Joumal of Mediterranean Archaeology 22.1 (2009) 101-126
ISSN (Print) 0952-7648
ISSN (Online) 1743-1700
Beyond Ethnicity: The Overlooked Diversity of Group Identities
Naoise Mac Sweeney
Fitzwilliam College, University of Can:\bridge, Storey's Way, Cambridge CB3 01X3, UK
E-mail: nm277@cam.ac.uk
Abstract
This article challenges the current teruiency in archaeology to assume an ethnic basis for group identity.
Archaeology has rehabilitated the concept of ethnicity over the last decade, embracing a theoretically
sensitive model of it as both socially constructed and socially constructing, as flexible, embodied and
hybridised. The success of this model has been such that group identities are often assumed to be ethnic
without investigation. Group identity, however, can relate to many types of perceived commonality and
we must kam to look beyond ethnicity, viewing it as only one amongst many potentia/i;y saiient social
factors. This article advocates the active investigation of group identity and the complex social rationales
that lie behind it. h takes as a case study the site of Beycesultan in western Anatolia, illustrating
the d;yTkimism of group identities in the Late Bronze and Early Iron Ages, and /lighiighting how group
identity crystallised only at particular historical moments aTid around social rationaies which were not
primarily ethnic.
Keywords: ethnicity, group identity, Anatolia, Late Bronze Age, Early Iron Age
Introduction 1999). Iti general, LBA-EIA westem Ana,„,
A 1 1 . 1 > TT tolia is seen as a patchwork of diverse but
Western Anatolia is a place in-between. His- . . . u • u u i i
, , * , , , distinct groups, each with an ethno-cultural
torians and archaeologists alike characterise ., . I L L I .. J I I J '
^ identity which can be plotted along a sliding
the region as being caught at the interface ^^^j^ somewhere between 'Greek' and 'Near
between east and west, Europe and Asia, the ^^^^^^^^, j ^ - ^ perspective does not take into
Greek world and the Orient. The societies ^^^^^^^ ^^^ ¿.^^^^^ ^^^^^ ^^ identities which
that lived here are usually considered to be ^^^^^ g^^^^p^ potentially may have had, nor
mixed and peripheral, and are often discussed ^^^^ ¡^ consider that what initially appear
in terms of hybridity and cultural encounter. ^^ ^e social units may not have had any
This is especially true of the Late Bronze conscious group identity at all. The case of
and Early Iron Ages (hereafter LBA and western Anatolia therefore highlights a seriElA),
when it is assumed that the region was QUS problem with the way that archaeological
pulled between two poles: the Mycenaean/ theory currently approaches the issue of group
Greek Aegean to the west, and the Hittite/ identity—its narrow focus on ethnicity.
Phrygian states of central Anatolia to the The study of past ethnicities is one of the
east (e.g. Bryce 2006: 77-86; Macqueen 1999: pillars of contemporary theoretical archaeol-
39; Márchese 1986: 24; Mee 1998; Niemeier ogy, and a general consensus has been reached
© The RniciforMediierraneanArchaeology/Equmox Publishing Ltd.. 2009 ' doi: 10.1558/jmea.v22Íl.l01
102 Mac Sweeney
conceming how ethnicity should he defined
and approached (Diaz-Andreu et al. 2005;
Gamble 2001; Insoll 2006). It is now widely
accepted that ethnicity is related to notions
of putative descent rather than 'facts' of biological
descent; that it is actively constructed
rather than naturally 'given', flexihle rather
than fixed, disparate rather than monolithic.
This dynamic way of thinking about ethnicity
has proved extremely fruitful for archaeology,
and has been the basis for many important and
innovative studies over the last decade (e.g.
Dusinberre 2003; James 2001; van Dommelen
1998; 2002).
The current approach has proved so successful
that ethnicity now dominates much
thinking about group identity in archaeology.
Just as rank and gender have become the most
commonly cited forms of identity dividing
social groups, ethnicity is the most frequently
discussed form of identity that pulls social
groups together (Diaz-Andreu et al. 2005: 6;
Insoll 2006: 2). In fact, ethnicity has become
such a compelling interpretive paradigm that
recent overviews of archaeological theory
have explicitly stated that other types of group
identity should be considered under its broad
banner (Hodos 2006: 17; Jones 2006: 44; Lucy
2005:101; Meskell 2006: 25). At the moment,
therefore, one sees a trend in archaeology
whereby group identities are often assumed to
be ethnic, even if there is no specific evidence
to suggest this (e.g. Barber 1999; Bryce 1989;
Marcus 2000). The case of western Anatolia
is one example where this tendency is particularly
common, but this is a trend that has also
been identified and critiqued more broadly
(Broodbank 2004; Shanks and Tilley 1987:
93-94).
The widespread prevalence of this ethnic
assumption is unjustified, given comparanda
from modern and historical societies. Contemporary
experience tells us that groups can rally
around an almost infinite range of factors, from
professional occupation to political persuasion,
from sporting allegiance to religious conviction,
from sexuality to class. In short, ethnicity
does not necessarily have to be invoked
for a group identity to form. In archaeology,
however, this variety of differing identities has
been largely overlooked. It seems that while
the revival of ethnicity in archaeology has
clearly been a valuable development in itself,
it may well have dampened our receptivity to
other types of group identity.
In this study, I argue that it is not satisfactory
simply to assume that group identities are
ethnic in nature. Instead, we should set out
to investigate the conceptual basis for each
specific group identity, determining whether
this might be religious, linguistic, ethnic or
related to any other type of social criteria. In
the following section, I examine the reasons
why ethnicity has thus far eclipsed the study
of other forms of group identity in archaeology,
and then explore the diverse nature of
group identities more generally and the way
in which tliey are constmcted. 1 then return
to the case of LBA-EIA western Anatolia,
exploring the fluid and changeable nature of
group identities at the site of Beycesultan, and
considering the diverse social rationales which
may have informed them.
Beyond Ethnicity—Group Identities
Ethnicity and Archaeology I
Ethnicity can be defined as a form of group
identity where membership is determined
by putative kinship and beliefs about shared
descent. While not all definitions include
this as an essential criterion, those which
exclude it are unhelpfully vague, and fail to
distinguish between ethnicity and other forms
of group identity (Emherling 1997: 302-303;
Hall 1997: 26-27). Given this hasic definition,
it is generally agreed that ethnicity is both
constructed through the active use of cultural
traits, and influenced by the normative effects
of culture on the individual (Eriksen 1993;
© The FundforMtditerranean Atchaeolngy/Equinov Publishing Ltd., 2009
Beyond Ethnicity 103
Jones 1997; Weber 1978). This conception of
ethnicity was developed in the late twentieth
century as part of a wider movement in the
social sciences that viewed identities as social
constructs, and is now common across a wide
range of disciplines, including archaeology
(Jones 1997), anthropology (Eriksen 1993),
literary studies (Bhahha 1994) and history
(Pohl and Reimitz 1998).
As already noted, in archaeology it is frequently
presumed that group identities are
formed on a primarily ethnic basis. This
assumption can be partly explained by the
success of the archaeology of ethnicity as an
interpretive model. The roots of this assumption,
however, stretch further back into disciplinary
history. Well before a coherent
archaeology of ethnicity emerged, the term
'ethnicity' was explicitly being used as a general
catch-all expression for a wide range of
group identities articulated in material culture
(Hegmon 1992: 527; Shennan 1989: 14;
Stemer 1989). At the most basic level, this
recurring tendency to privilege ethnicity over
other forms of social identity stems from the
fundamental sociological models espoused by
archaeologists and other social scientists over
the last two centuries (Smith 2006).
Until the mid-twentieth century a modernist
model dominated social thinking, rooted
in the ideas of race (Malik 1996), nation
(Hobsbawm 1990) and environmental determinism
(Peet 1985). Tlie model promoted a
view of human societies as naturally divided
by biological race, with each racial group
allotted a different primordial 'homeland'.
The prevailing social thought of the time
therefore asserted that identities of descent
and geographical territories should map onto
each other in a direct one-to-one relationship.
Tliese concepts were at their most influential
during the late nineteenth and early twentieth
centuries when archaeology was first emerging,
and so this geo-ethnic model of human
societies made a deep imprint in the discipline
as part of the 'culture-historical' school (DiazAndreu
2007; Jones 1997; Trigger 1989; Watkms
2008).
