. 4
( 7)


Hosler (1993) has discussed the importance of cosmological ideas related
to creativity and to solar and lunar deities in in¬‚uencing the production
technologies in historic west Mexico. She notes the use of alloying to
create shimmering colours, and the signi¬cance of the kinds of objects
made of metal “ especially bells and rattles, which are associated with the
creative power of thunder and rain. Lechtman (1984), too, has noted
that in order to produce metals of certain colours, the Peruvian Mochica
developed a complex set of alloying techniques. These techniques were
developed so that the critical ingredient, usually gold or silver, was incor-
porated within the body of the metal. This was because of the belief that
the ˜essence of the object, must also be inside it. The object is not that
object unless it contains within it the essential quality™ (Lechtman 1984,
30). Saunders (1999, 246“7) underlines the culturally speci¬c nature of
these metallurgical concepts in his discussion of the exchange of precious
metals between native Americans and Europeans. He discusses the way
in which gold was impregnated with copper amongst a number of peo-
ples in historic Mesoamerica. Due to their luminosity, both metals were
related to the power of the sun and each metal was considered to have
been embellished by this alloying process. However, alloyed gold caused
considerable consternation to Spanish traders, who considered it to be
adulterated or impure. This example forces us to re¬‚ect on the cultural
speci¬city of technologies and provides an important warning against the
simple rationalist description of technological histories. We need to be
aware of the likelihood of culturally speci¬c technologies in both his-
toric contexts, such as those discussed above, and prehistoric contexts.
Indeed Hamilton (1991) has successfully demonstrated the speci¬city
A biography of things 93

of differing traditions of copper metallurgy in prehistoric Europe using
proton-induced x-ray emission spectroscopy (PIXE).

The social organisation of production
If we take on board the notion that technologies are humanly related
through webs of interaction, as mediums by which the social, symbolic
and material are interwoven and used as a means of de¬ning or main-
taining different social identities, then we need to consider how technolo-
gies are produced and reproduced. How are the webs of interaction be-
tween the social and material created? Here we need to consider the social
relations involved in production. When considering this issue, we again
need to be aware of the scale of analysis at which we operate, since it is at
the local level of the interaction between people and things in technical
production that microscale analysis comes into its own. As Dobres and
Hoffman (1994, 213) note, a microscale approach enables us to ˜model
the dynamic social processes involved in ongoing, day-to-day technologi-
cal endeavours, and to consider the differential participation of the actors
and groups involved™.
In the domain of ceramics studies a number of innovative archaeologi-
cal studies were undertaken to examine the proposition that the patterns
observed in the production of material culture relate to particular sets of
social relations (Deetz 1968; Hill 1970; Longacre 1981, 1985). These
studies were concerned with examining how the production of ceramics
may be related to particular sections of a social group. Speci¬c character-
istics of ceramics were mapped both spatially and temporally, enabling
a concomitant mapping of social groupings. For instance, Deetz (1968)
was concerned to suggest that material culture quanti¬ed in such a way
enabled particular kinship con¬gurations to be mapped spatially, while
the rules of descent associated with such kinship groups could also be
determined through the study of the patterning of material culture char-
acteristics through time. This approach was realised in its most detailed
form in Hill™s study of Broken K Pueblo, Arizona (1970). Here, through
the quanti¬cation of speci¬c design features on pottery and the observa-
tion of the patterning of these design features throughout the settlement,
Hill asserted that the communal organisation of pottery production was
related to two moieties, with matrilineal descent rules, through the trans-
mission of knowledge from mother to daughter.
More recently, Arnold (1989) has investigated learning networks
through ethnoarchaeological ¬eldwork. Realising that there are problems
with modelling descent and residence groups from material culture pat-
terns, he sets out to demonstrate that a kinship model of learning can
94 Archaeological theory and scienti¬c practice

account for the transmission of ceramic style. While he demonstrates
that residence and kinship groups may be one means by which pottery
styles may be transmitted, he suggests a dichotomy between the fabrica-
tion of pottery “ which involves the long-term learning of particular motor
habits “ and the decoration of pottery, which is derived from cognitive
knowledge. This conclusion is useful in indicating that the production
of material culture may be transmitted through kin-based learning net-
works. However, it does not explain why the transmission of knowledge
associated with the production of material culture is the same from one
generation to another. It is necessary to explain both the maintenance
and change of material forms.
There are, of course, problems with these approaches in that they per-
ceive an unproblematic relationship between material culture patterning
and social organisation (Allen and Richardson 1971). If we are to em-
ploy some of these insights we need to consider two issues. First, how are
we to conceive of learning networks; and second, how can we examine
them using materials science? Learning involves the social demonstration
of bodily techniques related to the working of speci¬c materials (Dobres
1995, 2000). Connerton (1989) argues that because social knowledge is
embodied, the production process is one means by which memories are
evoked and channelled. Since the physical form of artefacts embodies the
techniques of previous generations, the repetitive production of artefacts
therefore involves a process of recall that draws on existent artefacts as a
template. So, the act of production may involve drawing relations of af¬n-
ity with objects associated with past kin or lineage members, or it may
involve a change or alteration in the production techniques associated
with past kin members.
Social relations are also constructed through acts of production. This is
an especially important consideration where an artefact is multi-authored,
since there may be a tension in what is being expressed by each author
(Mackenzie 1991). The production of artefacts is thus a powerful means
for expressing speci¬c social relations and, as Munn (1986, 141) notes
in relation to the construction of canoes on Gawa, the use of materi-
als considered to be symbolically male and female may be perceived to
materially objectify speci¬c social relations. It is the social decision to
alter or maintain the techniques of artefact production that lies at the
heart of learning networks. It is the operation of this process that is ar-
chaeologically visible in patterns of similarity and difference, as noted by
Deetz (1968), rather than the simple transmission of styles (Wobst 1977).
Rather than viewing artefact patterning as the simple and uncontentious
result of residence pattern and descent rules, the production of artefacts
may be used to express a number of different possible relationships.
A biography of things 95

Materials science has an obvious role to play here since the distinc-
tion between the composition and construction of artefacts may be re-
¬ned through materials science analysis. In order to provide adequately
detailed data concerning the distinction between artefacts as discussed
above, we need to consider both the resolution of our sampling strate-
gies as well the resolution of our characterisation techniques. If we are
to answer questions regarding the spatial and temporal differences in the
composition of artefacts, and thence social organisation, our sampling
strategies must again be conducted at the microscale level in relation to
intra-site contexts.

Exchange, consumption and materials science
So far we have considered the way in which social networks shape the
alliances made between people and things “ in this sense people are made
by the technological conditions in which they operate. I will develop
this idea by suggesting that we also need to be aware of the modes by
which artefacts shape social relations. One of the more important ways in
which artefacts shape social relations is through their mobility, and this
mobility enables social in¬‚uences to be extended spatially and temporally
(Battaglia 1991; Gell 1998; Strathern 1998). If we are to consider the
in¬‚uence of artefacts in the creation and maintenance of social relations,
then we need to consider the processes of exchange and consumption.
I will begin with the issue of exchange. Traditional accounts of ex-
change within archaeology have tended to operate within macroscale
frameworks. According to these macroscale perspectives, formal typolo-
gies for exchange were established and attempts were made to link these
types to particular distributional patterns of artefacts in the ˜archaeo-
logical record™. These patterns were then, in turn, used to de¬ne ideal
social types (Sahlins 1972). For example, a strong correlation was made
between the exploitation of certain types of resources, the exchange of
these resources and the development of ranked societies (Renfrew and
Shennan 1982).
There are considerable problems with the development of typologies
related to mechanisms of exchange. The least of these is that it is dif-
¬cult to distinguish from the evidence of patterns in the archaeological
record alone between one mechanism of exchange and another. Another
problem is that the creation of types of exchange relies heavily upon
Western concepts of utility and economy. An alternative approach
would be to note that ˜exchange in non-western societies is really a form of
diplomacy, and for this reason it cannot be understood in purely
“economic” terms™ (Bradley and Edmonds 1993, 11“17). Exchange here
96 Archaeological theory and scienti¬c practice

