in the distribution of artefacts and animal bones
( J. Thomas 1996b: 212–22). Again the adoption of Beaker pottery did not
involve a radical change.
A more obvious development affected some of the megalithic tombs in
Scotland, as they were selectively reused for Beaker burials. There is little sign
that their structures were modified, although certain architectural details may
have been copied in building a new series of monuments (Bradley 2000a:
220–31). The reused tombs contain human remains and pottery, but there are
few other artefacts, and some were blocked during this phase. Other monuments
may have been ‘closed’ in a different way. This could have happened
to some of the timber and stone circles which were surrounded by ditches or
palisades at a late stage in their history. In northern Scotland earlier monuments
might also be closed by the erection of a ring of monoliths (Bradley 2005a,
100–6).
If the first deposits of metalwork were in Ireland, by 2000 BC they were
common in Scotland, although there were few in England and Wales. That
change is significant as it spans the period in which bronze working developed
in Britain (Rohl and Needham 1998). Some of the raw material was imported
from Ireland, but now this was probably alloyed with Cornish tin. The British
metalwork was more varied than its Irish counterpart which was dominated
by large numbers of axes and halberds. These are the kinds of material that are
found in hoards and in natural places like bogs and rivers. It emphasises the point
that during the Beaker ceramic phase one group of artefacts might be associated
with the dead, whilst another was employed in a quite different sphere. Stuart
Needham (1988) suggests that the deposits in graves were connected with the
individual, and those in hoards with the wider community.
These new developments articulate with the practices discussed in the first
part of this chapter. Irish copper axes were deployed in the same fashion as
the stone axes that they replaced, and Beaker pottery took on many of the roles of Grooved Ware. There was no change in the pattern of settlement,
and people still lived in insubstantial round houses, as they had done for over
five hundred years. In the same way, Beakers were associated with structural
changes at British henges which perpetuate, and even enhance, their traditional
architecture. They were deposited according to the same conventions
as Grooved Ware, and so was the material associated with them. It was only
the development of a new tradition of individual burial that marks a radical
departure, and even this can be explained as the visible outcome of the social
divisions promoted by the building and use of large monuments ( J. Barrett
1994: chapter 1). People formed new alliances to highlight their own positions,
and in doing so they also adopted a new technology. No doubt that
process involved the movement and exchange of personnel, but its roots were
in local developments that extended far into the past.
DISTANCE AND ENCHANTMENT
These questions have been considered in some detail because the same issues
are important in the Earlier Bronze Age. How was British and Irish material
culture related to that in Continental Europe? What is the significance of the
types of artefacts shared between these islands and the mainland, and does the
evidence of burials and metalwork hoards provide an accurate picture of life
during this period?
There can be little doubt that approaches to these problems have been influenced
by the archaeology of a few areas and, in particular, Wessex. They have
also placed an undue emphasis on the objects found in graves (Annable and
Simpson 1964). That was understandable in the early development of Bronze
Age studies, for it was only by identifying those objects with the widest distributions
that there was any prospect of relating insular material culture to the
historical chronology established in the Mediterranean. At the same time, a
small selection of exceptional artefacts has come to dominate the discussion
for reasons which are explained by the history of fieldwork. Long before settlement
excavation was considered possible or useful, well-preserved barrows
were investigated for grave goods. The work focused on the deposits beneath
the centres of these earthworks and rarely extended to the monuments as a
whole. That was because antiquarians believed that it was where the richest
material would be found. Since they did very little to record the structures of
the mounds, the contents of a biased sample of graves provided the material
available for study.
Those early excavations were devoted to standing monuments which survived
in great numbers on the chalk of southern and northeastern England
because both regions had been used as grazing land. In the nineteenth century,
there were people with the financial means and social connections to conduct large numbers of investigations. For a long time research focused on the contents
of their collections (Annable and Simpson 1964; Kinnes and Longworth
1985), and even now the results of their fieldwork provide a misleading impression
of the Earlier Bronze Age.
Such biases in the record can be remedied in several ways. More sophisticated
excavations have treated individual mounds, and even the areas outside them,
as cemeteries that might contain a whole range of burials of different ages and
types. The development of aerial photography, and more recently of contract
excavation, redresses another imbalance in the record. It no longer seems as if
the main concentrations of Bronze Age barrows were located on high ground,
as was once supposed. Rather, they are only the surviving part of a wider
distribution of monuments associated with major valleys. The remainder of
this distribution has been obliterated by cultivation. Such work also shows that
the great concentrations of burials long recognised in areas like Salisbury Plain,
the Dorset Ridgeway, or the Yorkshire Wolds have their counterparts in areas
where few earthworks survive. For example, many round barrows are located
along the rivers discharging into the Fenland (Evans and Knight 2000), whilst
the density of round barrows on the island of Thanet in southeast England is
similar to that in most parts of Wessex (Field 1998).
In both the Beaker period and the Earlier Bronze Age finds of high quality
metalwork are known well outside the areas which have usually been studied.
Among recent discoveries in England are a group of gold ornaments associated
with a barrow at Lockington in the East Midlands (Hughes 2000). There
is a gold vessel from an enigmatic monument at Ringlemere and a number
of other finds from Kent (Varndell and Needham 2002; Champion 2004).
Elsewhere, in the Midlands, two barrows, at Raunds and Gayhurst, are associated
with enormous collections of bones, suggesting that hundreds of animals
had been slaughtered and consumed in the course of feasts (Davis and Payne
1993; A. Chapman 2004). There is little to distinguish the most complex grave
assemblages at lowland monuments like those at West Heslerton in northeast
England from the burials in the better known cemeteries on the chalk
(Powelsland, Haughton, and Hanson 1986; Haughton and Powelsland 1999).
The real difference is that these monuments did not survive above ground and
had been protected from antiquarian activity.
There have perhaps been fewer new developments in Irish archaeology, but
this is because some areas are poorly suited to air photography. There has
been a marked increase in discoveries as large areas of topsoil are stripped in
the course of developer-funded excavations. Like the east coast of Scotland,
Ireland contained many flat cemeteries in which the dead were buried in stone
cists (Waddell 1990). Here barrows or cairns were less often built and the sites
of cremation pyres are occasionally identified in fieldwork. Some of the graves
recently discovered in Scotland are just as noteworthy as the better known
examples in the south of England. Indeed, one result of contract archaeology has been the recognition of burials in pits and flat graves even in barrowdominated
landscapes.