This social paradigm was undermined in the
mid' to late twentieth century, when issues
of self-definition, identity and social meaning
came to the fore. While these new models still
divided human populations into descent-based
units, they no longer presumed that descent
was a natural 'given' but instead viewed it as
socially constructed (Anderson 1983; Bhabha
1990). Homelands became viewed less as
bounded geographical territories and more
as imagined space (Appadurai 1996; Smith
1999). Cultural traits came to be seen as the
means through which identities are constantly
made and remade, and equally as the socialising
influences acting upon the individual hahitus
(Bourdieu 1977). Overall, ethnicity came
to be viewed as hybridised, fragmented, and
dynamic, gaining meaning through encounter
with others and relating flexibly to real and
imagined space. This approach suggests that
identities of place and descent are linked in
a series of flexible many-to-many relationships,
rather than in a simple one-to-one correlation.
Archaeology caught up with these
developments in social theory relatively late.
It was not until 1997 that several important
works were published which applied contemporary
theory to the ancient world (e.g. Jones
1997; Emberling 1997; Hall 1997). Following
on from these, the last decade has seen an
exciting revival of ethnicity in archaeology,
with much valuable and groundbreaking work
being done (see above).
Within the westem academic tradition,
therefore, there is a deep-rooted sense of
a connection between descent-based social
units and geographic space. This is particularly
true of archaeology, because it is by nature a
spatial discipline. In archaeology therefore,
there is a recurring tendency to assume the
existence of a group identity from spatial
variations in material culture, to presuppose
© The Fund ior Mediterranean Aichaeulogy/Equinux Publishing Ltd., 2009
104 Mac Sweeney
an ethnic basis for it, and then to use this as
a starting position ÍTom which to investigate
the ethnic dynamics of the past. There are two
serious prohlems with this: first, inferring the
existence of a conscious group identity from
geographic clusters of similar cultural traits
and, secondly, assuming this group identity
relates to putative descent and ethnicity.
A new approach is needed to broaden this
focus, allowing a wider variety of group identities
to be explored through the archaeological
record. Such an approach must establish first
whether or not there is a conscious sense of
collective identity within a given social group,
and only then go on to consider what the
basis for this sense of collectivity might be.
hi many archaeological studies of ethnicity,
these two distinct operations—determining
the existence of a group identity and interpreting
its social meaning—are conflated into a
single process. This is clearly not a problem
in instances where an ethnic consciousness
is already attested in texts, ethnography or
through other media. In many archaeological
contexts, however, no such ethnic consciousness
can bt; presumed from the outset. In these
cases, the two processes should be separated,
and the social meaning of a group identity
should only be interpreted after its presence has
been properly established. It is time to unpack
the idea of group identities and consider the
full range of their overlooked diversity.
By explicitly considering group identity
rather than ethnicity, the central focus of
study necessarily shifts, moving from one
particular type of social reasoning that can
offer a basis for group identity to the dynamics
of group identity itself. A group identity
approach must take as its starting point the
sense of collective 'us' which is the essential
criterion for group identity, addressing the
psychological and social processes involved
in the ongoing creation and recreation of this
sense of scilidarity. Only once this is established
can we turn our attention to the various
social factors (of which ethnicity is one) that
may inform group identity.
Group Identities and Archaeology
The essential logic of a group identity lies
in a collective 'us-ness' in relation to or versus
'them' (Barth 1969; Buber 1947; Cohen
2002). This sense ot communal 'us' crystallises
around some kind of perceived commonality—something
that 'we' share in common but
which 'they' do not (Cohen 1985: 116). This
perceived commonality is the 'social rationale'
of group identity; it is the ideology on which
group belonging is based, the social logic that
determines membership or exclusion. It is
therefore simultaneously inward- and outwardlooking,
concerned with both the commonality
between group members and with their
differences from non-members. Although the
ideas of 'them' and 'us' are essentially two sides
of the same coin, to date more scholarly attention
has been paid to the former rather than
the latter (e.g. Cohen 2000). There have been
some moves towards correcting this imbalance
in anthropology and sociology (Jenkins 2004:
108), yet the archaeological focus remains
firmly on issues of boundary, distinction, and
the sense of'them' rather than the sense of'us'
(Knapp 2008: 34-35).
In archaeology recent theoretical trends
have concentrated on identities that stress differentiation,
including rank, gender, sexuality
and status (e.g. Diaz-Andteu et ai 2005; Gamble
2007; Insoll 2006). The concepts of agency
and phenomenology have contributed greatly
to this discussion, highlighting in particular
the ways in which personhood and the individual
are constructed through social practice
(Dobres and Robb 2000; Fowler 2004; Knapp
and van Dommelen 2008; Tilley 1994). This
work highlights how social identities can act
to separate and distinguish, teasing out the
way differences between people can be played
up. In such cases, these social identities do not
necessarily require any sense of collective 'us'
© Tile Fund for Meditenanciin Archacology/Et|uiniix Publishing Ltd., 2009
Beyorui Ethnicity 105
to be actively stressed, as they cati be espoused
singularly by individuals. The main determinant
of identity here is the negotiation of difference,
and the sense of'them'.
Group identity, in contrast, necessarily
involves a set of people, and cannot he
espoused by lone individuals. It is a form of
social identity that emphasises a sense of
togetherness just as much as a sense of distinction,
that relies on a feeling of 'us' just as much
as a feeling of 'them'. While some scholars
have used 'social identity' to refer to this sense
of collective 'us', 1 would argue that this term
is too broad. Frequently it is used to refer to
the composite identities attached to individuals
that incorporate diverse factors such
as gender, status and ethnicity at the same
time. I therefore use the term 'group identity'
here. Investigating group identities requires an
understanding of how broader social identities
can be used to unite people, and how social
differences can be played down strategically to
create this sense of solidarity. Such collective
identities are flexible and often fleeting social
constructs; they become salient in specific historical
situations for a specific set of social reasons.
The moments in time at which a group
identity crystallises, and then subsequently
becomes less salient, are therefore of crucial
interest. When and why does the idea of the
group as a corporate whole come to the fore
over the various different forms of identity
that intersect it, such as rank, gender or age?
What kind of circumstances lead to ideologies
of communal togetherness trumping those of
intemal social division? And in what situations
do individual identities and the desire
for social differentiation return to being more
salient than communal group identity?
Far from assuming the non-existence of
individual identities and social differentiation,
such an approach actively questions how
and why these identities can be temporarily
subsumed beneath a wider sense of communal
togetherness. It also accepts that group identity
can crystallise around a broad range of
different social rationales. Ethnicity, religious
inclination, political preference, economic
capacity, linguistic similarity, occupational
activity, or even a comhination of these factors
may be relevant. The potential diversity
of group identities is almost infinite.
Investigating this diversity in the archaeological
record presents some inevitable prob'
lems. A sense of collective identity does
not automatically arise from the experience
of kinship, co-residence, or frequent social
interaction. Instead it is constructed in the
dialogue between social practice and social
experience. The idea of identity as a social
construct has been discussed widely elsewhere
(e.g. Barth 1969; Canuto and Yaeger 2000;
Cohen 1985; Diaz-Andreu et al 2005; Jones
1997; Knapp and van Dommelen 2008), and
will not be treated in detail here.
Group identity (like other forms of social
identity) is therefore a form of ideology that
must be actively created, negotiated and
ascribed to through social practice. Such
practices both act to produce and simultaneously
are the products of a sense of 'us', and
it is only through the archaeological traces of
these practices that a collective identity can
be inferred. What Yaeger (2000: 125) calls
'practices of affiliation', however, can take
many varied forms and can have many different
levels of archaeological perceptibility.
At one end of the scale, textual or epigraphic
evidence sometimes preserves explicit verbal
statements of group unity, as in the classical
Greek world (Hall 1997). Slightly less easy to
interpret is the evidence for highly symbolic,
large-scale group activities, which can serve
to promote a sense of corporate helonging
through participation in socially significant
events such as religious ritual or feasting
(Wright 1994: 61; DeMarrais et al. 1996:
31). Harder again to unravel is the use of
linguistic or material culture styles within and
between groups. Such styles, when deliber©
The Funi! for Meilitetr.uiean Archaeolt^/Equinox Publishing Ltd., 2009
106 Mac Sweeney
ately deployed in specific social situatiotis, can
act as visual symbols marking the commonality
between group members and articulating
their difference from outsiders (Bennet and
Davis 1999; Terrell 2001).