is concerned with the ˜creation™, protection and manipulation of social
relationships™ (ibid., 12). So instead of considering exchange wholly in
terms of the transaction of socially neutral commodities, we also need to
consider exchange in terms of the transaction of socially valued gifts.
But what do we mean when we talk about consumption? As Gosden
(1999, 163) succinctly puts it, consumption is the process of ˜using things
in social acts™. The notion of consumption captures the point that the
way in which we consume need not relate to ideas of rationality and
necessity (Douglas and Isherwood 1979), but may be viewed as an active
means of creating cultural order and of de¬ning oneself. The process of
consumption is therefore bound up with the construction and display of
particular kinds of social and cultural identities (Friedman 1994; Miller
1995). In essence, consumption is a creative process that is determined
both by the cultural perceptions of the consuming group and by the kinds
of identities that consumers wish to construct for themselves.
This perception of consumption was pioneered by the work of Mary
Douglas, who examined the way in which meals were used as a form of
cultural expression (Douglas 1984). The important components of this
expressive activity concerned what was consumed, the mode by which it
is consumed and with whom (Douglas 1973). Meals, for instance, may
be seen as structured activities that express ideas of cosmology, identity
and speci¬c types of social relations (Deetz 1977; Johnson 1994, 1995;
Orlove 1994) on a number of planes. Our focus on food suggests a degree
of ¬nality, but in fact food is only one substance that may be used as an
expressive medium of communication “ many other forms of artefacts
are also used. For example, in an archaeological context a number of au-
thors have considered the deposition and destruction of stone (Thomas
1996; Tilley 1996), bronze (Barrett and Needham 1988) or iron (Bradley
1990) artefacts as mediums for display and consumption. If we are to dis-
tinguish the cultural speci¬city of acts of consumption, then we need to
consider the type of artefacts consumed and the manner in which they
are consumed. When we discuss consumption, then, we are not sim-
ply considering the use and discard of objects. Rather the involvement
of an artefact in speci¬c consumption practices is a critical element of that
artefact™s biography; it determines how the artefact is culturally perceived
and socially deployed. Artefacts may be consumed in many different ways,
and in different contexts, over the course of their lives.
Importantly, when we consider consumption processes, we need to
remember that in many pre-industrial societies there is a close relation-
ship between production and consumption (Longacre 1981). The choices
made in production are related to the way in which the artefacts are con-
sumed. More importantly, if we are to consider the role of materials
A biography of things 97

science in consumption studies, the way in which objects are categorised
structures the way in which they are consumed. So one way of ˜getting
at™ the issue of consumption archaeologically is to look at the differences
in the construction and subsequent use of artefacts in different contexts.
When we consider consumption, we are required to consider the physi-
cal dimensions of artefact function, but crucially we also need to be aware
that, like technology, the notion of ˜function™ is socially constructed. The
relationship between form and function is not determined by universal
rules of common sense; it is embedded within culturally speci¬c symbolic
structures. There is a whole series of physical criteria by which artefacts
may be categorised and consumed, including form, colour, texture, hard-
ness, etc. However, we need not assume that the consumption of arte-
facts is dependent on universal notions of functionality and performance.
As Miller™s (1985) study of pottery production in Central India demon-
strates, the consumption of pottery is structured by a series of variables
such as colour and morphology and their association with structuring
principles such as caste, gender and the structured consumption of food.
Rather than a simplistic correlation between form and function, material
culture may be categorised according to a complex cultural framework
that involves the categorisation of material according to a series of differ-
ing symbolic dimensions. As Miller (1985, 53) notes, the simple correla-
tion of certain vessel features with function is confounded by the use of
the same vessel form for entirely different functions both within and be-
tween social groups. Several dimensions of the artefact will be drawn on
for distinct cultural events. The categorisation of vessels in use is there-
fore context speci¬c. This study indicates that we are required not only
to consider the mechanical and physical properties of artefacts, but also
to tie this consideration to an understanding of context.
If we are to consider issues of consumption from the perspective of ma-
terials science, I believe we need to commence with a formal view of func-
tionality. Here the use-life perspective comes into its own; many studies
of the functional aspects of artefacts have been undertaken, and we need
to think in terms of which dimensions, physical qualities and properties of
artefacts will be important. For example, studies of ceramic function have
been useful in drawing out the importance of discussing factors such as
thermal shock resistance (Schiffer 1988) and the mechanical properties
of ceramic wares in terms of distinct functional capabilities (Rye 1980;
Schiffer 1988; Skibo 1993). Our attention should also be drawn to fun-
damental differences in artefact construction, such as volume (Barrett
1980; Woodward 1995; Woodward and Blinkhorn 1997). Aspects of the
physical properties of objects, such as colour, may be considered more
fully through a clear understanding of the nature of these properties in
98 Archaeological theory and scienti¬c practice

relation to modes of production (see Andrews 1997). Furthermore, at
this scale of analysis it is also useful to consider each artefact as a compo-
nent of a wider assemblage. Although this is fraught with problems “ how
are we able to determine if we have a ˜complete™ assemblage (DeBoer and
Lathrap 1979) “ we are able to distinguish between the broad differences
in the use characteristics of different components of artefact assemblages.
In terms of ceramics, simple macroscopic analysis of use-wear may re-
veal information concerning consumption practices (Skibo 1993). More-
over techniques such as GC/MS are able to provide a reasonable degree
of resolution in terms of ceramic use (Dudd et al. 1999; Evershed et al.
1990; Pollard and Heron 1996). For stone tools, microwear techniques
will also provide information concerning usage, while similar microscopic
techniques enable us to distinguish wear patterns in the use of metal
tools. Importantly many of these techniques enable us to intersect with
the domains of expertise associated with other archaeological scientists,
especially plant macrofossil specialists and faunal specialists (Meadows
1997). For example, Meadows (1997) uses the spatial distinctions in dif-
ferent kinds of faunal remains and associated artefacts from the Late Iron
Age and Early Roman period at Barton Court Farm, Oxfordshire, UK
to provide an integrative account of changing consumption practices as-
sociated with the process of ˜Romanisation™. This study highlights the
critical point that, in order to consider the social nature of consumption,
it is essential to take account of the spatial and temporal distinctions per-
taining to artefact use on a given site. This allows us to understand how
consumption practices are culturally performed.
In order to consider consumption practices at an intra-site level, we
again need to consider techniques that will provide a fair degree of res-
olution in terms of the characteristic properties of artefacts. Given that
we are able to resolve and describe differences within a site from this
perspective, it should be possible to consider how we might relate studies
of consumption practices at the local scale with a concern for exchange
practices at a wider geographical level. As noted previously, our ability to
characterise artefacts in terms of wide-scale exchange practices is fairly
well re¬ned. If we wish to consider exchange in terms of the consump-
tion practices that motivate exchange relations, we need to set our con-
sideration of the intra-site characteristics of artefacts alongside a wider
scale of analysis. The issues of exchange and consumption neatly illus-
trate the problems of macroscale and microscale analysis. Rather than
creating all-encompassing, rigid typologies of exchange patterns at the
macroscale level, Bradley and Edmonds (1993) propose that we examine
broad patterns of exchange alongside the microscale contexts of localised
consumption practices (see Hodder 1982b).They argue that we need to
A biography of things 99

be aware that we are not able to understand either wider exchange or lo-
calised consumption practices without an awareness of the reciprocal rela-
tionship of the other. We are required to oscillate between an awareness of
exhange at the macroscale level and an awarness of speci¬c consumption
practices at the microscale level if we are to understand the nuances and
interrelationships of either practice. For instance, Bradley and Edmonds™
(1993) analysis of stone axe consumption and exchange practices in
Neolithic Britain operates at local, regional and national scales, since the
data provided by their work at the site of production in the Lake District
were complemented by an examination of exchange and consumption
patterns in northern England, which works with the detailed petrological
data compiled in Britain since the beginning of the twentieth century.
In essence, we need to set our detailed understanding of localised con-
sumption against our knowledge of larger-scale exchange patterns and
continue to tack back and forth between the kind of information provided
by local contexts of interaction and that provided by wide-scale exchange
studies. In each case we are required to provide an adequate characterisa-
tion of artefacts at each scale of analysis. Studies that consider both local
and wider scales of analysis should operate with comparable techniques
of analysis. However, such studies may favour a multiscale contextual
approach to sampling, one which seeks simultaneously to examine dis-
tinct local contexts of difference, in which the sampling of artefacts is
extremely detailed, and distinctions of difference at much larger regional
scales, in which the sampling of artefacts is reasonably coarse.