The contents of Earlier Bronze Age burials pose problems. It is obvious
that unusual artefacts were obtained from a distance and that local products
imitated exotic types. In the past this was explained in two ways. The first has
proved especially tenacious. True to the intellectual climate of the time, Stuart
Piggott’s classic account of the rich graves of the ‘Wessex Culture’ drew on
Continental parallels to postulate an invasion of southern England by an elite
from Brittany (Piggott 1938). This was based on the close similarities between
the grave goods in these two areas (Fig. 3.28). Although Piggott took a cautious
line in later life, Sabine Gerloff followed the same interpretation nearly forty
years afterwards (Gerloff 1975: 235–46). However, most of the pottery found
in association with exotic grave goods in Britain and Ireland is of insular forms,
suggesting that this approach may be misleading.
A similar relationship was proposed with the burials of the Aunjetitz Culture
with its focus on what is now central Germany (Gerloff 1975). It is true that,
like Wessex, this region contains some exceptionally rich burials. They were
associated with round barrows, but in fact most of the graves in this tradition
are found in flat cemeteries, and even more artefacts were deposited in hoards.
There are stylistic links between the metalwork from this region and Britain, but
these are equally apparent from the distribution of other objects. Amber beads
are distributed from Britain and Scandinavia as far as the Mediterranean, and those made of faience are found even more widely, with distinct concentrations
in Greece, Central Europe, southwest France, Britain, and Ireland (Sheridan
and Shortland 2004).
For a long time studies of the Earlier Bronze Age also supposed that Britain
and Irelandwere on the outer edge of ancient Europe and that its corewas in the
Mediterranean. This was taken as the ultimate source of inspiration for a number
of developments, from monument types to the design of Stonehenge. Some
of these links were implausible – British and Irish passage tombs are Neolithic
and cannot be compared with the tholos tombs of Greece; Stonehenge bears
no resemblance to Mycenaean architecture – but it may be that the entire
approach was wrong. Why should these islands have been dependencies of the
European mainland? It is ironic that the closest link between the study area and
the wider world should involve movement in the opposite direction. Amber
which had probably been imported from Scandinavia was used to make jewellery
inWessex, and a few of the finished artefacts were deposited in the shaft
graves at Mycenae (Sheridan, Kochman, and Aranauskas 2003).
A useful perspective on such long-distance contacts is provided by the faience
beads which were in use in Britain and Ireland between about 1900 and 1500
BC (ibid.). Might they provide more convincing evidence of international
trade? It was once considered that the Mycenaeans were exchanging southern
faience for northern amber. It is true that these finds were linked by a similar
technology, but now it is obvious that it was the technology that was adopted
in these different areas: the beads were not exported from a single source, and
the British examples made use of tin from southwest England. At the same
time, there must have been contacts from one end of the island to the other,
for a number of the artefacts with this distinctive composition were made in
the north of Scotland.
The same applies to the adoption of metallurgy, and here a core–periphery
model is equally inappropriate. Ireland contains gold, and some of the objects
produced there were exported to the Continent. Copper mining was established
in the southwest of the country from 2400 BC, and new sources were
exploited when Ross Island went out of use. Around 1900 BC other mines
were established inWales and parts of northern England and remained in operation
for about three hundred years, and in one case for even longer (Rohl
and Needham 1998). It is clear from metal analysis that Cornish tin was also
being exploited from an early stage. In Britain, bronze was used before it was
adopted in neighbouring regions of the European mainland.
It seems more consistent with the evidence to suggest that the concentrations
of richer burials in Europe are closely related to the proximity of metal sources.
That would explain the prominent position played by Britain, Ireland, and
northwest France, with communities in Wessex well placed to control the
cross-Channel movement of tin from southwest England and its distribution
across both these islands (Sherratt 1996). Similarly, the growth of the Aunjetitz Culture is surely related to the mining of Alpine copper and the growth of
fortified settlements whose inhabitants could have controlled the movement
of the metal (Shennan 1995). In each case similar processes seem to have been
at work, and so it is hardly surprising that communities in these different areas
formed alliances with one another.
A new study by Stuart Needham (2000) suggests an even more satisfactory
way of thinking about this evidence. Although he was specifically concerned
with relations between southern England and Brittany, his ideas have a wider
application. He accepts that there were close connections between individual
artefacts on either side of the Channel and that objects might have moved
in both directions. They were particularly important because they referred to
links with distant areas. They could also be made out of materials with unusual
physical properties, like amber, gold, and jet. The work of the anthropologist
Mary Helms (1998) is relevant here. She discusses the role played by knowledge
of distant places and unfamiliar practices. Travel may be a source of social power,
and the acquisition of exotic items can assume a cosmological significance. The
very fact that some of those connections were with remote places lends them
their special power. That may be why drawings of axeheads and daggers feature
on a number of stone monuments in Britain, including a series of burial cists
in the west of Scotland (Simpson and Thawley 1972).
There were significant changes over time. These are illustrated by the movement
of portable artefacts, but metal analysis provides a still more general picture
(Rohl and Needham 1998). After the discovery of Cornish tin the inhabitants
of southern Britain may have exported raw materials, both as native ores and
in the form of bronze, but by about 1650 BC the occupants of these islands
were making less use of local metals and the European mainland was becoming
an increasingly significant source of raw material. That was especially true in
the south. The sequence is consistent with studies of ancient seafaring. Before
the first exploitation of copper and tin, travel across open water may have been
dangerous and infrequent, and it is surely significant that it was then that the
earliest sewn-plank vessels seem to have developed. They remained important
from that time onwards, but they may not have been limited to a purely practical
role, for their remains have been found in the Humber Estuary not far from
a number of the richest burials in northern England (van de Noort 2003). Sea
travel may been a way of obtaining foreign valuables but could also have been
a method of winning prestige.
Not all the links were between these islands and the European mainland.