The precise form that such practices take
necessarily varies with the needs and nature of
the group in question. To promote the ideology
of shared belonging, such practices must
emphasise the social rationale of group identity,
accentuating whatever perceived commonality
lies at its root, and elevating it above the
other social factors that might otherwise work
to differentiate between members of the group.
Investigating gtoup identity, then, involves a
close examination of 'practices of affiliation'.
Only from these cati we infer a conscious group
identity, and only ftom these can we determine
the social rationale that underpinned
this identity. Adopting this approach avoids
the current problem where group identities are
unthinkingly ascribed to patterns in material
culture, and where ethnicity is assumed to lie
at their root. As mentioned above, this problem
is particularly notable in the case of LBAEIA
westem Anatolia, where sites are often
assigned to positions along a Greek-Oriental
scale of ethnic identity. Westem Anatolia is
therefore an especially apt case ftom which to
consider the both the changeability and the
diversity of group identities.
Group Identities at Beycesultan in Westem
Anatolia j ; \
Be}icesuitan and Westem Anatolia
LBA-EIA western Anatolia was populated
by diverse, interrelated, and constantly shifting
groups. Political groups made and broke
alliances, ethnic groups expanded and contracted,
and elite groups rose and fell. In contemporary
Hittite documents, the region is
portrayed as a fluid patchwork of independent
polities, engaged in complex and ever-changing
networks of trade, communication and
diplomacy (Beckman 1996; Bryce 2003; 2006;
S e a
Figure 1. Map of Anatolia and the Aegean.
©Tlie Fund for Mediterranean Archaeobf^/Equinox Puhliáitng Ltd., 2009
Beyond Ethnicity 107
Figure 2. Beycesultan viewed from the south.
Heinhold Kramer 1977; Hawkins 1998). The
archaeological record largely supports this
picture, as both material culture styles and
cultural practices show dense micro-regional
patterning (Basedow 2002; Bayne 2000; Mee
1998; Mountjoy 1998). Despite the conventional
view that western Anatolia was caught
between the Greek world and the Near East
(see above), it seems that the politics of group
identity was both complex and nuanced.
Against this colourfxil backdrop, the site of
Beycesultan emerges as an independent and
dynamic settlement, with a population that
coalesced around various group identities at
different times. The site is located on the
Upper Maeander river valley along a major
route between the Aegean coast and the central
Anatolian plateau (Eigure 1), in the midst
of a rich archaeological landscape (Abay and
Dedeoglu 2007). Between 1954 and 1959,
Beycesultan's twin mounds {höyük) were
excavated to reveal a complex urban centre
stretching back to the Chalcolithic period,
where the LBA and ElA are represented by
four architectural levels—Levels III, II, Ib and
la (Eigures 2 and 3). Recording conventions
of the time meant that only a selection of
the finds from these levels was documented.
This selection was designed to include the full
range of material found rather than preserving
relative quantities, and to focus primarily on
the more eye-catching of the small finds and
fine ware ceramics (Lloyd 1955; David and
Lisa French pers. comm.). This study therefore
relies heavily on unpublished material
in the form of archived documents and finds
in museum storage (see acknowledgments).
Eortunately, the excavators' selection biases
were constant in all phases. This permits
comparison between different levels and an
exploration of change over time, but not a full
quantitative analysis of the assemblages.
In the following section, I shall discuss the
four LBA-EIA levels in order, examining each
for evidence of a conscious .sense of collective
group identity, and considering this evidence
in its wider social and historical context to
make inferences ahout the nature of any group
identity identified.
Level III: The Late Bronze Age
The LBA at Beycesultan was a period of stability
and prosperity, with a strong centralised
authority at the site that organised central
urban planning, and with extensive external
contacts that brought in imports from across
the eastern Mediterranean (Figure 4; Lloyd
1972; Mellaart and Murray 1995). There
© The Fumi kv Mediterraiiean Archaeolt^/Equinox Publushing Ltd., 2009
108 Mac Sweeney
-g
o
g
-o
c
%
î
4-
O
f—l
u
•S
Q.
a
u©
The Flint) for Nicdit uiniK PublLshing Lui., 2009
Beyond Ethnicity 109
edge of excavations
edge of excavations
street
Figure 4. Plan of east mound Level III Beycesultan, adapted from Lloyd 1972.
is some evidence for deliberate practices of
group affiliation at Beycesultan, especially in
the form of communal dining.
Commensality can be inferred for two reasons.
First, fine ware dining vessels constitute
a conspicuously large proportion of the overall
ceramic assemblage recovered (Table I). Even
allowing for the bias in recording, the ceramics
from this phase include an unusually large
number of shapes associated with drinking,
serving and pouring. Secondly, these dining
vessels were found concentrated in specific
areas within the site, implying the substantial
scale of the dining group. Most notably,
almost half of all dining vessels were found
within one of the six structures excavated, and
in some of the structures no dining shapes at
all were recorded (Table 2). Taken together,
these two factors suggest that feasting or at
least commensality was practiced in LBA Bey-
cesultan.
Feasting, however, offers a potential venue
not only for the expression of group solidarity
but also for the competitive articulation of
©The Fund tbr Mediterranean Archaeology/Equinox IHiillshingUd., 2009
lio Mac Sweeney
Function
Dining
Of which:
sen ice
comm. serv.
drinking
pouring
Storage
0} which:
small
large ^ccde
uncertain
Cooking
Ritual
Unclear
TOTAL
Level III
#
111
44
20
24
23
9
2
7
5
1
I
127
%
87.4
34.6
15.7
18.9
18.1
7.1
1.6
5.5
3.9
0.8
0.8
100%
Level 11
#
158
51
33
34
40
43
1
30
13
19
-
2
111
%
71.2
23.0
14-9
15.3
18.0
19.3
0.5
13.5
5.9
8.6
-
0.9
100%
Level Ib
#
108
55
14
22
17
13
8
3
2
9
2
5
137
%
78.8
40.1
10.2
16.1
12.4
9.5
5.8
2.2
1.5
6.6
1.5
3.6
1000%
Level la
#
38
u
4
5
6
4
3
1
1
1
5
49
%
11.^
46.9
8.2
10.2
12.2
8.2
6.1
2.0
2.0
2.0
10.2
100%
Table 1. Ceramic vessels by functional category.
individual identities {Dietler 1990; Treheme
1995). In this case, however, it appears that
group solidarity rather than differentiation
was emphasised. This is suggested by the
high level of homogeneity within the dining
assemblage, which not only conforms to a
strictly standardised set of shapes (Figure 5;
Mellaart and Murray 1995) but also to a very
narrow range of decorative styles. Nearly 75%
of all fine wares from this phase make use of
warm-coloured decorative schemes (Types Cl
and C2, see Table 3). It therefore seems that
in LBA Beycesultan commensality brought
together groups of people and stressed a sense
of collective 'us' through standardised and
homogeneous material culture features. In this
way, the differences between group members
would have been played down, and a sense of
cohesion between them played up.
Thus we could argue that evidence exists at
Beycesultan for some form of conscious group
identity. Nonetheless there are also signs of
social differentiation at the site, indicating
that such a group identity was not open to all.
A shrine building was in use during this period,
which implies exclusivity in both its small size
Level III Structures
Structure A
Structure B
Strucrure C
Structure D
Structure E
StnictLire F
Not assoc. with any structure
TOTAL
# of vessels
2
I
28
48
0
0
12
111
% of vessels
L8
0.9
25.2
43.2
0
0
28.8
99.9%
Table 2. Level III dining vessels by excavated structure.
© The FundforMediCenanean Archaeok^/Equinox [\i)lishing Ltd., 2Û09
Beyond Ethnicity 111
Figure 5. Collection of drinking vessels frotn Level III Beycesultan.
and narrow access route (Lloyd 1972). In
addition, while the concentration of dining
vessels point to commensality, the absolute
number involved does not indicate that this
commensality included the entire population
of the site. It therefore seems that the group
identity evident at LBA Beycesultan relates
to an elite group, which demarcated itself from
others through exclusive ritual practice and
communal dining, and articulated an ideology
of similarity between group members by
encouraging a high degree of standardisation
in the material culture used in these activities.