Fragmentation: the science of deposition
As noted at the beginning of this chapter, our accounts of the life of an
artefact necessarily work backwards from the ¬nal stage of the life of the
artefact “ its discard. It is a truism that it is at the end of an artefact™s life
that the artefact becomes most archaeologically visible. Throughout this
chapter we have examined the relationship between the various stages of
the life of an artefact and the scales used in its analysis from a materials
science perspective. Given the critical importance of deposition to our
understanding and analysis of the life of an artefact, I will commence
by considering how we characterise archaeological deposits. As indicated
in the opening chapter, the characterisation of archaeological deposits is
subject to debate. From a rationalist perspective, deposits consist of the
traces of past physical processes. However an alternative perspective is to
view the deposits as structured, a concept predicated upon the idea of
the archaeological record as the result of intentional action. Structured
deposition was initially correlated with structured activity such as ritual
100 Archaeological theory and scienti¬c practice

(Richards and Thomas 1984), in part since many structured deposits
appeared to be on sites of a non-functional or ritual nature. However we
limit ourselves if we simply equate structure with ritual, and the lack of
structure with non-ritual activity. A more developed view may be that no
deposits of an anthropogenic nature can be dismissed as simple casual
discard, given that all forms of depositional activity are of a cultural na-
ture and therefore are informed by culturally speci¬c systems of logic. We
need to move beyond the simple recognition of differential patterning in
artefact assemblages (Brown 1991; Cleal 1991; Pollard 1992, 1995) to
consider how these deposits may once have been deployed in the con-
struction and expression of social relations.
Critically, Hoffman (1999) notes that differing modes of destruction or
discard may be considered to be important ways by which social groups
both de¬ne themselves and create social relations. He notes examples in
contemporary contexts “ such as graf¬ti “ and in prehistoric contexts “
the breakage of metals “ as two differing cultural modes expressing both
the de¬nition and the creation of social identities. The recognition of the
deliberate breakage of objects prior to deposition has been noted for
some time. For example, in the ceramic ethnographies of Africa (Barley
1995; Sterner 1989) and in the deposits within the facade of Danish and
Swedish passage graves (Tilley 1984, 1996; Shanks and Tilley 1987) acts
of breakage seem to be both periodic and deliberate. Here processes of
destruction may be understood according to a number of aspects associ-
ated with the biography of an artefact. The artefact, like any animate life
form, may be considered to contain ˜spirit™, and therefore its corporeal
form is destroyed at the end of its use-life. Alternatively, the destruc-
tion of commodities acquired through exchange may be one means of
gaining social prestige which may occur through a number of mecha-
nisms, notably through the staging of feasts or as a votive deposition. Or
the destruction and deposition of objects may be linked to the activities
with which artefacts are related, for instance mortuary rituals. There are,
therefore, a series of socialised activities that may result in the destruction
and deposition of material (Bradley 1985, 1990), as opposed to the sim-
ple causative principles of wear or accident (Arnold 1991; DeBoer and
Lathrap 1979; Nelson 1991; Schiffer 1976). We might consider acts of
destruction or deposition within the broader parameters of consumption
practices, and these kinds of activities should be considered as achieving
similar social ends.
John Chapman (2000b) has recently taken this point further in his in-
vestigation of depositional practices in Eastern European prehistory. He
considers artefact depositional practices in terms of states of fragmen-
tation, wholeness, juxtaposition and accumulation. He proposes that we
A biography of things 101

might consider the social relations represented by depositional practices
in terms of fragmentation and accumulation. He suggests that the frag-
mentation of objects, the accumulation of sets of objects and the recom-
position of fragmented objects are critical to the creation of social rela-
tions, since the act of breaking and sharing material culture establishes
af¬liation between people. Similarly the act of accumulating objects and
the act of creating composites out of distinct fragments harnesses the
relations established in sharing, through cementing and articulating to-
gether shared social bonds, and thereby re-articulates a new set of social
relations. Depositional practices may therefore involve very complex acts
in which social identities are clearly signalled.
So if we embrace the view that depositional practices are the result of
intentional and meaningful activity, how are we to relate this to a materi-
als science perspective? Are we to abandon the view of the archaeological
record as physical trace? It is important to realise that while a contextual or
structured approach to the archaeological record may allow us to discuss
how artefacts were employed to construct and live in past social worlds,
a notable failing of such an approach is the tendency to overlook the ma-
terial qualities of past material culture. We need to develop an approach
that views the material properties of artefacts to be critical elements in
enabling us understand the taphonomic nature of the site. We also need
to be aware that the formation processes involved in the creation of the
site were the result of culturally speci¬c depositional practices.
First, if we are to examine the meaningful nature of artefact deposi-
tion, we need to consider the mechanical properties of different materials
(Cotterall and Kamminga 1990). This enables us to consider the attri-
tion of artefacts at both pre- and post-deposition stages (Schiffer 1976).
Microscopic analyses of artefacts enable us to consider the level of wear
on artefacts and so allow us to consider the processes involved in the
destruction of artefacts prior to deposition. For example, we have to
consider questions such as are the edges of artefacts worn, suggesting
pre-depositional wear, or are their edges freshly broken, suggesting de-
liberate destruction prior to deposition? If this information is combined
with simple metrical analysis of artefact size (Bradley and Fulford 1980;
Schiffer 1976), and knowledge of the likely mechanical breakage patterns
of distinct materials, an analysis of this nature will allow us to consider
the detailed contextual differences in the nature of depositional practices.
One further process involved in analysing the practices that lead to
the creation of on-site deposits is the investigation of artefact re¬tting
patterns (Lindauer 1992). This is obviously easier with certain classes of
artefact than others. However, the task of re¬tting may also be aided by
characterisation studies. If we have an adequate characterisation of the
102 Archaeological theory and scienti¬c practice

distinct composition of an artefact, then the process of re¬tting may be
clari¬ed. For instance, the relationship between fragments of stone tools
and pottery may be clari¬ed by petrological or elemental analyses. Again,
a combination of approaches that attend to the ¬ne-grained analysis of
artefacts alongside a materials science approach should provide fruitful

A biographical approach to artefacts allows us to chart the trajectories of
an artefact™s life from the raw materials of production through to its ¬nal
deposition. However, it is important to remember that biographies are
¬‚uid. What I have presented here is simply a broad narrative structure “
we must not forget that as an artefact progresses through its life, it is likely
to change its meaning and status. Attention to its material status at various
stages of its life, through the application of materials science techniques,
will enable us to draw out this ¬‚uidity. As we shift away from a view of
the object as a static entity and begin to consider how we can chart its
changing trajectory as it moves from context to context, we are required
to engage with this idea using the imaginative application of materials
science techniques. In the following two chapters I will demonstrate how
such an approach to materials science and artefact biographies might be
6 A biography of ceramics in
Neolithic Orkney

The next two chapters provide an extended case study that illustrates
some of the ways in which we might articulate the methodologies of ma-
terials science with the concerns of interpretative archaeology. Before
commencing with this, I want to reiterate a point that recurs throughout
this volume: we are required to consider the interpretative framework
within which we operate prior to undertaking our analyses. Dobres artic-
ulates this well with regard to technological studies when she argues that
˜explicit consideration of the sociopolitical nature of technologies cannot
be done after the material facts are settled; one cannot simply insert sym-
bolism, questions of value, or the dynamics of social differentiation into
a pre-existing materialist pot that, by de¬nition, discounts or downplays
them as constitutive elements™. She then goes on to point out, ˜These
intangible processes clearly play a structural role in shaping and chang-
ing technologies . . . and if we are to understand how they did so in the
past, they must be central to our conceptual frameworks rather than added
after the facts are in™ (Dobres 2000, 118, original emphasis).
In this chapter and the next, I will present an analysis in which con-
cerns of a theoretical nature play a central role in determining what and
how material is analysed and how this analysis ultimately relates to wider
theoretical concerns. My case study focuses on the analysis of a pottery
assemblage from the later Neolithic settlement site of Barnhouse, Orkney,
Scotland. Before examining the details of the case study, I will situate the
analysis within the broader framework of Neolithic studies.