The best illustration of this point is the abundance of metalwork deposits in
northeast Scotland, a region which also includes the moulds for making axes.
What is extraordinary is that the copper seems to have been introduced from
a part of Ireland 750 km away, and tin from equally distant sources in western
England. Needham (2004) suggests that the two regions may have been linked
to one another, not just in economic terms but by the distinctive orientation of the local Scottish stone circles which are often directed towards the southwest
or south-southwest. These are the directions of the copper and tin sources,
respectively. The monuments may also have been aligned on the positions of
the moon or the sunset (Bradley 2005a: 109–11), but Needham’s argument
emphasises as well as anything else the peculiar character of such long distance
connections.
RELATIONS WITH THE DEAD
At various points in this chapter the text has referred to ‘single’ or ‘individual’
burials. The terms have been used interchangeably, but now it is necessary to
consider them in detail. They are imprecise, and yet they are impossible to
eliminate from archaeological writing.
On one level the idea of single burials was an artefact of early fieldwork
(Fig. 3.29). Antiquarians thought that round barrowswere memorials to important
individuals. Each mound would cover a grave containing their remains,
accompanied by a variety of artefacts (Ashbee 1960). The first excavators targeted
what was sometimes the earliest of a whole series of burials. If their finds
were to be studied as a coherent assemblage, it was important that they had
been deposited at the same time, and yet one grave might actually include
the remains of several people. They could have been buried simultaneously,
some bodies might be incomplete, or the grave might have been reopened
for the reception of other corpses or groups of bones. Renewed excavation of
these monuments shows that such features were often overlooked. In any case
there are too few graves to represent the entire population and it is clear that
isolated body parts continued to circulate, as they had during the Neolithic
period.
The term ‘single’ burial cannot refer to the deposition of one body per barrow,
as more recent excavations demonstrate that this was an important feature
in some phases and less significant in others, but not every grave contained
a mixture of human remains in the manner of an earlier long barrow. Most
intact bodies were accompanied by some selection of artefacts. That is why
those objects seemed to be associated with specific ‘individuals’.
There are problems with using that term. In modern Western thought the
‘individual’ carries specific connotations relating to identity and agency (Br¨uck
2004). Those ideas developed during the Enlightenment, and there are societies
in which this approach is not appropriate: people are not thought of
as autonomous entities but are defined in terms of their relationships with
others. The idea of an individual burial might be equally misleading because
the portrayal of the deceased created during the funeral was composed by the
mourners who placed particular objects in the grave. It is not clear whether
these had been the property of the dead person, and certain items, including
gold ornaments, may have been made for the occasion (Coles and Taylor 1971). Some artefacts were so fragmentary that they might have circulated over several
generations (Woodward 2002), yet the presence of flowers in Scottish graves
must be explained as gifts (Tipping 1994b). The corpse could not determine
how it was displayed in death; the image encountered by the modern archaeologist
reflects the relationships between the deceased and those who undertook
the burial.
Perhaps the grave was an arena for negotiating relationships between the
living and the dead, but that is not how the evidence has been understood
until now, for more emphasis has been placed on the contrast between ‘richer’
and ‘poorer’ graves (Pierpoint 1980: chapter 9). The distinction between adult
and child burials has been studied and so has that between men and women,
although gender distinctionswere inferred on the basis of grave goods as barrow
diggers rarely retained the excavated bones. In each case the objective has been
the same: to use the contents of the burials to reconstruct social organisation.
From the work of William Stukeley in the eighteenth century, there have
been two main trends. Archaeologists have employed the grave assemblage
to infer social status and to identify the activities associated with particular
people in the past. Just as Stukeley identified Kings’ Barrows and Druids’
Barrows on Salisbury Plain, recent writers discussed the graves of warriors,
archers, leather workers, smiths, and shamans (Case 1977; Woodward 2000:
119–22). This approach makes many assumptions. The dead were equipped
by those who survived them, and the deceased might be accompanied by the
material associated with more than one of these roles. Thus the man who
has become known as the ‘Amesbury Archer’ might have been a hunter, but he was portrayed also as a warrior, a flint worker, and a smith, for his grave
contained several sets of objects that might otherwise have been distributed
between different burials (Fitzpatrick 2003).
Of course there were contrasts between the contents of these graves, but
the British and Irish evidence seems surprisingly uniform. There are a few
exceptional deposits, but the range of variation is comparatively limited considering
how many burials have been excavated. The quantity of metal items
is hardly impressive compared with the contents of Aunjetitz hoards like those
at Dieskau (von Brunn 1959), and it is overshadowed by the large number
of objects consigned to rivers during the Later Bronze Age (Bradley 1998a).
Some burials are accompanied by a locally made pottery vessel, and even the
necklaces of jet, faience, and amber might be composites made out of material
which was already worn and broken (Woodward 2002). There were episodes
of lavish consumption and display, but these were mainly concerned with the
building of cremation pyres and with funeral feasts ( J. Barrett 1994: chapter 5;
Davis and Payne 1993). Few burial mounds attained the proportions of the
larger Neolithic monuments, and some Bronze Age barrows developed incrementally
over a long period of time. The presence of exotic items in the graves
illustrates the importance of connections with distant areas, but the portable
wealth of this period seems rather insignificant compared with what is known
from Central Europe. There seems little justification for postulating a rigid
social hierarchy.
The same point is illustrated by the distribution of mortuary monuments.
Passage tomb cemeteries and the largest henges had been widely spaced across
the landscape, as if they provided focal points for a large area around them.
Earlier Bronze Age cemeteries occur at much smaller intervals, although they
are sometimes close to ceremonial monuments. They are so common, and
their contents are so consistent, that they are best regarded as the burial places
of local communities (H. S. Green 1974). If higher-status monuments are to be
identified, then they are more likely to be those associated with long-established
sites like Stonehenge (Exon et al. 2000). Indeed, Rosamund Cleal (2005) has
pointed out that the burials close to that monument are significantly richer
than those near the equally impressive henge at Avebury.
In any case burial rites changed over time. They have been studied by Paul
Garwood, whose work makes use of a large number of radiocarbon dates
(Garwood in press). These are mainly from England andWales, but it is possible
to suggest some points of comparison with Scotland and Ireland.