It is also significant that the standardised
material culture favoured by the group only
used characteristically local styles, despite the
presence at the site of several external pottery
styles, including Mycenaean (Mellaart
and Murray 1995). This situation suggests a
deliberate rejection of extemal stylistic influences,
and a strong identification with the
immediate local area—the Upper Maeander
river valley.
Such a phenomenon should be seen in its
wider historical context. During the early
years of the thirteenth century Bc; and around
the time of the construction of Level III, parts
of western Anatolia formed a political and
military alliance known as the Arzawa confederacy
(Figure 6). The Hittite king Mursili
II attacked this alliance and subsumed it into
the Hittite sphere as several separate vassal
states (Beckman 1996: nos. 10-12). Although
official Hittite accounts of Mursili's 'Arzawa
catnpaigns' cannot be taken at face value
because they are pieces of imperial rhetoric,
it seems that there was some Hittite military
intervention around this time, and that afterwards
several westem Anatolian polities did
enter into official alliances with the Hittite
Empire. Hawkins (1998) has convincingly
argued that the Upper Maeander region can
be identified with one of these allied polities,
Kuwaliya. Wliile there is no certain way of
confirming this argument, it is highly likely
that Beycesultan, situated on a route between
© The Fund tor Mediterranean AichacoloBy/Eqiiiunx l\ibl¡shing Ltd.. 2009
112 Mac Sweeney
Type
Al
A2
Bl
B2
Cl
C2
D
E
F
G
H
X
TOTAL
Level III
#
4
16
1
72
24
-
-
I
1
5
Î
127
%
î.i
12.6
-
0.8
56.7
18.9
-
-
0.8
0.8
3.9
2.4
100%
Level II
#
17
55
-
1
52
31
12
26
-
16
12
222
%
7.7
24.8
-
0.5
23.4
14-0
5.4
-
11.7
7.2
5.4
100.1%
Level Ib
#
15
8
5
4
13
34
19
3
10
-
18
S
127
%
10.9
5.8
3.6
2.9
9.5
24.8
13.9
2,2
7.3
-
13.1
5.8
99.9%
Level la
-
1
1
15
21
1
6
-
-
3
1
49
%
-
-
2.0
Z.O
30.6
42.9
2.0
12.2
-
6.1
2.0
99.9%
Key: Al: Silver and gold lusrroiis slipped wares; A2: Coppery-red lustrous slipped wares;
Bl: Linear painted decoration on a lustrous surface; B2: Linear painted decoration on
a matt surface; Cl: Warm-coloured burnished wares; C2: Warm-coloured slipped
wares; D: Grey and black wares; E: White and pale-slipped wares; F: Plastic decoration
or incision; G: Mycenaean; H: Undecorated; X: Unknown decoration.
Table 3. Ceramic vessels by decorative type.
the central Anatolian plateau and the Aegean
coast, would have heen one of the western
Anatolian communities that now came under
Hittite influence.
It is significant that at this time an elite
group at Beycesultan was emphasising its
intemal cohesion and collective identity, and
rejecting external influences in order to anchor
this collective identity in the immediate
locale. Although it is likely that Beycesultan
came under Hittite political influence, Hittite
material culture is almost unknown at Beycesultan.
Tl\e Hittite language of power—the
artistic motifs, visual styles and cult practices
used in the imperial heartland on the Anatolian
plateau—were adopted enthusiastically
in southeast Anatolia and the Amuq as convenient
expressions of elite status (Hawkins
© Tlie Fiind fbr Mcditemuiean Archaeology/Equinox Publishing Ltd.. 2009
1995; Gates 2001) hut are conspicuously
ahsent from Level III at Beycesultan. Instead,
elite group identity is articulated only through
items bearing specifically local associations.
Given this historical context, perhaps it is to
be expected that links with central Anatolia
were deliherately eschewed. What is surprising
is that there are no clear signs of solidarity
with other western Anatolian communities.
After all, this was a time when the region was
first politically united, and then forcibly partitioned.
It is often argued that since Arzawans
were mostly Luwian-speaking peoples, they
would have shared a common ethnic identity
that underpinned their political alliance
(Bryce 2006: 77-86; Macqueen 1999: 39).
Yet the fact that the Beycesultan group constructed
a fiercely local identity precisely
Beyond Ethnicity 113
e a
M e d
Figure 6. Map .showing a rctonstrucfion ot Hittite geography.
during the period of the Arzawa confederacy
undermines this view. If anything, it suggests
that that the alliance was based on calculated
political expediency in the face of a powerful
common enemy, rather than in a deeper sense
of underlying ethnic commonality hetween
westem Anatolian groups.
Level U: The Bronze to hon Age Transition
The following phase at Beycesultan spans the
final decades of the LBA and the first few of
the EIA, and was designated by the excavators
as 'Level IF (Figure 7). The remains show
signs of increased prosperity, as well as continuity
in terms of overall urban layout and in
the use of shrine building. Yet there is also
a sharper sense of social differentiation and
more centralisation in the control of resources
{Lloyd 1972). Amongst the domestic buildings
in particular, there is much more variation
in both architectural elaboration and
size than during the earlier Level III, with the
largest house occupying more than eight times
the total floor space ofthe smallest (Table 4).
While some bouses bad developed into ornate
multi-room complexes with central 'public'
rooms, courtyards and lustral areas, others
continue to be built on tbe standard threeroom
model. This sense of hierarchy and elite
differentiation is also evident in the reduction
of the space given over to domestic buildings
on the summit of the mounds, and the
construction of a stabling complex for horses
instead. The further centralisation of resources
is also implied by the substantial storage facilities
attached to one particular domestic unit
on the east mound (the 'shops'—Lloyd and
Mellaart 1955).
Deliberate expressions of group solidarity
are not readily apparent in this more hierarchical
and vertically structured social milieu.
Although dining vessels still constitute a substantial
proportion of the overall ceramic
assemblage, this is not beyond the proportion
© The Fund for Meditemincui Archaeolüj.'v/Eijuinox PublLshing Ltd., Z009
114 Mac Sweeney
Figure 7. Plan of east mound Level II Beycesultan, adapted frotn Lloyd 1972.
that might be expected, given recording biases.
Similarly, while some of these vessels have
been found in clusters that suggest commensality,
neither the absolute numbers nor the
proportions suggest that such commensality
extended beyond a very limited household
group. Unlike the situation during the previous
phase of occupation, therefore, there is no
indication of large-scale group activities that
would have encouraged a sense of group solidarity
or collective 'us-ness'. Supporting this
overall picture, the ceramic assemblage of this
phase was less standardised. The repertoire of
vessel shapes remains more or less the same,
but the decorative schemes on fine wares
diversified. As well as the warm-coloured
schemes that so dominated the previous phases
(Types Cl and C2), lustrous schemes (Types
© The Fund tbr Medirenanean Archaeobi.'v/Eijiiuiox Publishing Ltd., 2009
Beyor\d Ethnicity 115
Level
Level III
Level 11
Level Ib
Level la
Domestic Unit
A
R
C
D
E
F
A
B
C
D
E
F
A
B
C
n
A
Total floor area (m^)
45
62
68.25
uncertain
55.25
42
122.5
313.75
127.5 (min)
Í5.75
74.22
65.31
35.75
48.75
40.37
38.75
102 (megaron & court)
Table 4. Floor space of domestic units.
Al and A2) also became popular (Table 3).
Although this does not represent an especially
wide range of different styles and still implies
some level of conservatism, it is nonetheless
significant that elite ceramics were no longer
visually homogeneous. ^X4lile Level 111 vessels
encouraged a sense of similarity and standardisation,
this levelling factor is absent ÍTom the
Level n assemblage.
In Level II at Beycesultan, therefore, there
no longer seems to be a coherent elite group,
articulating its sense of collective identity visà-vis
non-elites. Instead there seems to have
been a much more structured social hierarchy,
dividing individual households along a rising
scale of status and wealth. Overall, it seems
that people at this time were much more concerned
with differentiating themselves from
each other than with constructing a sense of a
collective 'us'.