Problems in Neolithic archaeology
Traditionally the onset of the Neolithic is characterised by a number
of signi¬cant events. These include the change from a shifting hunter-
gatherer economy to an economy based on sedentary agriculture and the
adoption of a suite of novel forms of material culture, including pottery
and polished stone tools and the construction of substantial stone, tim-
ber or earth monuments. Although the term ˜Neolithic™ has undergone

104 Archaeological theory and scienti¬c practice

considerable alteration and rede¬nition (Thomas 1993), until recently
the role of the economy was considered to be critical to the process of
becoming Neolithic (Higgs and Jarman 1972). The use of new forms
of material culture was assumed to be allied to this economic change,
with pottery and polished stone axes viewed as consequences of the pro-
cesses of sedentism and deforestation that arose from an economy based
on agriculture. Monuments too were assumed to be ¬nanced by agricul-
tural surplus (Case 1969; Sherratt 1990).
However, this view of economic primacy has increasingly been criti-
cised. For instance, Thomas (1991, 7“8) argues for the inseparability of
issues of economy with the social changes that constitute the Neolithic.
He considers the Neolithic to be a historical process bound up with a
changing set of relationships that occurred both between people, and be-
tween people and their environment (see also Hodder 1990). This point
is echoed by Whittle, who notes that the Neolithic in Europe was based
˜on a set of beliefs, values and ideals™ (1996, 355). He goes on to say that
these beliefs concern ˜the place of people in the scheme of things, descent,
origins and time, and about relations between people™ (ibid.). We might
reverse the equation then and suggest that we view the Neolithic not as a
social effect brought on by the adoption of agriculture, but as an effect of
changing social relations and beliefs which made the idea of agriculture
possible. This proposal has been substantiated by careful examination of
the chronological sequence relating to the adoption of agriculture and
the construction of monuments.
In many areas of Neolithic Europe the chronological evidence suggests
that monuments are found alongside agriculture or that monuments ac-
tually precede agriculture (Bradley 1993), so reversing the traditional
causal emphasis. Rather than viewing monuments as a by-product of
agriculture, we may view them as central to the process of becoming
Neolithic. If we are to consider the Neolithic as a process that engendered
an alteration of beliefs and social relations then monument construction
appears to be a critical element in this process. Monuments evoke an
altered conception of both time and place; they embody an alteration of
the natural world, and their construction involves the creation of a new
kind of place in the landscape which, by their very nature, they endure.
This perception of the world may be allied to the perceptions required
of the agricultural regime, but there is no necessary relationship. The
relationship between food production and monumentality is therefore
Curiously, although the adoption of novel artefacts at one stage de-
¬ned the Neolithic, recent assessments of the period have emphasised
either agriculture or monumentality. As I have argued above, both of
A biography of ceramics in Neolithic Orkney 105

these aspects of the Neolithic are critical to our understanding of the pe-
riod. However, we have seen less emphasis placed on the signi¬cance of
material culture in the process of becoming Neolithic. I would argue
that a reconsideration of the role of material culture is essential if we
consider this process to be bound up with the recon¬guration of the re-
lationship between people and the world they occupy. As we have seen in
the previous chapter, material culture is central to the construction and
maintenance of social relations. An examination of the manner in which
material culture is employed during the Neolithic should enable us to
understand in more detail how relations between people and their world
are con¬gured during this period.
Importantly, while monuments altered the experience of time and
place, the creation of monuments will also have drawn on the culturally
categorised landscape. Plants, animals, and other elements of the land-
scape will be meaningful components of the culturally classi¬ed landscape
prior to monument construction (Jones 1998, forthcoming a). If we are
to understand the Neolithic as a series of changing relationships between
people and the natural world, then we are required to take into account
not only the temporal and spatial experience invested by monuments.
We also need to consider the temporal and spatial experience associated
with plants and animals and with the production, use and deposition of
artefacts. The aim, then, is to examine one aspect of the Neolithic, in this
case pottery, as a means of illuminating the complex relationships that
existed between material culture and agriculture and monuments. Each
of these elements forms the conditions through which the Neolithic was
constructed and lived. What interests us here is how each element was
deployed in the process of constructing and living in the Neolithic.

Pots and people
The typological classi¬cation of pottery remains the primary tool for ar-
chaeologists seeking to understand the chronology of a given site. Pots
have consistently been employed by archaeologists as a ¬ne-grained in-
dicator of the presence of speci¬c cultural groups. Archaeologically, pots
appear to equal people. However this relationship requires further ex-
amination. We need to carefully assess what relationship exists between
pottery and people, and ask why pottery is assumed to be such a precise
indicator of cultural groups. While an interpretative analysis of a pottery
assemblage comprises the core of this case study, the study has involved
the use of a series of scienti¬c analytical techniques in order to char-
acterise the nature of the pottery assemblage. The speci¬c techniques
of thin-section petrology and Gas Chromatography coupled with Mass
106 Archaeological theory and scienti¬c practice

Spectrometry have been drawn on in order to examine and de¬ne the
relationship between pottery, social practices and social identity.
Typologically, pots have been traditionally studied as objects divorced
from their cultural context. This is a general problem of artefact analysis.
Pots are used for the storage, preparation, cooking and consumption of
food, amongst other things, and it is essential to view pots not simply
as passive decontextualised sherds, but as actively produced and used
according to the culturally speci¬c motivations of people. In order to do
this it is essential to study not just individual aspects of pots such as their
function (Braun 1983); production (Wardle 1992); use (Evershed et al.
1995) or deposition (Richards and Thomas 1984), but beyond this to
realise that all of the above aspects are important. Pots are at all times
linked with each ¬eld of activity, since pots are made by people who are
embedded within a particular social structure and cultural framework.
Just as the processes of production, use and deposition of pots are linked,
so the functional (Rice 1996) and symbolic (Hodder 1982c; Tilley 1984)
or metaphorical (Gosselain 1999) aspects of pottery cannot be separated
(see Boast 1998).
Given the assumed status of pots in relation to people and in order
to fully understand how this relationship is brought about, it is essen-
tial to examine the way in which pots are produced, used and deposited
by people. Here it is important to expand the way in which we might con-
sider these practices. Rather than examining these activities as ciphers
for static social identities, I prefer to consider identities to be created
through practice, a process in which the performance of social practices
in distinctive ways is productive of different kinds of subjects or people
(see Butler 1990; Strathern 1988). This perspective allows us to move
away from an equation that simply relates speci¬c forms of material cul-
ture with speci¬c kinds of identity, and towards an examination of the
way in which identities are instantiated by the contextual relationship of
artefacts associated with different practices (see Jones 1997, 106“44).
An excellent example of this approach to material culture is Mackenzie™s
(1991) analysis of the gendered identities associated with the production
of woven bags in Papua New Guinea. Here bags are made by women and
embellished by men, so the bags are ˜androgynous™ “ they are associated
with both men and women. Yet as objects they serve to materialise the
complex construction of gender relations in this area.
The aim then is to examine how social relationships are expressed
through the practices associated with the production, use and depo-
sition of Later Neolithic Grooved ware pottery, and how social iden-
tities are formed out of these relationships. This approach allows us
to conceptualise how identities might shift over time as artefacts are
A biography of ceramics in Neolithic Orkney 107

transferred from one context to another and enter into fresh contextual
relationships with other artefacts. It is the dynamic associated with the
creation of fresh or novel contextual relationships over time that allows
us to see how “ through contextual juxtaposition “ artefacts are used to
express a range of different kinds of social identities (for example see
Thomas 1996, 141“83).

Introducing Grooved ware
Grooved ware is a class of Later Neolithic pottery whose distribution
spreads from the Orkney Isles in the north of Scotland to southern Eng-
land. Many Grooved ware assemblages derive from ˜ritual™ sites such as
henges or structured deposits within pit clusters found throughout main-
land Britain. The de¬ning characteristic of Grooved ware is its unusual
decoration, which lends the pottery its name. It has long been noted
that these decorative motifs are analogous to many other forms of later
Neolithic material culture (see Fig. 6.1) especially passage grave art
(Bradley 1984, 1989; Bradley and Chapman 1986; Cleal 1991; Piggott
1954; Shee Twohig 1981; Wainwright and Longworth 1971), the art
found on objects such as the Folkton drums (Kinnes and Longworth
1985), the Garboldisham macehead (Edwardson 1965) and the carved
stone balls of northern Scotland (Edmonds 1992; Marshall 1977).
Due to its unusual decoration and its association with apparently non-
domestic contexts, Grooved ware has been characterised as an ˜exotic™
(Bradley 1984; Cleal 1991) or ˜ritual™ (Richards and Thomas 1984)
ceramic. Meanwhile, other aspects of Grooved ware, such as its function,
have largely been ignored (see Cleal 1992 for a general discussion of
studies of function in relation to British Neolithic ceramics). This state
of affairs is due, in part, to the fact that many studies of this class of
pottery have emphasised its distribution in the southern half of Britain,
where it is more usually found on ˜ritual™ sites. This is somewhat paradox-
ical since the earliest radiocarbon dates are derived from the Orkney Isles
(see Bradley 1984). While Grooved ware in the rest of Britain and Ireland
is typically associated with non-domestic and non-mortuary contexts, in
Orkney the pottery is primarily found in settlements, as well as in sites
that might be considered to be more overtly related to ˜ritual™ activities,
such as passage graves and henges.