The first Beaker round barrows were quite simple and have already been
discussed. Between about 2150 and 1850 BC burial mounds became much
more diverse, and many structures underwent a sequence of transformations.
In particular, barrows might be associated with a larger number of deposits
and with a wider variety of people. They included a greater proportion of
women’s graves and those of children. There was also a wider range of variation in the treatment of the body, and cremation became more important at this
time.
This contrasts with the period between about 1850 and 1500 BC. It saw the
development of the major cemeteries that are often supposed to characterise the
whole of the Earlier Bronze Age. Although the mounds were often large and
might assume specialised forms, they were frequently constructed in a single
phase, and the central graves may contain only one body. The linear barrow
cemeteries of the Wessex downland seem to date from this period and are
unusual in being organised in such a formal manner. This is not found widely
in other areas, although it is sometimes considered as the norm. Cremation
was widely practised, and certain of the individual mounds contain the richest
grave groups of this period.
A basic distinction is between the use of an individual mound for a whole
series of burials (such as a cemetery barrow/cairn), and the development of
barrow cemeteries, which are groups of barrows, each containing one or
more interments. Both have to be considered here. A good starting point
is with the mortuary monuments of northeast England, where any mound
may include several graves and where each grave could contain more than one
body (Petersen 1972). Such cemetery mounds occur very widely. In the south,
for instance, well-excavated and well-preserved monuments inWessex contain
the remains of roughly fifteen people, while the cairn at Bedd Branwen on
Anglesey included a similar number (Lynch 1971). Mounds in Ireland include
as many as nineteen, (O’Sullivan 2005: 169–70), although the median is six
(Waddell 1990). There was considerable variation from one region to another.
The round barrows of central southern England contain an unusual variety of
artefacts, but these sites cannot typify the range of mortuary practices found in
these islands. English barrows are commonly organised into small groups, but
in other regions, most obviously in North Wales and Ireland, such mounds
are often isolated (Lynch, Aldhouse-Green, and Davies 2000: 121–8; Waddell
1990; J. Eogan 2004).
There were other ways of commemorating people without constructing a
monument. In the East Anglian fens bodies might be placed in bogs or pools
together with Earlier Bronze Age artefacts (Healy and Housley 1992). Their
remains are also found in caves in northern England (Barnatt and Edmonds
2002), and across large parts of Ireland, areas of western Scotland, and along
the coast from the River Tees to Aberdeenshire there were cemeteries of flat
graves, most of which lacked any mound or cairn (Waddell 1990; Cowie and
Shepherd 2003). In the west of Ireland they are much less common, and here
wedge tombs were used instead. Like Neolithic monuments, they include a
mixture of burnt and unburnt bones (O’Brien 1999). Such variation is by
no means unusual. Apart from the many people whose remains have left no
trace, the deposits found in graves include inhumations, cremations, and smaller
groups of disarticulated bones. Although inhumation was favoured at an earlier stage than cremation, the use of both rites overlapped and deposits of each kind
can even be found in the same grave.
There is a way of thinking about such evidence which combines several of
the points made so far. Not all the graves thatwere once described as ‘individual’
burials contain the remains of one person, although this is frequently the case.
Male burials adopt a different body position from those of women (Tuckwell
1975; Greig et al. 1989: 79–80), but within these deposits artefacts are associated
with particular people and are often placed in specific locations in relation
to the corpse. At the same time, other burials are found on the same sites.
This discussion has highlighted the way in which the image of the dead was
manipulated by those conducting the funeral, for they were responsible for the
selection of the artefacts that were to be buried. Perhaps the organisation of
the cemeteries also expressed the relationships amongst the dead themselves.
In 1972 Petersen criticised the idea that Earlier Bronze Age burials were
organised around single graves. He studied the records of a large number
of excavations, mostly in northern England. It was clear that many graves
contained the remains of several people. Following the chronology that was
accepted at the time, this could be compared with the Neolithic practice of
collective burial. It was less obvious how this evidence was related to other
regional traditions during the same period. That subject was taken up by Koji
Mizoguchi nearly twenty years later (Mizoguchi 1993). Like Petersen, he paid
particular attention to the round barrows of northeast England. He observed
that there were certain consistent relationships between successive burials in the
same graves. Often the original excavation had been reopened. That means
that its precise position was known, but when it happened other features of the
original burial were taken into account. Where the first burial was that of an
adult man, the next interment was generally that of a woman or a young person
who would be laid out either on the same alignment as the previous burial or
at right angles to it. That process might be repeated. At the same time where
inhumations and cremations were found together, the unburnt corpse was the
first one committed to the ground, and the cremation was generally placed
in the grave either in a subsidiary position or at a higher level in its filling.
At times it is clear that relics were removed. Not all the earlier skeletons are
complete, and it is possible that artefacts were also taken away. For example,
careful excavation showed that the central grave at Gayhurst Quarry had been
recut on five occasions (A. Chapman 2004).
The same processes apply to the siting of different graves. They may be juxtaposed,
they can be aligned on one another, or their positions may respect one
another with such accuracy that this must have been intended. It would have
required detailed knowledge, and such information even extended to the configuration
of the corpse. This is clear from Jonathan Last’s analysis of the burials
from a round barrow at Barnack in eastern England (Last 1998). Another example
of these patterns is the cemetery at Keenoge in the east of Ireland which
includes a series of graves, most of them in pits or cists; the inhumations were often accompanied by cremations (Mount 1997). With few exceptions, the
burials followed approximately the same alignment, but the bodies in adjacent
graves tended to be organised in pairs, one with its head to the east and the
other laid out in exactly the opposite direction. The only divergence from
this pattern concerned a few graves with a northsouth axis, but the inhumations
were paired in a similar manner (Fig. 3.30). It resembles the process
described by Last, but this time the principle extended to a flat cemetery.