But why should the occupants of Level II
Beycesultan be less concerned about banding
together than they had been during the previous
phase.' The removal of external threats
may go some way cowards explaining this.
During the LBA, the sense of corporate group
identity at Beycesultan seems to have crystalUsed
in response to Hittite imperial expansion
from the east and the integrating pressures
from the Arzawans to the west. In this Level
II phase, however, when these powers were
declining in influence or collapsing completely,
there would have been less external
impetus for solidarity at Beycesultan. The people
at Beycesultan no longer needed to differentiate
themselves from the 'other' without;
©The Fund tor Mediterranean Anrhaeologv'/Equinox Publishing Ltd., 2009
116 Mac Sweeney
they could now tum their attentions to differentiating
themselves from each other within.
The complex multi-leve! hierarchy of Level
II society shows how far Beycesultan went in
abandoning the idea of collective solidarity.
Level îb: The Early iron Age
This imptession strengthens as the EIA
progresses. Level Ib occupation starts some
time during the twelfth century BC, and shows
evidence for continued economic prosperity
and social continuity on the site as a whole
(Figure 8; Lloyd 1972: 17-18). At this point,
however, the similarity between Levels II and
Ib ends. The entrenched social asymmetry of
Level II seems to have been replaced by flexibility
and instability in the social order. First,
the domestic units of this phase are much more
equal in terms of architectural elabotation and
in the amount of space they occupy. Although
a small sample, the four preserved houses are
all two-to-three room structures with a single
main room for most household activities, and
all cover a similar amount of total floor space
(Table 4)- Social hierarchy was no longer
enshrined within the domestic built environ-
ment.
Hierarchy also seems to be absent from
the non-domestic built environment. Significantly,
in Level Ib there are no obvious public
or central structures such as the shrine buildings
or the stable complex, suggesting less centralisation
and a less rigidly structured social
organisation. Rather a large open area to the
east of the domestic buildings was plastered
over with a thick layer of white gypsum and
appears to have been used for ritual purposes.
The large-scale slaughter of animals seems
to have been a feature of this ritual activity,
as two circular pits have been found which
were filled with broken up animal bones and
covered with flat stone slabs (Lloyd 1972:18).
open nlual area
covered in
Willie plaster
Figure 8. Plan of east mound Level Ib Beycesultan, adapted from Lloyd 1972.
© The Fund for Wedirerranean ArchaeokjRy/Eipjinox Publishing Ltd., 2009
Beyorxd Ethnicity 117
Ritual activities on this kind of scale, and in
this kind of open space implies a very different
set of social relations from those suggested by
the shrine building of Levels III and II. A large
open area such as this would encourage inclusivity
and public display, as opposed to the
confined shrines of earlier periods that imply
exclusivity and selective control of participation
(Wright 1994).
The lack of rigid social distinctions and
hierarchy, however, does not necessarily imply
that there was a sense of collective group identity
or corporate solidarity. Instead it seems
that social distinctions were just as important
in Level Ib as in Level II—-the difference,
as implied by the lack of rigid distinctions,
was that status was more flexible and open
to negotiation. While the open ritual area
mentioned above could possibly have been
used for staging practices of group affiliation to
foster a sense of collective 'us', it seems more
likely in this case to have been the setting
for social competition, public display and the
active contestation of status.
There is a new emphasis on jewellery and
bodily adornment in Level Ib, suggesting a
greater focus on the body and personal status
of the individual when compared to previous
phases {Table 5). Not only can a greater proportion
of small finds be classed as 'jewellery'
in this phase than in any other, but it is also
the first time that such items seem to have
been used for the bodily adornment of individuals
within the community (Figure 9). In
Levels III and II, all such objects were deposited
as offerings within the shrine building. If
'jewellery' items had been present in domestic
contexts in these phases, we would expect at
least some of them to have been found in situ,
given the sudden destruction of Level II. In
Level ib by contrast, while some ritual deposits
have been discovered in the open plastered
area, all 'jewellery' items come from domestic,
rather than ritual contexts (Lloyd 1972). This
interest in the body of the individual suggests
a concem with the active creation and contestation
of status within a more flexible and
unstable social structure.
Status competition is also indicated in the
ceramic record, as there is a sudden diversification
of the decorative types in use (Table 3).
Level Ib exhibits the greatest breadth in the
ceramic repertoire of all the phases considered
here. There seems to have been no established
or dominant elite style; instead the wide
variety of aesthetic schemes supports the idea
of competitive display. The previously popular
lustrous and warm-coloured decorative
schemes are joined in this phase by greater
numbers of linear painted vessels (Types BI
and B2), black and grey wares (Type D), and
pale and white slipped wares (Type E). In
Function
Tools
Weapons
Admin
Jewellery
Ritual
Other
Total
Level III
#
36
13
I
10
2
8
70
%
51.4
18.6
1.4
14.3
2.9
11.4
1000%
Level II
#
35
23
2
8
7
16
91
%
3Ö.5
25.3
2.2
8.8
7.7
17.6
100.1%
Level Ib
#
31
6
2
10
2
11
62
%
50.Û
9.7
3.2
16.1
3.2
17.7
99.9%
Level la
#
2
1
-
2
1
1
7
Table 5. Small finds by functional category.
© The Fluid fur Mediterranean Archaeology/Equinox Publishing; Ltd.. 2CO9
118 Mac Sweeney
Figure 9. Jewellery items from the Lt-vcl Ih houses (Melhuirc aiiJ Murray 1995, reproduced with the kind permission
ofthe BIAA).
addition, it also seems that the traditional
Type Cl burnished wares themselves diversify
in terms (if colour, including what Mellaart
evocatively described as "chocolate brown',
'deep plum red', and 'rich camel-leather buft^
(Mellaart and Murray 1995i 56).
Wbile the excavators have assumed that
Level Ib was a period of EIA 'decline' (Lloyd
1972: 5), this evidence suggests that it should
perhaps better be understood as a time of transformation
in the community's social structures,
when old hierarchies were contested, reshaped
and remodelled. Parallel developments elsewhere
in the eastern Mediterranean show a
wide range of responses to this turbulent time
(Bachhuher and Roberts 2008; Knapp 2008),
but Beycesultan's development in the EIA
is characteristically idiosyncratic. While not
immediately affected adversely by the economic
collapses that occurred around 1200 BC,
there seem to bave been some delayed or longterm
consequences of the LBA-EIA transition
at the site, involving deeper changes in scxial
structure and the fundamental concept of elite
identity. The failure of centralised, rigidly hierarchical,
and long-established social systems
in otber parts of the eastern Mediterranean
might have encouraged a more gradual déstabilisation
of the same ideas at Beycesultan,
offering an opportunity for status to he redefined
and reformed in new and different ways.
Any remaining vestiges of group identity or a
collective 'us' seems to have been lost in this
competitive quest for status.
Not only is there little sign of a sense of collective
'us' at Level Ib Beycesultan, but even
bad there been one it seems unlikely that
it would have been informed by ethnicity.
© The FundforMediterranean ArcKaeikigv/Equinox Publishing Ltd.. 2009
Starkly contrasting with the localised preferences
of Level III, the Level Ib assemblage
embraced a broad range of different external
influences, readily incorporating imported
goods and exotica from across the eastern
Mediterranean. Foreign styles and imported
items therefore .seem to have been seen less as
indicators of ethnic affinity, and more as tools
in the competitive discourse of status negotia-
tion.
Level la: The Later Iron Age
Relatively little remains of Level Ia, the final
architectural phase before a gap in occupation
at Beycesultan. A single large megaron house
was found on the east mound, flanked by a
number of outbuildings set around a courtyard
to the west and a large circular haking oven to
the east (Figure 10). The excavators suggested
5 metres
Beyorui Ethnicity 119
that the population had shrunk to very small
proportions, perhaps consisting of only one
extended family (Lloyd 1972: 6). This level,
however, was severely disturbed by later building
and quarrying activity, and traces of other
domestic remains may have been completely
destroyed. The population of Level la may
therefore have been considerably larger than
suggested by the surviving remains. Indeed,
the forms of both the megaron house and the
baking oven suggest that this was so, and both
hint at the return of a sense of group identity
or collective 'us' which had been absent from
Levels II and Ib Beycesultan.