Grooved ware and Neolithic Orkney
The Orkney Isles are situated ten miles north of the Scottish mainland
(Fig. 6.2). The archipelago comprises around seventy islands of variable
size. The isles are situated at latitude 59—¦ north, in the North Atlantic
108 Archaeological theory and scienti¬c practice

Figure 6.1 A selection of Neolithic material culture exhibiting similar
curvilinear motifs

Ocean, and this northerly location provides them with contrasting light
conditions over much of the year. In the summer months, the hours of
daylight are numerous, with only around two to four hours of darkness; in
the winter months this is reversed and the hours of darkness are numer-
ous, and only two to four hours of daylight are experienced. This situation
is important for understanding a variety of aspects of Later Neolithic life
(Richards 1990a).
The topography of the islands provides a contrast between the land
and the sky, while the treeless nature of the landscape means that the sky
is also an ever-present feature of the Orcadian horizon. All these aspects
Papa Westray
North Ronaldsay





Hoy Burray

South Ronaldsay Scotland

Figure 6.2 Map showing Orkney archipelago
110 Archaeological theory and scienti¬c practice

go to make up a picture of a fertile but dramatic landscape, unchanged in
terms of vegetation since the Later Neolithic, and dominated by coastal
stacks of rock and lowland lochs. Geologically the islands are comprised
almost entirely of Middle and Upper Old Red Sandstone. The sandstone
bedrock laminates into easily worked slabs that comprise much of the
building stone used during the Neolithic. Probably the most important
geological features of the islands, in relation to this study, are the intrusive
igneous dykes that outcrop intermittently around the inland and coastal
shores of the islands (Mykura 1976).
The Later Neolithic period in Orkney is characterised by a series of
stone-built architectural forms, including settlements, passage graves and
henges. Interestingly each of these is constructed with a similar empha-
sis on circular space. Richards (1990a) has noted that the circular form
of the Later Neolithic house is organised according to a cruciform axis.
The house is focused around a central hearth, with a ˜dresser™ or set of
shelves towards the rear of the house, and ˜box-beds™ or stone boxes sit-
uated either side of the hearth, with the entrance itself completing the
cross-shaped arrangement of space (Fig. 6.3). For Richards, the consis-
tency of this arrangement is related to speci¬c cosmological principles

Figure 6.3 The spatial layout of the Later Neolithic house in Orkney
A biography of ceramics in Neolithic Orkney 111

of classi¬cation. Not only are hearths arranged in order to face speci¬c
cardinal points, related to events of calendrical signi¬cance, especially the
midwinter and midsummer sunrise and sunset, but most of the houses
at Barnhouse, Skara Brae and Rinyo lie on a northwest/southeast axis.
This arrangement of space, as well as having an underlying symbolic
logic related to ideas of concentricity and circularity, applies not only to
the house, but also extends to other monumental constructions such as
henges and passage graves (Hodder 1982d; Richards 1993a), as well as
the landscape itself (Richards 1996).
It is notable, then, that a number of homologies exist between the
construction of houses and other monument forms (Fig. 6.4). Passage
graves are constructed with a central chamber with a series of side cells
exiting this space, and an extensive passage exiting the central chamber
enabling access to the exterior of the monument. Henges, such as the
examples at the Ring of Brodgar and the Stones of Stenness, are de¬ned
by a circular bank and ditch which surround a circle of monoliths. In the
case of the Stones of Stenness, the relationship between house and henge
is emphasised by the construction of a large hearth in the centre of the
monument (Richards 1993a). Each of these architectural constructions
is related through similar principles of order. Moreover, each form of
architecture provides the context for the use and deposition of Grooved
ware pottery. If we are to understand how the biography of Grooved
ware unfolds in Later Neolithic Orkney, we are required to begin with an
examination of both the house and the settlement.
Orkney constitutes one of the few areas of Europe with substantial up-
standing evidence for Neolithic habitation. Although houses are stone-
built, the occupation sequences on many Orcadian Neolithic settlement
sites provides evidence of continuous episodes of building and rebuilding.
Neolithic settlements in Orkney often have a tell-like form with long se-
quences of occupation throughout the Neolithic. Due to this occupational
history, and due to the practice of depositing midden material in close
proximity to the house and within the wall core of the house, archaeolo-
gists have been able to establish pottery sequences for much of the Orkney
Neolithic. The primary sequence was established by Gordon Childe
(1931, 130“2) after his excavation of the celebrated site of Skara Brae:

Class A Relief/applied decoration; A1 simple applied cordons; A2 cordons ap-
plied with slip. A1 was found in all phases, A2 was found only in phase 2.
Class B Relief decoration augmented with incisions on grooves. Does not occur
beyond phase 2.
Class C Grooved decoration incised into slipped surface.Found in phases 1 and 2.

This broad scheme, based on differences in the technique of decora-
tion, with incised or grooved decoration at the beginning of the sequence
112 Archaeological theory and scienti¬c practice

Figure 6.4 The spatial homology between passage grave, house and
A biography of ceramics in Neolithic Orkney 113

Figure 6.5 The distinction between incised Grooved ware and applied
Grooved ware

followed by applied cordons (Fig. 6.5), accords with more re¬ned se-
quences for Orkney Grooved ware (Hunter and MacSween 1991). Later
Neolithic settlements therefore provide well-strati¬ed pottery sequences,
enabling us to examine the changing nature of social practices over time.
Moreover, the high density of well-preserved settlements across Orkney
also enables us to examine the nature of interactions between contempo-
rary settlements within the islands.
The principal aim in this study was to examine the nature of the bi-
ography of Grooved ware as it was constructed through social practices
in different kinds of context, such as the settlement, henge and passage
grave. A further aim was to examine the way in which “ at a regional
scale “ such biographies were related to the construction of settlement
histories. In order to achieve these aims, a detailed study was made of a
large Grooved ware assemblage from the settlement site of Barnhouse.

Introducing Barnhouse
The Barnhouse settlement itself is situated on a promontory in the centre
of Mainland Orkney (Richards forthcoming). This area is topographically
low lying, and forms the centre of a natural bowl bounded by two lochs
and surrounding hills. It is located in the centre of a remarkable concen-
tration of Neolithic monuments (Fig. 6.6). Within sight of the settlement
Figure 6.6 Map of the central area of Mainland Orkney indicating position
of principal monuments
A biography of ceramics in Neolithic Orkney 115

to the east is the passage grave of Maes Howe, and the Stones of Stenness
henge is located some 150 m away. Also visible from Barnhouse is the
earlier Neolithic chambered tomb at Unstan. Nearby lies the immense
henge and stone circle, the Ring of Brodgar, and some distance north
of this is the Ring of Bookan, a probable passage grave, as well as the
chambered tomb of Bookan. On the farm of Bookan, to the northeast of
Brodgar, there is also a possible Later Neolithic settlement represented
by a rich artefact scatter (Callander 1931).
Since the construction of houses in Orcadian Neolithic settlement sites
is ¬‚uid, it is dif¬cult to divide the site into phases, although various con-
structional episodes can be isolated. The initial stage of construction
involved laying a complex system of drains. Despite the apparently mun-
dane nature of this activity, it would seem that this initiated and solidi¬ed
the spatial structure of the settlement. Two discrete systems of drains are
arranged in two concentric arcs. The inner arc of drains connects house
6 as well as later houses 1, 11 and 12 in the centre of the settlement. The
second arc of drains connects houses 2, 3, 5 and 9.
The spatial arrangement of the settlement therefore consisted of two
concentric arcs of houses surrounding a central space. This central area
is important for our considerations of the spatial organisation of pottery
production. In the earliest phase of settlement around seven houses were
built, including houses 2, 3, 5, 6, 7, 9 and 10, followed by houses 11 and
12 (Fig. 6.7). Most houses appear to have been, at some time, demolished
and then built over, in approximately the same position and orientation.
This is particularly clear in the case of structures such as house 5, rebuilt
over four times. What is most notable is that both houses 2 and 3 appear
to continue in use through much of the life of the settlement, and the
architecture of these houses conforms with excavated examples from the
earliest levels of Rinyo and Skara Brae (Childe 1931; Childe and Grant
Not all houses at Barnhouse are constructed in precisely the same way.
The architecture of house 2 draws on an arrangement of space similar
to other Later Neolithic houses; however, the house also has a double
cruciform arrangement with a total of six recesses, an architectural form
that recalls the plan of the passage grave at Quanterness (Renfrew 1979).
House 2 stands out from the rest of the settlement and, as we shall see,
the activities conducted within it mark it out as unusual.
The ¬nal use of the Barnhouse settlement is marked by the construction
of a monumental building, structure 8. This building was constructed
south of the main area of settlement and consisted of an external clay
platform on which a large square building was constructed (see Figs. 6.7
and 6.19). The platform contained a series of hearths and was surrounded
116 Archaeological theory and scienti¬c practice