Those relationshipswere expressed on an intimate scale andwere presumably
played out over a short period of time, so that newer burials might be related to those of the recently dead. The same principle applies to people who had
died long before, but it is more difficult to see it at work. Perhaps the best
evidence comes from earthwork monuments. For a long time it was assumed
that these could be reduced to a few distinctive forms, so that there would
be a consistent relationship between the appearance of any particular barrow
and the character of the principal burial or burials (Ashbee 1960). That was
unduly optimistic as it is clear that, except in the late group of single-period
barrows discussed by Garwood (in press), the shape of any individual mound
is no guide to the deposits associated with it. Impressive earthworks may be
found with sparsely furnished graves – or with none at all – and more elaborate
graves may occur beneath smaller mounds or outside them altogether. That is
because monuments were rebuilt or extended over time, so that in principle
they might begin with an unmarked grave and pass through a series of structural
stages during which they changed from one ‘type’ of monument to another
( J. Barrett 1994: chapters 2 and 5). It is misleading to pay too much attention to
their outward appearance, for it merely marks the point at which that process
stopped. Sometimes the remains of simpler constructions are buried beneath
the outer mantle of a larger earthwork.
A similar process extended to the relationships between the mounds in
the same group, although Paul Garwood’s study suggests that the complex
linear cemeteries which are so common in Wessex date from the end of this
period and include some barrows with only one grave. They show a number
of relationships. Although the separate monuments could be some distance
apart, it is quite common for them to be orientated on certain focal points
or for the monuments to be arranged in lines so that every barrow had its
neighbour or neighbours (Fig. 3.31). In that way one earthwork might refer to
the position of another, and sometimes they were even combined as formerly
discrete structures coalesced. Just as mounds could be linked together, they
could also be kept separate, and some of the larger cemeteries can be subdivided
into several clusters or rows of mounds. Recent excavation suggests that such
principles extended to the spaces in between these structures. This process was
followed at Monkton in Kent, where the cemetery includes a post alignment
(Bennett andWilliams 1997). At Radley in the Thames Valley, there were two
parallel lines of round barrows, but there were also some flat graves, including a
rowof urned cremations which followed the long axis of the cemetery (Barclay
and Halpin 1999: 128–33). This is one of the barrow groups whose siting was
influenced by the presence of a Neolithic monument, and it reinforces the
suggestion that the organisation of these complexes was one way of relating
those who had recently died to the dead of earlier generations. Garwood (in
press) suggests that some of the linear cemeteries were also directed towards
the setting sun, thus linking the fortunes of the dead to the annual cycle
of the seasons. That may have been the case at Snail Down on the Wessex
chalk, where two lines of barrows were established between about 1800 and 1500 BC.  One may have been directed towards
the sunrise, and the other towards the sunset. This formal pattern seems to
have been imposed on a scattered distribution of older mounds, two of which
had been built over the remains of a settlement.
This discussion began with the suggestion that the identities of the dead
were constituted by their relationships with the living and that these were
expressed through the choice of materials placed in the grave. That may well
have been true, but it seems at least as important to recognise that not all the
Earlier Bronze Age cemeteries contain the burials of ‘individuals’. Rather, they
are organised around the relationships between the dead themselves: not only
those who had just died but also the burials of their forbears. The same idea
is expressed by the circulation of heirlooms or even human bones, and their
eventual deposition in the grave. The result is rather like a genealogy in which
the placing of the grave, its orientation, and its contents locate any particular
person within a wider network of social relationships extending into the past.
That might be expressed quite informally in the organisation of a flat cemetery
or a single mound, but it could also be set out on a massive scale through the
development of a large barrow cemetery.
That is where Garwood’s analysis is so helpful, for it draws attention to
important changes in the ways in which these processes operated over time.
Between 2150 and 1850 BC round barrows were built and reconstructed on a
large scale. These monuments were often located in relation to existing features
of the landscape, including Beaker mounds, but the most important way of
signifying relationships between the past and the presentwas by reopening older graves and adding new deposits. In the same way, where the first barrows might
have covered only one grave, now such monuments included a wider variety
of burials whose positions within the mound seems to have been determined
by real or imagined relationships with the dead (Barrett 1994: chapter 5).
The same concerns were important in the last major phase of barrow building,
between 1850 and 1500 BC, but it was the spatial relationship between
different monuments that assumed more importance. That is particularly obvious
in Wessex where linear cemeteries developed, many of them built out
of specialised forms of mound that were constructed in a single phase. They
include the burials of individual corpses associated with a rich array of grave
goods, but this time they are located in relation to a more complex history in
which the placing of any one mound was related to the positions of all the
others. As Garwood (in press) points out, these lines of monuments could be
read as a sequence of individual graves, and the same argument has been pursued
by Mizoguchi (1992). Now those relationships were displayed to everyone,
and the mounds had even greater authority because they were orientated
not only on the remains of the past but occasionally on the position of
the sun.
Genealogies provide one way of codifying relationships, but there were
other, less tangible connections that were displayed through the use of artefacts.
One is the relationship with previous generations epitomised by the
transmission of relics. The other is the pattern of long-distance alliances that
linked people in Britain and Ireland with those in Continental Europe. This
was shown by the use of unusual or exotic artefacts, and it may be no accident
that they played a particularly prominent part in Earlier Bronze Age Wessex,
the very region in which formal barrow cemeteries found their fullest expression
towards the end of this period. Rather different practices were followed
in other areas, and it remains a priority for future research to explore the wider
significance of these distinctions.
This account has been concerned with relationships, but there is one which
has not been considered so far. That is the relationship between the dead and
the places in the landscape where they were buried. There are three points to
make here. The first is that the distribution of standing monuments is incomplete.
It is all too easy to suppose that round barrows were a major feature of
the uplands, but even the surviving monuments overlook lower ground, and
aerial photography has shown that they are often located in valleys and beside
watercourses (Field 1998). The impressive earthwork barrows on the hills of
Wessex and the YorkshireWolds have created a misleading impression for they
were accompanied by similar monuments that have been ploughed out. There
is no real difference between these regions and the river terraces where many
other burials are found.