The building itself is large and well constructed,
resembling the elaborate megaron
houses of Level II, and may have been a
central building for a modest community.
Although 'monumental' would perhaps be an
\
open courtyard
baking oven
O
Figure 10. Plan of Level Ia Beycesultan, adapted from Lloyd 1972.
© The Funii for Mtxlitemuiean Archaeokigy/Equindx Publishing Ltd., 2009
120 Mac Sweeney
exaggeration, the megaron is nonetheless a
large and imposing huilding, and is more likely
to he a central meeting place rather than the
domestic quarters of an individual household.
A fragment of a libation arm and a terracotta
basin were also found inside the megaron,
implying that some ritual activities may have
taken place within the building. Indeed, the
architecture of the megaron is deliberately
designed to maximise inclusivity and access
for a reasonably large group of people. The
first chamber is not accessed through a doorway;
instead, one whole wall takes the form of
an open pillared porch. The doorway from the
first into the second chamber is also wide—it
occupies two-thirds of the overall wall space.
The design of the megaron therefore betrays
a concern with inclusivity and participation,
rather than exclusivity and control of access.
This indicates both a group using the building
(even if not all of this group were resident
at the site itselO, and a deliberate concem
with ptomoting collectivity and togetherness
amongst this group. The central location and
sheer size ofthe baking oven (c. 3 m in diameter)
may also suggest the same. The scale
of food production implied by the size of the
oven is beyond that of the strictly domestic,
and may even indicate commensality or at the
very least a strong sense of collective interest
based on shared subsistence provisions.
Both the megaron house and oven therefore
hint at the existence of a coherent group
identity at Beycesultan. The sense of group
solidarity that emerges at Beycesultan in the
later Iron Age, however, is harder to interpret
than the presence or absence of collective
group identities in previous phases. This is
because ofthe relative lack of evidence for the
period, both in the disturbed archaeological
record of Beycesultan and also more generally
in the history and archaeology of the eastern
Mediterranean. This is the time traditionally
characterised as the 'Dark Ages', when social
complexity and interregional exchange are
© The FundforMediterranean Archaeology/Equinox Publishing Ltd., 2009
thought to have decreased, and when local
communities are said to have slid into backwardness
and insularity.
The remains of Level la Beycesultan belies
the first of these characterisations. Social complexity
is implied by the megaron—such a
building could not have been constructed
without the existence of some form of social
organisation (involving authority and specialisation,
if not necessarily hierarchy). Similarly,
it appears that Beycesultan continued to have
access to long'distance trade routes, as this
allowed for the continued import of metals to
the site. It seems clear that social complexity
and interregional exchange did not disappear
in the later Iron Age in this part of westem
Anatolia. Insularity, however, may indeed
have been a feature of Beycesultan's society at
this time. This is meant not in the sense of a
lack of external contacts (as these are attested
by the continued metals trade) but rather in
terms of no longer having an outward'looking
perspective. Despite the site's continued access
to imports and trade routes, the wide range of
extemal styles and influences embraced in the
previous phase narrows dramatically in Level
la, and the decorative schemes used for ceramics
retum to having a distinctly local flavour.
Hie narrow material repertoire is therefore
the result of a conscious decision on the part
of the site's inhabitants, rather than a lack of
options. ¡
Given the lack of evidence about this period,
it is impossible to come to any firm conclusions
about why a conscious group identity
crystallised in the later Iron Age at Beycesultan.
It may have been due to the community
coming together in the face of an external
threat as had happened in previous phases,
or it may be an expression of the population's
insular and inward-looking perspective.
Until more is known about the wider regional
dynamics of this period, however, the specific
social dynamics at Beycesultan will remain a
mystery.
Beyond Ethnicity 121
Conclusions
The most obvious conclusion that can be
drawn from this consideration of the LBA
and EIA site of Beycesultan in westem Anatolia
is that group identities were not a
constant. There was no sense of collective
'us-ness' that consistently bound together
the inhabitants of the site, or even certain
portions of the overall population. During
the LBA phase of occupation, there are signs
that people at Beycesultan were keen to foster
a sense of the collective 'us', downplaying
differences between themselves to promote
cohesion and community. At the time, external
threats such as Hittite imperial expansion
may have encouraged this strongly local sense
of group identity. In the following two phases,
however, this external pressure was relaxed,
and it seems that the people of Beycesultan
became less intetested in maintaining group
solidarity than in negotiating internal status
distinctions. Any sense of a collective 'us'
in the LBA-EIA transitional phases seems
to have been eclipsed by a move towards
greater social differentiation, first rigid and
formalised and later flexible and contested. It
was not until the later Iron Age that a sense
of group identity seems to have re-emerged
in Beycesultan, although it is not yet clear
whether this should be related to the instability
of the time or to an insular, inwardkioking
perspective on the part of the site's
inhabitants. Given the sporadic emergence
and dissolution of group identity at Beycesultan,
it seems that a conscious sense of group
identity only ctystalUses at certain times and
for certain reasons.
This conclusion calls into question the
common assumption in archaeology that sites,
communities, or population groups can be
treated as if they inherently embed coherent
group identities. Part of the problem is
the enduring concern in archaeology about
constructions of the extemal 'them' in social
identities—relatively little attention is paid to
constructions of the ctillective 'us'. While it is
frequently assumed that the former inevitably
leads to the appeatance oi the latter, this is
not necessarily always the case—after all, lone
individuals can construct a sense of 'them'
without any collective 'us' being ptesent. We
cannot therefore assume that any social group
we distinguish from material culture patteming
will necessarily have a conscious sense of
'us'. The Beycesultan example illustrates this
vividly, showing that this sense of'us' does not
necessarily appear as a result of co-residence
in a given settlement, comparable elite status,
or even sharing the same forms o{ material
culture. Instead, the social and historical circumstances
play a major role in whether people
consciously choose to see themselves more
as a collective or more in terms ot distinction
and difference.
A second conclusion to be drawn from this
case study is that even when conscious group
identities do coalesce, they do not necessarily
do so around ideas of shared descent
or ethnicity. Conventional wisdom tells us
that Beycesultan was ethnically related to its
neighbours in the west, and that this ethnic
affinity was translated into a political alliance
in the LBA. The evidence from Level
IU Beycesultan, however, suggests otherwise.
The people here seem to have rejected an
ethnic-based sense of identity, celebrating
instead a sense of group identity which was
informed by a strong sense of geography and
the immediate locale. Ethnic concerns do
not seem to have played an important role in
later phases either; the Level Ib inhabitants
welcomed imports and styles hailing from all
directions, and the occupants of the Level la
phase reverted back to a closely local identity.
In none of the four occupational levels does it
appear that ethnicity was a determining factor
in the social calculations of the people at Beycesultan.
The social dynamics at the site were
far more complex than an approach focused
© The FundforMediterranean Archaeok^/Equinox Pubtuhing Ltd., 2009
122 Mac Sweeney
purely on ethnicity would bave detected. By
consideriníí group identities, bowever, it can
be seen tbat the inhabitants of the site were
actors in a constantly changing social system,
where ethnicity may well have been an important
factor but nonetheless did not dominate.
Thus we cannot assume that ethnicity lies
at the root of all collective group identities.
While the archaeology of ethnicity has offered
many important insights over the last decade,
it is only part of a much broader picture. From
its inception, the proponents of an archaeology
of ethnicity argued for a contextual
viewpoint, and warned against abstracting
ethnicity from the other identities that both
cut across and intertwine with it (Emberling
1997: 319; Jones 1997: 133). In practice however,
the last ten years have shown that this is
easier said than done. Adopting an approach
that explicitly considers group identity may
be one way of moving towards this goal.
This group identity approach recognises that
while ethnicity may be one of the factors that
frequently inform a sense of 'us', it does not
always necessarily do so.
TTie third and final conclusion of this paper
returns to LBA-EIA western Anatolia, and
tbe theme of 'between-ness'. The case of Beycesultan
shows that western Anatolia should
not be considered as merely being 'between' or
peripheral; rather it was home to societies of
interest in their own right. In the LBA, Beycesultan
seems to have exercised substantially
more independence than is usually assumed—
from both the Hittites and from its western
neighbours. But along with this autonomy of
action seems to have come an independence of
outlook, a preference for keeping itself distinct
and maintaining a kind of'splendid isolation'.