Figure 6.7 Plan of the Later Neolithic settlement at Barnhouse
A biography of ceramics in Neolithic Orkney 117

by a stone wall, and the central building was entered through an elabo-
rate porch. This building draws on the spatial arrangement of other Later
Neolithic houses, having a large central hearth and dresser, while also hav-
ing similarities to the architecture of the passage grave (with an external
platform) and henge (a large central hearth and porch-like entrance).
The occupation of the site spans a four-hundred-year period as esti-
mated by calibrated radiocarbon dates. The earliest calibrated dates for
the settlement are 3600“3110 BC and 3500“3100 BC, while the latest
dates are 3270“2920 BC and 3090“2910 BC.
With this detailed chronology it is possible to examine the changing
production and use of Grooved ware in relation to different houses and
at different stages during the life of the settlement. While the ¬rst two
phases of building associated with the settlement are fairly ¬‚uid, the con-
struction of structure 8 marks the ¬nal phase of occupation, and it is
possible to contrast the Grooved ware associated with this later building
with that from the earlier houses. Like other Later Neolithic settlements,
the earlier phases of building are predominantly associated with Grooved
ware decorated by incision, while the later phases of occupation are gen-
erally associated with Grooved ware decorated by cordon.

Analysing the Barnhouse Grooved ware
As noted earlier, my interest in the Barnhouse Grooved ware assem-
blage relates to the way in which social identities were instantiated dur-
ing episodes of procurement, production, use and deposition. I was also
interested in the way in which these activities helped to shape certain
social identities both within the settlement and beyond. Here I will de-
scribe how the interpretative approaches taken to the examination of this
pottery assemblage were allied with speci¬c methodological objectives.
My initial examination of the Barnhouse Grooved ware assemblage
was concerned with the macroscopic characterisation of the pottery. As
indicated in chapter 3, traditional accounts of pottery from many British
prehistoric sites commence with an attempt to characterise pottery as-
semblages with the ultimate aim of relating the pottery with other similar
assemblages. This practice is, of course, an artefact of earlier culture-
historical approaches allied to an interest in the use of artefacts as a
dating mechanism. I found this pursuit to be largely fruitless with re-
gard to my investigation of pottery and social identity, since the ultimate
product of such an activity is a static catalogue of annotated sherds. A
number of ˜diagnostic attributes™ of pottery are traditionally recruited for
this task, the most obvious being decorative motifs, decorative technique,
and rim and base morphology. However, my interests related to the active
118 Archaeological theory and scienti¬c practice

construction and use of pottery. While decoration and rim and base mor-
phology are crucial elements of pottery characterisations, I was equally
concerned to consider the fabric of the assemblage since this allowed me
to consider differences in the choices made in production technologies
(see Cleal 1996; Sillar and Tite 2000). Rather than placing emphasis on
the traditional diagnostic attributes of pottery, I wished to examine the
individual components of the pottery assemblage in terms of a range of
variables, which emphasised not only their morphological differences but
also their functional differences (Braun 1983; Cleal 1992).
I was interested in considering each component as part of a functioning
pottery assemblage. As both Miller (1985) and Boast (1990, 181“2) have
indicated, categories of pottery are created through the manipulation of
various ˜dimensions of variability™. Such ˜dimensions of variability™ may
include fabric, wall thickness, decorative motifs, decorative scheme, etc.
It is the variation in these ˜dimensions of variability™ which distinguish
one category of vessel from another. The aim, then, was to distinguish
the various dimensions along which Grooved ware vessels varied, and
thereby characterise the differences that constitute the various categories
of Grooved ware at Barnhouse.
In order to consider the dimensions along which these individual com-
ponents varied, I needed to examine a range of attributes. While these
included morphological variables, I was also concerned to examine dis-
tinctions in other attributes related more closely to the performance
characteristics of pottery (Schiffer and Skibo 1987), including volume
(Woodward 1996), fabric (Cleal 1996), use-wear (Skibo 1993), sooting
(Hally 1983) and organic residues (Evershed et al. 1992). This perspec-
tive, which has seen less emphasis in the context of British ceramic studies,
led me to examine the internal differences in the pottery assemblage.
Having examined the different categories in the assemblage, my anal-
ysis was directed towards the construction and use of these different cat-
egories of pottery in relation to the different contexts within the settle-
ment. Here my theoretical in¬‚uence was twofold. First, I was struck by
ideas promoted by the early generation of processualists concerning the
construction of kinship networks from material culture patterns. While
I found these ideas appealing, they carried with them the problems asso-
ciated with treating the archaeological record as a direct and unmediated
record of past events (see Patrik 1985 and Barrett 1994, 2000 for critical
discussions). While in¬‚uenced by these approaches to artefact pattern-
ing, I preferred the approach adopted by Dobres (1995, 2000) and Jones
(1997) concerning the relationship between the practical organisation of
production and the more subtle expression of social differences. I also
took on board the relationships established by Arnold (1989) between
A biography of ceramics in Neolithic Orkney 119

learning networks, motor habit and identity. Adoption of these perspec-
tives led to the petrological examination of a representative sample of pot-
tery from each house at Barnhouse. Rather than concerning myself with
problems of supralocality (contra Sheridan 1991), I was more concerned
with the internal constitution of the assemblage and with the relationship
between differences in the assemblage and different procurement strate-
gies (see Howard 1981). I therefore applied a microscale approach that
took account both of differences in the assemblage and of differences in
context in the sampling procedure.
This microscale approach to sampling enabled me to de¬ne differences
in the primary tempering strategies from different contexts in the site.
These differences led me to consider the relationship between the use of
these resources and the signi¬cance of the locality of these resources in the
lived landscape. Here the framework developed by Arnold (1985) relat-
ing to the in¬‚uence of resource availability on procurement strategies was
critical. However, discussions concerning the signi¬cance of the cultural
categorisation of the landscape (Tilley 1994) and the relationship between
place and identity (Casey 1987; Gow 1995) in¬‚uenced my approach to
the magnetometer survey of the area surrounding the Barnhouse settle-
ment and my subsequent survey of the environs of Quanterness passage
Having examined the distinctions between place, production and iden-
tity, I was concerned to examine the distinctions between the use of
different categories of the assemblage in relation to context. Here at
a primary level I was interested in the kinds of arguments concerning
food and consumption developed within anthropology (Douglas 1966,
1973; L´ vi-Strauss 1970). However, after Friedman (1994) and Miller
(1995), I was also interested in the role that consumption practices play
in the construction of identities (see Orlove 1994 for a good example).
It was this in¬‚uence that led to the selection of speci¬c categories of
pottery (related to the overall demography of pots in each house) from
speci¬c contexts for Gas Chromatography/Mass Spectrometry (GC/MS)
analysis. For this reason more samples were taken from the more fre-
quent medium-size vessels than from the small or large vessels. Analysis
here was meant to be representative of the excavated assemblage from
each house. This analysis cohered with the known information derived
from faunal and botanical studies from Barnhouse and other sites in
Orkney. This allowed me to construct GC/MS standards for different
species against which to match the Grooved ware data, but moreover
it allowed me to articulate the analysis of food and Grooved ware
with a contextual study of animal and plant deposition (Jones 1998,
120 Archaeological theory and scienti¬c practice

The combined data derived from each of these forms of analysis en-
abled me to construct biographical trajectories for distinct categories of
pots from their place of production, through use, and to relate both of
these factors to their place and manner of deposition. My concern with
the examination of the biographical process of artefacts was primarily
related to recent discussions in anthropology (Battaglia 1991; Thomas
1991; Weiner 1990) and archaeology (Dobres 1995; Edmonds 1995;
Thomas 1996; Tilley 1996). My interest in biography led to a more de-
tailed microscale approach to petrological sampling, and it was the speci-
¬city of tempering strategies de¬ned by petrological analysis that allowed
me to work back from pots deposited within settlement middens, henge
ditches and passage graves to de¬ne their place of origin.