Having said that, it is probably true that the basic pattern was influenced by
the continuing attraction of a variety of monuments, particularly henges and cursuses, which were already established during the Neolithic period. These
are associated with mortuary monuments of later date, some of which include
particularly striking collections of grave goods. This is obviously the case in
southern England, but it is equally obvious around the stone circles of Orkney
(Sheridan, Kochman, and Aranauskas 2003). Flat cemeteries developed in the
vicinity of henge monuments like Broomend of Crichie in northeast Scotland
( J. Ritchie 1920), and in Ireland one of the largest groups of Bronze Age burials
was in the mound covering the decorated passage tomb at Tara (Newman
1997a: 147–8; O’Sullivan 2005: Chapter 4). Just as wedge tombs were used
during this period, older megaliths in northern and western Scotland were
selected for the burial of the dead before their chambers were sealed (Bradley
2000a: 221–4).
Lastly, there are a number of barrows and cairns which fall outside these
trends. These are prominent monuments that were built on exceptionally high
ground, especially hills and mountains in north Wales and northeast England.
They are not always easy to identify from below, but they command extensive
views and can be seen from one another (Pierpoint 1980: 266–70; Gibson
2004b: 156–9). They raise important questions about the place of Earlier Bronze
Age monuments in the wider pattern of settlement and may have commemorated
people with a special position in the community (Fig. 3.33).
MONUMENTS AND THE SETTLEMENT PATTERN 2000–1500 BC
It is frustrating that settlements can be difficult to find and that the candidates
which are suggested are not always easy to date. For that reason more attention
has been paid to the distribution of monuments, as this can shed some light
on where and how people lived. But much of the difficulty arises because it is
commonly supposed that certain areas were used as ‘ritual landscapes’ and given
over to the commemoration of the dead. That cannot be taken for granted,
and the argument needs to be substantiated.
In fact there is some evidence for the relationship between cemeteries and
living areas, but it takes different forms in different regions. In the Upper
Thames, it seems as if settlement focused on the lower river terraces, whilst the
largest barrows were on slightly higher ground (Bradley et al. 1996: 18–19).
Around Raunds in the English midlands, that relationship was reversed (Healy
and Harding in press), but in each case the two zoneswere not far apart. Close to
Stonehenge a third pattern can be identified (Peters 2000). Here the smallest
barrows, which were associated with poorly furnished graves, were usually
located on the lower ground and occur in areas where there is evidence for the
clearance of surface stones – presumably such regions were being ploughed.
The larger barrows were located in more prominent positions on the higher
ground. These were associated with a more complex series of graves and with
a wider range of funeral gifts. Just as important, these monuments made use of
turf, suggesting that they had been built on grazing land. In northern Scotland the evidence takes yet another form. Here it has been shown by field walking
that the artefact scatters left by occupation sites were set apart from Neolithic
chambered tombs but located near to the Earlier Bronze Age Clava Cairns
(Phillips and Watson 2000).
In other cases barrows were directly superimposed on settlements. This was
particularly common in East Anglia (Gibson 1982: 27–48), but it is found
widely. Indeed, at a few sites, including the Brenig in North Wales, mortuary
monuments were built over the remains of older dwellings (Lynch 1993:
chapter 13). That may have happened in other cases, although the structural
evidence is not always convincing. On Dartmoor, it has been suggested that
collapsed round houses were reused as mortuary cairns (Butler 1997: 137–8),
but this idea needs to be checked by excavation.
Where barrows and settlement areas were further apart there might still be
a direct connection between them. At Roughan Hill in the west of Ireland
a series of wedge tombs were constructed in between a number of enclosed
settlements (Fig. 3.34). These were interspersed with an irregular system of
field walls. Excavation has shown that they were associated with round houses
and with Beaker pottery. It seems likely that activity continued into the second
millennium BC (C. Jones 1998). In northeast England the main areas with
evidence of Earlier Bronze Age land use are overlooked by chains of barrows
and cairns running across the higher ground. These enclose the valleys which
were suitable for settlement, and Don Spratt argued that they marked the outer
limits of small estates or territories (1993: 116–20).
There are many reasons why domestic sites are so difficult to find. Excavation
on the Scottish island of Arran showed that the sites of Earlier Bronze Age houses had been ploughed when they went out of use (Barber 1997).
Another problem arises in field survey, for in upland areas the remains of
stone-built houses resemble the enclosures known as ring cairns which were
built in increasing numbers during this phase. They played various roles as
ceremonial sites, cremation cemeteries, and pyres, but on the ground they can
be hard to distinguish from the remains of houses or settlements. The link may
have been intended, for there are indications that abandoned buildings were
enclosed within a mantle of rubble so that they resembled monuments of this
type (Barnatt, Bevan, and Edmonds 2002). A similar problem applies to many
of the small cairns found in the uplands, a number of which certainly date from
this period. They seem to have accumulated around prominent boulders or
outcrops which would have been difficult to move when the land was tilled.
For that reason such features are interpreted as clearance cairns and have been
taken as evidence of agricultural settlement. On the other hand, excavation has
demonstrated that a number of them covered human burials or formal deposits
of artefacts. Does this mean that they had been misinterpreted? An interesting
way of thinking about the problem was suggested by Robert Johnston
(2000), who argues that the resemblance between mortuary monuments and
field cairns was no accident, because the dead were integrated into the working
of the land.
Timber monuments illustrate a similar problem. Just as the timber circles
inside henges look like enlarged versions of ordinary domestic dwellings, a
number of graves were ringed by circles of posts or stakes before barrows
were built on the same sites. They raise similar problems to the ring cairns,
for a few examples could be construed as the remains of dwellings, especially
where such structures are associated with hearths or concentrations of artefacts
(Gibson 1982: 27–48). Perhaps these features were considered as the houses of
the dead, and their forms referred to those of domestic buildings. Again there
is a significant overlap. Nowhere is this more obvious than at West Whittlesey
in the English Fenland, where a row of three circular structures was identified
in excavation (Knight 2000). Two of them were round barrows, whilst the
third was an Earlier Bronze Age henge enclosing a circle of posts. Fifty metres
away there was a settlement of the same date which included the remains of
a small round house. In such cases the cross-reference between these different
types must have been intentional.