Similarly in the EIA, Beycesultan's experience
following the LBA-EIA transition does not
fit the accepted patterns. While 'peripheral'
sites are assumed either to have declined
(alongside the collapsing LBA states), or to
have enjoyed a post-collapse boom (stepping
into vacant political and economic niches),
it seems that Beycesultan did neither. Instead
the transitions had little immediate impact,
and it was only later in the EIA when gradual
changes began to be felt on a more ideological
level, in the shape of greater hierarchical
flexibility and social mobility. The evidence
from Beycesultan prompts us to reassess how
we understand both the LBA and the EIA
in Anatolia, and may have implications for
contemporary developments elsewhere in the
eastern Mediterranean. Neither peripheral
nor 'between', neither Greek nor Near Eastem,
western Anatolia emerges as a region of
independent communities and autonomous
groups. These groups espoused identities that
were constantly crystallising, dissolving and
changing, and while their identities may have
owed something to ideas of ethnicity, they
also went far beyond it. '
Acknowledgments | >
I
This paper draws upon work initially undertaken
for my PhD dissertation at tbe University
of Cambridge, and would not have
been possible without generous support from
the AHRC (Arts and Humanities Research
Council). I owe a great debt to my supervisor,
Robin Osbornc, for his continuing help and
encouragement. 1 would also like to thank
John Robb, Roger Matthews, Lisa French and
Nicholas Postgate for helping me to develop
the ideas contained in both the dissertation
and this paper. I am also grateful to the editors
and to four anonymous reviewers for their
comments and suggestions. Analysis of the
Beycesultan material was kindly facilitated by
the Turkish Ministry ofCulture; Huseyn Baysal
and Salim Ugulmar of the Denizli Provincial
Museum; and Hugh Elton, Gülgün Girdivan,
Gina Coulthard and Yaprak Eran of the British
Institute of Archaeology at Ankara.
© T h e Fund tor Mixiiterranain ArchaH)l(i(jv/Einiii><)X Publishing Ltd., 2009
Beyond Ethnidty 123
About the Author
Naoise Mac Sweeney is a Researcb Fellow at
Fitzwilliam College, University of Cambridge.
Her main research interests are the Bronze and
Iron Ages of Anatolia, and the dynamics of
group identities including ethnicity, community
identity, and migration and its rhetorical
constructs. She currently works at the Kilise
Tepe project (Turkey), and as a consultant for
the Intemational Labour Organisation and
Donor Committee for Enterprise Development
on contemporary ethnic conflict.
References
Abay, E.. and F. Dedeoglu
2007 Yih Çivril Ovasi Yuiey Ara§tirmasi 2005.
Arojtirmd Sonuçlan Topiantisi 24: 277-92.
Anderson, B.
1983 Imagined Communities: Reflections on the Origin
and Spread of Nationalism. London; Verso.
Appadurai, A.
1996 Modernity at Large: Cuituroi Dimensions of
Globalization. London and Minneapolis: University
of Minnesota Press.
Bachhuber, C, and R.G. Roberts
2008 Forces of Transformation: The End of the Bronze
Age in the Mediterranean. Oxford: Oxhow.
Barber, R.L.N.
1999 Hostile Mycenaeans in the Cycladesí In R.
Laffmeur (ed.), Polemos: Le contexte guerrier
en Egée à l'Âge du Bronze. Aegaeum 19: 133-
39. Liège: Université de Liège.
Biirth. F. (ed.)
1969 Ethnie Groups and Boundaries; The Social
Organization of Culture Differeru:e. Boston;
Little, Brown.
Basedow, M.A.
2002 Cemetery and ideology in the west Anatolian
coastal region. In R. Asian, S. Blum, G. Kastl,
F. Schweizer and D. Thumm (eds.), Mauerschau:
Pestsckrift für Manfred Korfman, 469-
74. Remshalden-Grunbach: Verlag Bernhard
Albert Greiner.
Bayne, N.P.
2000 The Grey Wares of North-West Anatoiia ¿n
the Middle and Late Bronze Age and Early Iron
Age and their Relation to the Early Greek Settiements.
Bonn; Asia Minor Studien.
Beckman, G.
1996 Hittite Diplomatic Texts. Atlanta. Georgia:
Scholar's Ptess.
Bentict, J., and J.L. Davis
1999 Making Mycenaeans: warfare, territorial
expansion, and representation of the other
in the Pylian Kingdom. In R. Laffineur (ed.),
Polemos: Le contexte guerrier en Egée à l'Age du
Bronze. Aegaeum 19: 105-120. Liège; Université
de Liège.
Bhahha. H.K.
1990 DissemiNation: time, narrative, and the margins
of the modem nation. In H.K. Bhahha
(ed.), Nation and Narration, 291-322. London
and New York: Routledge.
1994 The Libation of Guitwre. London and New
York: Routledge.
Bourdieu, P.
1977 Outline of a Theory of Practice. Cambridge:
Cambridge University Press.
Broodbank, G.
2004 Minoanisation. Proceetings of the Gamijridge
Philoíogícaí Society 50: 46-91.
Bryce, T.R.
1989 The nature of Mycenaean involvement in
westem Anatolia. HL'itoria 38: 1-21.
2003 Life and Society in the Hittite World. Oxford:
Oxford University Press.
2006 Trojans and Tlieir Neighbours. London and
New York: Routledge.
Buber, M.
1947 Between Man and Man (translated by R.G.
Smith). London: Kegan Paul.
Canuto, M.A., and J. Yaeger (eds.)
2000 The Archaeology of Communities: A New
World Perspective. London and New York:
Rouledge.
Cohen, A.P.
1985 The Symbolic Construction o/Gommunitv. London:
Tavisrock.
2000 Signifying Identities: Anthropoh^cai PerspeC'
tives on Boundaries and Gontestcd Values. London
and New York; Routledge.
2002 Epilogue. In V. Amit (ed.), Reaiii^ng Community:
Concepts, Social Relationships and
© The FuiiJ fiT Mediterrdiiean AKhaw)!ugv/Emúiiux Poblishuig Ltd., 2009
124 Mac Sweeney
Sentiments, 165-70. London and New York:
Roiitledge.
Dawson, A.
2002 Tht- mitimg community and the ageing body.
Towards a phenomenology of community?
In V. Amit (ed.), Realizing Community:
Concepts, Social Relationships and Sentiments,
21'Î7- London and New York: Routledge.
DeMarrais, E., L.J. Castillo and T. Earle
1996 Ideology, materialization, and power strategies.
Current Anavropohgy 37: 15-31. doi:
10.1086/204472
Diaz-Andreu, M.
2007 A World History of Nineteenth-Century Archaeology:
Nationdism, Colonialism, and the Past.
Oxiord: Oxford University Press.
Diai-Andreu, M., S. Lucy, S. Babic and D.N. Edwards
2005 The Archaeology of Identity: Approaches to
Gender. Age, Status, Ethnicity and Reiigion.
London and New York: Routledge.
Dietler, M.
1990 IXven by drink: the role of drinking in
the political economy and the case of Early
Iron Age France, journal of Anthropoh^cd
Archaeology 9: 352-406. doi;10.1016/0278-
4165{90)90011-2
Dobres, M.A., and J.E. Robb (eds.)
2000 Agency in Archaeology. London and New
York: Routledge.
Dusinberre, E.R.M.
2003 Aspects of Empire in Ac/iaemenid Sardis. Cambridge:
Cambridge University Press.
Embetling, G.
1997 Ethnicity in complex societies: anthropological
perspectives. ]oumal of Archaeological Research
5: 295-344. doi:10.1007/BF02229256
Eriksen, T.H.
1993 Ethnicity and Nationalism: Anthropob^cal Perspectives.
London: Pluto Press.
Fowler, C.
Z004 The Archaeology of Personhood: An Anthropological
Approach. London and New York:
Routiedge.
Gamble, C.
2001 Archaeology: The Basics. London and New
York: Routledge.
2007 Origins and Révolutions: Human Identity in
Earliest Prehistory. Cambridge: Cambridge University
Press.
Gardner, A.
2007 An Archaeology of Identity: Soldiers and Society
in Later Roman Britain. Walnut Creek, California:
Left Coast Press.