The Barnhouse Grooved ware
The Barnhouse Grooved ware mainly varied in size and capacity. Mor-
phologically, the same broad ˜bucket™ shape was employed at different
scales. While a continuum in size was observed, three broad clusters of
pottery size could be discerned: large, medium and small. These vessel
sizes were de¬ned by distinctions in three main ˜dimensions of variability™,
including fabric, wall thickness and decoration. In characterising vessels
of different categories, I will consider the process of production as a chaˆne
op´ratoire or a series of interconnected technical choices (see Sillar and
Tite 2000), analogous in many ways to the hierarchical design decisions
discussed by Friedrich (1970) and Plog (1980). The point here is that
each category is made up of a series of technical decisions related together
in a hierarchical or ordered sequence.
The primary factor that distinguishes vessels of different size cate-
gories is the manner in which pots were tempered. Fabrics can be broadly
grouped into the following categories:
Fabric A Rock-tempered with a frequency of inclusions between
10“30 per cent.
Fabric B Rock-tempered with a frequency of inclusions of
50 per cent or more.
Fabric B1 Rock-tempered with a frequency of inclusions of
50 per cent or more, also tempered with approximately 10
per cent shell (observed as voids).
Fabric C Shell-tempered with a frequency of inclusions of
10“30 per cent (these are observed as voids since the shell
has decayed in acidic soil conditions).
Fabric D Untempered, only non-plastics include naturally
rounded quartz inclusions.
Fabric E Untempered.
A biography of ceramics in Neolithic Orkney 121

Primary variation therefore depends on the type of temper used: shell,
rock or no temper. There is also further variation depending on the fre-
quency of the tempering agent. Vessels tempered with rock vary accord-
ing to the frequency of temper: for fabric A, 10“30 per cent; for fabric
B and B1, 50 per cent. Although fabrics D and E remain untempered,
the presence or absence of natural non-plastics within the clay also dis-
tinguishes them. Shell-tempered vessels are again distinguished by this
variable: fabric C vessels are distinguished by the high frequency of shell
inclusions; and fabric B1 is tempered with a lower frequency of shell
along with a high frequency of rock.
The reasons for this variation were initially dif¬cult to understand;
however, if we compare temper frequency in terms of fabric, against wall
thickness, then we see that variation in temper depends on the size of
vessel constructed (Fig. 6.8). Fabrics A and B include vessels tempered
with rock inclusions; vessels of fabric A are 10“15 mm in wall thickness;

Figure 6.8 Graph of fabric plotted against wall thickness
122 Archaeological theory and scienti¬c practice

while vessels of fabric B are typically over 20 mm in wall thickness. Shell-
tempered vessels of fabric C are 7“15 mm in wall thickness, while those of
fabric B1 are typically at the extreme end of the scale being over 25 mm
in wall thickness. Fabrics D and E are essentially part of a continuum
ranging from 5“9 mm thickness, with Fabric D occupying the upper end
of the scale, fabric E the lower. At the outset then vessels are distinguished
by fabric and wall width. These variables are related to the three distinct
sizes of Grooved ware vessel from Barnhouse. Since Grooved ware ves-
sels are similar in shape, these variables also provide an index of vessel

1 Large vessels (Fabric A, B and B1): Vessels with a wall thickness of
16“30 mm and a volume of 10,000“35,000 cc.
2 Medium vessels (Fabric A and C): Vessels with a wall thickness of
9“15 mm and a volume of 2,000“8,000 cc.
3 Small vessels (Fabric C, D and E): Vessels with a wall thickness of
5 to 9 mm and a volume of 2,000“3,000 cc.

The next major variable is decoration. We observe differences in the
nature of vessels of each size with a hierarchy of decoration; large ves-
sels are simply decorated with a single incision or cordon; medium-
size vessels are decorated with three curvilinear incisions or with a cor-
don that is altered in a serpentine pattern; and the small vessels are
decorated with curvilinear incisions or serpentine cordons. However,
with the small vessels we also witness the addition of passage grave art
motifs such as rings of dots or rosettes. While we can observe differ-
ences in decorative technique and the use of simple decorative elements
or motifs, we also note differences in decorative scheme on vessels of
different sizes. It would also seem that certain decorative schemes are
restricted to different categories of vessel (see Figs. 6.9, 6.10, 6.11
and 6.12).
In the later phases of settlement, associated with structure 8, we see
a simpler decorative scheme employed. Large vessels are decorated by
simple cordon. Medium-size vessels are decorated with a scheme that
involves the repetition of incisions. The incisions are used either deeply
or lightly on the surface of the vessel and cover the whole surface to give
an all-over incised appearance (see Fig. 6.20).

Production and procurement at Barnhouse
Having characterised the nature of the pottery assemblage at Barnhouse,
I will consider the nature of production strategies in relation to issues
A biography of ceramics in Neolithic Orkney 123

Figure 6.9 Large Grooved ware vessel from Barnhouse

of social identity. The primary step in this analysis involved examining
the contextual differences in pottery. First, I examined the ˜demography™
of vessels in each of the houses at Barnhouse. This indicated regularity
in the frequencies of vessels of each category (see Foster 1960). The
˜population™ of vessels for each house consisted of a single large vessel,
around ¬fteen medium-size vessels, and one or two small vessels. What
is more, the decoration on vessels from all houses in the earliest phase of
settlement was identical. Each house was using vessels that were deco-
rated using the curvilinear or serpentine decorative scheme for medium
vessels and the complex decorative scheme with the addition of dot mo-
tifs for small vessels. This immediately provoked questions concerning
the relationship between production and social identity. Was pottery pro-
duced at the level of the household or was the mode of production more
communal in nature?
124 Archaeological theory and scienti¬c practice

Figure 6.10 Two medium-size vessels from Barnhouse with character-
istic decorative schemes
A biography of ceramics in Neolithic Orkney 125

Figure 6.11 Medium-size vessel from Barnhouse with serpentine
applied cordons

I began investigating the problem of production by examining the de-
posits in the central activity area (Fig 6.13). The de¬nition of pottery
production sites in British prehistory is notoriously dif¬cult. In order to
de¬ne a pottery production site, Wardle (1992) lists a series of attributes
such as the presence of wasters, artefacts used in production, raw ma-
terials, structural evidence for the curing/mining of clay and the presence
of distinctive manufacturing assemblages. According to these criteria, the
central area at Barnhouse bore all the hallmarks of a pottery production
site. The deposits in this area were focused around a large stone slab with
a high magnetic susceptibility reading. In close proximity to this stone
was a clay-¬lled pit. Around the stone there were spreads of ash and
burnt bone, pieces of pumice, polished pebbles and large numbers of
pottery sherds. These sherds consisted of wasters, broken during ¬ring,
and abraded sherds. Radiocarbon dates from this area indicated the area
was used throughout the occupation of the earlier houses. However an ex-
amination of the fabric of the sherds from this area found that 89 per cent
were shell-tempered (fabric C). If this was a production site, then activity
at this site did not account for the production of all vessels.
126 Archaeological theory and scienti¬c practice

Figure 6.12 Sherds from small vessels from Barnhouse with passage
grave art motifs

Analysis of over 180 petrological thin-sections of Grooved ware vessels
from individual houses at Barnhouse showed that the stone and shell
used to temper pottery was employed in a number of speci¬c ways. Those
houses situated at the centre of the settlement, houses 1, 6, 11 and 12,
only contained pottery tempered with shell. This was in stark contrast
to the houses at the periphery of the settlement, houses 2, 3, 5 and 9,
that employed rock temper (see Fig. 6.14). So, pots of different fabric were
Figure 6.13 Plan of the central area at Barnhouse
128 Archaeological theory and scienti¬c practice

Figure 6.14 Schematic diagram indicating the distinction in the use of
temper between inner houses and peripheral houses at Barnhouse

related to different houses. At a more detailed level, petrological examina-
tion of the pottery from the peripheral houses “ those using rock temper “
indicated that each house was employing its own speci¬c temper ˜recipe™
of sandstone and rocks from igneous dykes (Table 6.1). This suggests
that in these houses pottery production was a household-based activity.
Having observed the use of distinct tempering strategies or temper
˜recipes™ in the houses at the periphery of the settlement, it remained to
examine the provenance of these rock sources. A provenancing project
was undertaken in the environs of Barnhouse. The rocks used in the

Table 6.1 Presence/absence of tempering agents in different
houses at Barnhouse

House 2 House 3 House 5 Structure 8

Sandstone Yes Yes Yes Yes
Siltstone Yes Yes No Yes
Mudstone Yes Yes No No
Dyke rock source 1 No No Yes Yes
Dyke rock source 2 Yes Yes Yes Yes
Dyke rock source 3 Yes Yes Yes Yes
Dyke rock source 4 No Yes No Yes
Dyke rock source 5 No Yes No No
Dyke rock source 6 No No No Yes
A biography of ceramics in Neolithic Orkney 129