As in the Beaker phase, some occupation sites survive because they are
deeply buried.Well-preserved dwellings have certainly been found among the
coastal dunes in theWestern Isles, for example on South Uist (Parker Pearson,
Sharples, and Symonds 2004: 43–52). These are exceptionally fertile areas, and
prehistoric ard marks are often recorded there. Earlier Bronze Age settlements
have also been found on the edge of the English Fenland, notably at West
Row Fen, Mildenhall (Martin and Murphy 1988), and others are identified in
areas that had not been settled before. In some cases their remains have been
preserved because such areas were not occupied for long. The distribution of barrows, cairns, and other monuments suggests that the Later Neolithic
expansion of settlement accelerated during this phase. That is certainly supported
by pollen analysis. The distribution of such monuments is much more
extensive than that of their predecessors and includes large areas of lowland
England which are currently covered by heathland, as well as tracts of moorland
in upland regions. Among these areas were the New Forest, the Weald,
the Breckland, the Pennines, and the North York Moors. Human activity
extended into a wider range of environments, but clearance and exploitation
caused significant changes to the soils which often became acidic and poorly
drained (Bradley 1978: 113–14). These examples have been taken from English
archaeology, but similar trends have been identified by field survey in other
areas. The difficulty is that certain types of monument have been investigated at the expense of others, so that it is obvious that barrows, cairns, and stone
circles were being built there during this period, whilst the evidence from
domestic sites has attracted less attention. That is not to deny that certain excavated
settlements do date from this phase, but others have been assigned to
the Earlier Bronze Age on the premise that the high ground would have been
uninhabitable during later periods because of climate change (Burgess 1985,
1992). That is controversial (Young and Simmons 1997; Tipping 2002), and
some of the field evidence that had been confidently assigned to this phase
dates from a later period.
The settlement evidence from the uplands often shows the same sequence.
The stone-built houses which are visible on the surface were not necessarily
the first to be built, and excavation has demonstrated that before the landscape
was cleared and surface rock was exposed domestic buildings might have been
made of wood (Fleming 1988: chapter 6). Not many have been excavated,
but they do seem to have been more substantial than the Beaker dwellings
considered earlier in this chapter. Unfortunately, where such structures were
not replaced in stone they have been difficult to find, although their positions
can be located where their floors were levelled into sloping ground, leaving
a distinctive circular platform behind. Otherwise they may be detected by
open patches within a wider distribution of boulders and small monuments. A
number of the houses and cairns were linked by low walls or banks of rubble
which define a series of irregular areas suitable for cultivation. Sometimes small
enclosures have been tacked onto one another as increasingly large areas were
cleared. They can incorporate clearance cairns, whilst others seem to be scattered
across the surrounding area. There are excavated fields in the Burren in
the west of Ireland (C. Jones 1998), in Shetland (Whittle 1986), in the Scottish
Highlands (McCullagh and Tipping 1998), and on the island of Arran (Barber
1997), but they seem to survive as surface remains in most upland regions.
Apart from the remarkable evidence from Dartmoor, which is discussed
in Chapter Four, there is little sign of more extensive or regular systems.
In a recent article Simon Timberlake (2001) suggests that the land use in
the uplands provides the context for early copper mining in Wales and the
west of Ireland. This was sometimes undertaken in remote areas that were
perhaps being visited for the first time. These sites were not obviously related
to nearby settlements or monuments and may have been used by small numbers
of people on a seasonal basis. The output of the mines was very limited and for
the most part their period of use was restricted to the Earlier Bronze Age. Like
comparable sites in northern England, they seem to have operated between
approximately 1800 and 1500 BC, after which they were abandoned. The one
exception was at Great Orme Head in NorthWales (Dutton and Fasham 1994;
Lewis 1998; Timberlake 2002). This was the site of the largest group of mines
in Britain or Ireland, and here activity continued into the Later Bronze Age.
Upland areas also contain some specialised monuments (Fig. 3.35). Space
does not permit the enumeration of all these separate types, and in any case they underwent constant modification. Like the round barrows considered
earlier, they could change their outward appearance from one phase to the
next, so that their surface remains provide little indication of how individual
monuments had developed over time (Bradley 1998b: chapter 9).
The great majority of these are arrangements of standing stones, but they
could also be combined with the walled enclosures known as ring cairns and
even with the Clava group of passage tombs which were perhaps the northern
Scottish counterparts of the late wedge tombs in Ireland (Bradley 2000a). In
addition to these varieties of circular monuments, there were also stone alignments
which may have been miniature versions of the longer avenues associated
with henges (Burl 1993). Most of these forms could appear in combination,
so that stone circles might enclose ring cairns and passage tombs, whilst stone
rows could run up to round barrows, cairns, or other settings of monoliths.
Some of these structures have well-defined local distributions. For example,
the Clava Cairns, which consist of ring cairns and passage tombs set within
graded stone circles, are confined to the inner Moray Firth (Bradley 2000a),
whilst in the neighbouring region of northeast Scotland similar rings of uprights
contain cairns but also include a massive horizontal block; these are known
as ‘recumbent’ stone circles (Bradley 2005a). In the same way, in eastern and
central Scotland there are small rectangular stone settings (‘four posters’) which
seem to include low cairns associated with cremation burials (Burl 1988). Local
traditions can be identified in other areas, too. At a site like Beaghmore in
northern Ireland there are circular settings of small boulders laid out in complex
geometrical patterns. They include numerous concentric rings but also feature
a series of straight alignments radiating out from a central cist (Pilcher 1969;
Foley and MacDonagh 1998). Again they can be found with other kinds of
structure, and these particular monuments are sometimes linked to stone rows.
It is easy to lose direction amidst so many specialised constructions, but
many of them can be interpreted as local manifestations of a series of quite
simple ideas that were widely shared across the two islands. Like their Neolithic
predecessors, the circles were often graded in height towards the south or
west, and sometimes they made effective use of differently coloured stones. At
Balnuaran of Clava, for example, two of the passage tombs were aligned on
the midwinter sunset (Bradley 2000a: 122–9). The stones employed on that
side of the monuments were often red and absorbed the light. In the opposite
direction, towards the rising sun at midsummer, the rocks were grey and white
and frequently contained inclusions of quartz which would reflect the light.
How were these monuments related to the mortuary rituals considered
earlier? They are often found near to Earlier Bronze Age barrows and cairns, and
in some cases they contain human cremations, which are presumably those of
people who could not be buried in a more formal manner. Other monuments
include the remains of pyres, but in this case the bones were usually removed
and deposited somewhere else. This may have happened during a secondary phase, and a common sequence is for an open enclosure, defined by a ring
of monoliths, to provide the site for a later mound or cairn (Bradley 1998b:
chapter 9). That might impede access to the interior, and on some sites the
earlier building was almost completely buried beneath the new construction
(G. Ritchie 1974). In such cases it is tempting to suggest that monuments
which could have been used by a large number of people were taken over as
the burial places of selected individuals. On the other hand, recent fieldwork
shows that was not the only way in which such sites developed. In north and
northeast Scotland it is clear that cairns, enclosures, and platforms might all
be built before the stone circles on the same sites. In that case it seems likely
that the erection of the ring of monoliths enclosed the older monument and
brought its use to an end (Bradley 2005a: 105–6, 112–16). That is similar to
what had happened with the earthworks of henges.
The relationship between such monuments and settlements changes from
one region to another. The Clava Cairns, for example, were built over, or
near to, occupation sites, and yet the stone circles of northeast Scotland, with
which they have much in common, were often placed in more conspicuous
positions on the outer edge of the settled land (Bradley 2005a: 108–9). The
stone settings at Beaghmore in the north of Ireland may even have been built
amidst the remains of an older field system (Pilcher 1969). The distinctive
stone rows of southwest England introduce another variation, for they seem
to run upslope from the areas with evidence of habitation towards mortuary
monuments located on the higher ground (Barnatt 1998: 99–102).
The monuments are sometimes embellished with the pecked motifs known
as cup marks, and many of them contain significant quantities of quartz (Burl
1981). Often they provide evidence of burning in the formof pits that had been
filled with charcoal (Lynch 1979). Some of these structures could have been
orientated towards the setting sun, but others may have been more directly
associated with the moon (Ruggles 1999). Such associations connect a series
of monuments of very different forms: the stone circles of northeast Scotland,
the Clava Cairns, four-posters, wedge tombs, and some of the stone alignments
found on the west coast of Scotland and in northern Ireland. These geographical
links are reinforced by the movement of copper and, to some extent, by
pottery styles.
Nearly all these elements refer back to Neolithic practice, although they
are often associated with Beakers and with Earlier Bronze Age ceramics. The
resemblance is not limited to their distinctive forms, as they illustrate the association
between stone monuments and the remains of the dead that was highlighted
by Parker Pearson and Ramilsonina (1998). They add some new elements,
but the structures themselves are smaller and occupy less conspicuous
positions in the landscape. They also illustrate an increasing concern with the
moon rather than the sun. Many provide evidence for the use of fire, and it is
possible that certain of these places were used at night (Bradley 2005a: 111–12). AT THE LIMITS
A number of separate issues have been discussed here and need to be brought
together now. They will be treated in greater detail in Chapter Four. Each
concerns the distinctive developments of the Earlier Bronze Age and suggests
that they were reaching their limits towards the middle of the second millennium
BC.
This period between 2000 and 1500 BC is usually thought of in terms of its
distinctive burials and the long-distance connections that they seem to illustrate,
but in some ways both were changing by this time. It has been customary to
think of Britain and Ireland as the periphery of Europe, but as long as local
materials were being used that was quite untrue – bronze was produced in
Britain at an unusually early date and Irish gold was exported. By 1500 BC,
however, it appears that most of the British and Irish copper mines had gone
out of use, although activity continued at Great Orme Head (Northover 1982;
Timberlake 2002). These developments have never been explained, but they
form part of a more general crisis in the distribution of metalwork which
saw quite rapid oscillations between the use of insular copper and a greater
dependence on Continental sources of supply. Those changes will be discussed
in the chapter which follows, but they suggest that access to metal artefacts
perhaps became more difficult.
There were also changes in the forms of barrow cemeteries. It has long
been recognised that great concentrations of mounds were built close to the
monuments of the Neolithic period and that a few of the burials associated with
them contain elaborate or exotic objects. What has not been so clear is that great
linear cemeteries, such as those around Stonehenge, were a late development
(Garwood in press). A landscape that was already permeated with images from
the past took on yet another layer of significance (Exon et al. 2000). It may
be no accident that these formal arrangements of monuments were among
the last to be built. The growing importance of cremation may be important
too, for it required a greater consumption of human energy than inhumation
burial. The barrow cemeteries that were built towards the end of the Earlier
Bronze Age seem almost excessively elaborate, and the sheer complexity of
references enshrined in these mounds and graves may no longer have had the
desired effect. Perhaps they stopped being an effective method of signally social
relationships. The same is true of the layout of Stonehenge, which may have
been modified at this time (Cleal,Walker, and Montague 1995: 256–65). Those
ideas might have been better expressed in a different medium. Chapter Four
documents that transformation.
It is easier to identify another way in which the existing system approached
its limits. Many monuments were built during the Later Neolithic and Earlier
Bronze Age periods, but there is little to suggest that their construction
depended on a major intensification of land use. Instead food production took place over an increasingly large area (Bradley 1991). That is indicated by the
distribution of artefacts, settlements, and monuments, and it is shown also by
pollen analysis. The problem is that the expansion took in regions that could
not sustain a long period of settlement. Theywere adversely affected by changes
to the status of the soils, and some of them became increasingly waterlogged.
Although there are certain exceptions, large parts of lowland Britain which
were first colonised during the earlier second millennium BC had been abandoned
by the Later Bronze Age. The same was probably true in the uplands,
and in each case its consequence was the same. A system which had been
maintained by a process of continuous expansion was becoming increasingly
vulnerable. Long before the first signs of climatic change, that process came to
a halt.
Those separate factors – the over-exploitation of marginal land, the excessive
elaboration of mortuary rituals, and changes in the supply of metals – are not
directly linked to one another, yet all three are tendencies that have been
identified in earlier parts of this account. In every case matters were reaching a
crisis at about the same time. Taken in combination, they resulted in the single
greatest change in these islands since the adoption of agriculture. That striking
transformation is the subject of Chapter Four.