Figure 6.15 The location of dyke rock sources in the Barnhouse

Barnhouse pottery were predominately from localised igneous dykes that
outcropped at discrete points on the loch (or lake) edge. The provenance
project indicated that the sources of the rocks used in each household
were located less than 1 km by foot or boat from the settlement (Arnold
1985, 45“6); a more detailed analysis of their source location indicated
that they were derived from a number of signi¬cant places (Fig. 6.15).
130 Archaeological theory and scienti¬c practice

Two sources were located in proximity to the earlier Neolithic chambered
tomb at Unstan, while another was located in proximity with the Later
Neolithic artefact scatter at Bookan (Calder 1931), which is likely to
represent another contemporary settlement site. In one instance, then,
stone was related to a place of the dead, likely to be associated with the
identity of a speci¬c ancestral group; in the other instance, stone was
linked with the identities of the members of another settlement. In both
of these cases, people are implicated in the land through rights of access
to, or ownership of, resources.
Crucially, it was not simply individual rock sources that were employed
in tempering vessels. Rather, rock sources from different locations in the
landscape were combined together in speci¬c ˜recipes™ in the production of
vessels within individual houses (see Table 6.1). This suggests that the
act of combination itself was constitutive of the expression of social iden-
tity and provided a metaphor for the creation of links between different
households and communities. This is most obvious when we examine the
Grooved ware from structure 8, which represents the latest phase of oc-
cupation at Barnhouse. The pottery associated with this house contained
all those rock sources that had previously been employed in each of the
earlier houses (Table 6.1). The production of Grooved ware in relation
to this later house would appear to represent the notion of communal
production and sharing in a very concrete manner.
As noted above, each house in the settlement utilised a similar suite
of vessels, and although shell-tempered pottery was a component of
the houses at the settlement™s periphery, no rock-tempered vessels were
found in the houses at the centre of the settlement. Despite the distinc-
tion between the organisation of production related to shell-tempered
and rock-tempered vessels, and the distinctions related to different house-
holds, each house decorated their pottery in precisely the same way. Im-
portantly, the common decorative scheme used on small (Fig. 6.12) and
medium-size (Fig. 6.11) Grooved ware vessels suggests the expression
of communal identity. This practice is not unique to Barnhouse; in fact,
during the earliest phases of the Later Neolithic it would appear that each
settlement in Orkney was decorating pottery according to a settlement-
speci¬c scheme. Although the primary production of pottery involved
the complex articulation of identities through the medium of tempering
strategies, it is important to note that through the subsequent process of
slipping, burnishing and decorating vessels the social relations involved
in primary production become hidden. The memory of these relations is
embedded in the fabric of the pot.
Social relations are therefore inscribed on the surface of the vessel.
Pottery decoration involves an active process in which similarity and
A biography of ceramics in Neolithic Orkney 131

difference are articulated through the medium of particular decorative
schemes. Through materials and techniques the technology of pottery
production at Barnhouse involves the complex articulation of notions of
identity. The production process is twofold, and two distinct ˜layers™ of
memory are bound up in primary and secondary production. The mem-
ory associated with primary production remains hidden and must be
retained and transmitted from one individual to another through teach-
ing and learning (see Arnold 1989), while the memory associated with
secondary production is highly visible and is communicated by visual ob-
servation. The hidden aspect of memory is associated with the dyke rocks
whose location is highly localised within the landscape, while the visible
aspects of memory are linked to the decorative motifs displayed on the
surface of the pot.

The consumption of Grooved ware at Barnhouse
The next stage of my analysis focused on the manner in which different
categories of pottery were related to different consumption practices. In
order to understand the nature of consumption practices across the set-
tlement, a number of samples of pottery of each category were analysed
using the technique of Gas Chromatography/Mass Spectrometry. Most
of these samples were derived from the houses, from midden deposits
close to the houses and from the central area.
Residue analysis relies on the identi¬cation of speci¬c biochemical
compounds, and in the case of the Barnhouse assemblage, the lipid or
fat group was considered to be the most amenable to preservation. Once
lipids migrate into the ceramic matrix of a vessel due to their hydrophobic
properties they are retained (for further details see Evershed et al. 1990,
1992; Jones 1986). The identi¬ed residues are generally in the form of
fatty acids, the basic ˜building block™ by which more complex lipids are
made up. The diagenesis of complex fats to fatty acids occurs as the re-
sult both of cooking and of natural decay over time (Evershed et al. 1992,
203). The identi¬cation of speci¬c fatty acids is feasible and may provide
a signature, or ˜taxonomic marker™ indicating the species origin of the
food. In order to provide a set of standards for identi¬cation, I analysed
the food remains from contemporary settlement sites in Orkney. This al-
lowed me to de¬ne a series of possible food species. Samples from each of
these species, both plant and animal, were then analysed using GC/MS,
and these provided known standards by which to compare the samples
derived from the Grooved ware.
A total of forty-¬ve sherds were examined, and most of them showed
positive evidence of use. The results are presented below, and in each case
132 Archaeological theory and scienti¬c practice

Table 6.2 Simpli¬ed results of GC examination of sherds from large Grooved
ware vessels at Barnhouse

Large vessels

Sherd number Context Contents

SF 1554 House 3, ash dump Milk
SF 1564 House 3, ash dump Wheat/barley
SF 1586 House 3, ash dump Unidenti¬ed plant material
SF 1589 House 3, ash dump Bark resins, milk
SF 1685 House 2, NW alcove No evidence for fatty acids
SF 1812 House 3, ash dump Milk
SF 1827 House 3, ash dump Wheat/barley
SF 2000 House 3, below dresser Barley only
SF 4227 Structure 8, exterior ditch Wheat/barley, milk
SF 4246 Structure 8, exterior ditch Unidenti¬ed plant material
SF 5053 Structure 8, platform Unknown
SF 5299 Structure 8, platform Milk
SF 5618 Structure 8, ditch Milk
SF 5662 Dump near central area Milk
SF 6218 Dump near central area Milk and unidenti¬ed sugar
SF 1839 Structure 8, interior Barley only
(Complete vessel) (set into ¬‚oor)

a simpli¬ed list of the food type found in each sherd is presented
(Tables 6.2, 6.3 and 6.4). This is not the context for the discussion of the
precise details of analysis, since more detailed discussions may be found
in Jones (1997) and Jones et al. (forthcoming). Here I simply want to
discuss the integration of these results with the wider analysis.
Overall the use of Grooved ware at Barnhouse is both complex and
structured. Large vessels appear to form a coherent group of vessels
(Table 6.2). The major food found within these large vessels is dry food
such as barley, although they also appear to be used for the temporary
storage of milk. The use of medium-size vessels is more problematic.
Many vessels seem to contain milk, and these are vessels of both fabric A
and C; however, it would also seem that cattle meat is consumed within
vessels of this size (Table 6.3). Plant material is also obviously utilised in
these vessels, but its origins are dif¬cult to determine. Given the fact that
the small vessels were sampled from three quite different contexts, these
vessels would appear to be clearly and singularly associated with barley
(Table 6.4).
In order to draw out how these foods were prepared and utilised, a
further stage of analysis examined the presence or absence of sooting as
an index of cooking (Hally 1983). The pattern for the site as a whole in-
dicates that sooting predominates on sherds of fabric C. However sooting
A biography of ceramics in Neolithic Orkney 133

Table 6.3 Simpli¬ed results of GC examination of sherds from medium
Grooved ware vessels at Barnhouse

Medium vessels

Sherd number Context Content

SF 10 Ash dump house 2 Cattle stomach
SF 165 Old land surface Unidenti¬ed plant
SF 1080 Old land surface Milk
SF 1577 Old land surface Unidenti¬ed plant
SF 1650 Pit in W recess house 2 Unknown
SF 1655 Ash around E hearth house 2 Cattle meat
SF 1665 Hearth ¬ll W hearth house 2 Cattle meat, milk
SF 1829 Ash dump house 3 Milk
SF 2032 Old land surface Milk
SF 2522/1905 Occupation deposits house 3 Unknown
SF 2547/3477 Occupation deposits house 3 Milk
SF 2578 Occupation deposits house 3 Unknown
SF 3727 Occupation deposits house 6 Unknown
SF 5511 Occupation structure 8 platform Unidenti¬ed plant
SF 5587 Occupation structure 8 platform Unidenti¬ed plant
SF 5607 Occupation structure 8 platform Unknown
SF 5697 Ash spread central area Milk, wheat/barley
SF 5855 Occupation structure 8 interior Unknown


. 4
( 7)