Gates, M.-H.
2001 Potmarks at Kinet Hüyük and the Hittite
ceramic industry. Varia Anatolica 13: 137-57.
HallJ.M.
1997 Hthnic identity in Greek Antiquity. Cambridge:
Cambridge University Press.
Hawkins, J.D.
1995 The political geography of north Syria and
south-east Anatolia in the Neo-Assyrian
period. In M. Liverani (ed.), Neo-Ass>rian
Geography, 87-101. Rome: Quademi di
Geografia Storica.
1998 Tarkasnawa King of Mira, 'tarkondemos',
Bogakoy sealings and Karabel. Anatolian Studies
48: 1-31. doi:I0.2307/3643046
Hegmon, M.
1992 Archaeological research on style, Annuaí
Revieu'of Anthropology 21: 517-36. doi:10.1146/
annurev.an.2L100192.002505
Heinhold-Kramer, S.
1977 Ar^aim: Untersuchungen TM seiner Geschichte
nach den hethitischen Quellen. Heidelberg: Carl
Winter.
Hobsbawm, E.J.
1990 Nations nnd Nati'onaiismsince 1780: Programme,
Myth, ReaUty. Cambridge: Cambridge University
Press.
Hodos, T.
2006 Locai Responses to Cohnization in the Iron
Age Mediterranean. London and New York:
Routledge. i ^
Insoll, T.{ed,) '
2006 The Archaeology of Identities: A Reader. London
and New York: Routledge.
james, S.
2001 'Romanization' and the peoples of Britain. In
S.J. Keay and N. Terrenato (eds.), Italy and
aie West: Comparative /¡sues in Romani^arion,
77-89. Oxford: Oxbow.
Jenkins, R.
2004 Social Identity. London and New York: Rout-
ledge.
Jones, S.
1997 The Archaeology of Eàinicity: Constructing
identities in the Past and the Present. London
and New York: Routledge.
© The Fund fbr MeiJicenaiieaii Archaeolc^/Equinox Publishing Ltd., 2009
BeyoT\d Eàinicity 125
2006 Discourses of identity in the interpretation of
the past. In T. InsoU (eJ.), The Archaeology 0/
identities: A Reader, 44-58. London and New
York: Routledge.
Knapp, B.
2008 Prehistoric and Protohistoric Cyprus; Insularity,
Connectivity and Identity. Oxford: Oxford
University Press. doi:10.1093/acprof:oso/978
0199237371.001.0001
Knapp, B., and P. van Dommelen
2008 Past practices: rethinking individuals and
agents in archaeology. Cambridge Archaeoh^caljoumal
19: 15-34.
Lloyd, S.
1955 New light on ancient people. The Times, 1
August 1955.
1972 Beycesultan, Volume 3, Part 1. Late Bronze
Age Architecture. London: British Institute of
Archaeology at Ankara.
Lloyd, S., and J. Mellaart
1955 Beycesultan excavations; first preliminary
report. Anatoiian Studies 5; 39'135.
doi:10.2307/3642324
Lucy, S.
2005 Ethnic and cultural identities. In M.S. DiazAndreu,
S. Lucy, S. Bahic and D.N. Edwards
(eds.), The Archaeology of Identity: Approaches
to Gender, Age. Status, Ethnidty and Religion,
86-109. London and New York: Routledge.
Macqueen, J.
1999 The Hittites. Revised edition. London: Thames
and Hudson.
Malik, K.
1996 The Meaning of Race. Basingstoke: Palgrave.
Márchese, R.T.
1986 The Lower Maeander Flood Plain: A Regional
Settlement Study. British Archaeological
Reports, Intemational Series 292. Oxford:
British Archaeological Reports.
Marcus, J.
2000 Towards an archaeology of communities. In
M.A. Canuto and J. Yaeger (eds.). The Archaeology
of Communities: A New World Perspective,
231-42. London and New York: Routledge.
Mee, C.
1998 Anatolia and the Aegean in the Late Bronze
Age. In E.H. Cline and D.R. Harris (eds.),
The Aegean and the Orient in the Second Milienni«m.
Aegaeum 18: 137-48. Liège; Université
de Liège.
Mellaart, J., and A. Murray.
1995 ße^cesuitan, Volume 3, Part 2. Late Bronze
Age and Phrygian Potters and Middíe and Late
Bronce Age Small Objects. London: British
Institute of Archaeology at Ankara.
Meskell, L.
2006 Archaeologies of identity. In T. InsoU (ed.).
The Archaeobgy of identities: A Header, 23-43.
London and New York; Routledge.
Mountjoy, P.
1998 The east Aegean—west Anatolian interface
in the Late Bronze Age: Mycenaeans and the
kingdom of Ahhiyawa. Anatolian Stiidie.s 48:
33-67. doi:10.2307/3643047
Niemeier, W.-D.
1999 Mycenaeans and Hittites in war in Asia Minor.
In R. Laffineur (ed.), Polemos: Le contexte guerrier
en Egée à l'Âge du Bronze. Aegaeum 19:
141-55. Liège: Université de Liège.
Peet, R.
1985 Tlie social origins of environmental determinism.
Annals of the Association of American
Geographers 75: 309-33. doi:10.llll/
J.1467-8306.1985.th00069.x
Pohl, W., and H. Reimitz (eds.)
1998 Strategies of Distinction: The Construction of
Ethnic Communities 300-800. Leiden: Brill.
Shanks. M., and C. Tilley
1987 Social Theory and Archaeology. Cambridge:
Cambridge University Press.
Shennan, S. (ed.)
1989 Archaeological Approaches to Cultural Identity.
One World Archaeology 10. London and
New York: Routledge.
Smith, A.D.
1999 Myths and Memories of the Nation. Oxford:
Oxford University Press.
Smith, A.T.
2006 Undisciplined theory. Archaeoh^cal Dialogues
13: 159-63. doi:10.1017/S1380203806272088
Sterner, J.
1989 Who is signalling whom? Ceramic style,
ethnicity and taphonomy among the Sirak
Bulahay. Antiquité 63: 451-59.
Terrell, J.E.
2001 Archaeology, Language, and History: Essays on
© The Bind for Mediterranean Archaeology/Equinox Publishing Ltd., 2(X)9
126 Mac Sweer^y
Culture and Ethnicity. Westport: Bergin and
Garvey.
Tilley, C.
1994 Space, place, landscape and perception:
phenomenological perspectives. In C. Tilley
(ed.), A Phenomenology of Larviscape, 7-34Oxford:
Berg.
Treheme, P.
1995 TTif warrior's beauty: the masculine body and
self-identity in Bronze-Age Europe. Joumal of
European Archaeology 3: 105-44Trigger,
B.G.
1989 A History of Archaeolo^cal T/iowg/it. Cambridge:
Cambridge University Press.
van Dommelen, P.
1998 On Colonial Grounds: A Comparative Study
of Colonialism ar\d Rural Settiemeni in First
Millennium BC West Central Sardinia. Leiden
University Archaeological Studies. Leiden:
Faculty of Archaeology, Leiden University.
2002 Ambiguous matters: colonialism and local
identities in Punic Sardinia. In C. Lyons
and J. Papadopoulos (eds.). The Archaeology
of Colonialism, 121-47. Los Angeles: Getty
Research Institute.
Watkins, T.
2008 Supra-regional networks in the Neolithic of
southwest Asia, jourrud of World Preiusurry 21 :
139-77. doi:10.1007/sl0963-008-9013-i
Weber, M.
Í978 Ethnic groups, in G. Roth and C. Wittich
(eds.). Economy and Society 1: 385-98. Berkeley
and Los Angeles: University of California
Press.
Wright, J.C. !
1994 The spatial configuration of belief; the
archaeology of Mycenaean religion. In R.
Osbome and S.E. Alcock (eds,). Placing the
Gods: Sanctuaries and Sacred Space m Anâent
Greece, 37-78. Oxford: Oxford University
Press.
Yaeger, J. '
2000 The social construction of communities in
the Classic Maya countryside: strategies of
affiliation in westem Belize. In M.A. Canuto
and J. Yaeger (eds.). The Archaeology of Commurúties:
A New World Perspective, 123-42.
London and New York; Routledge.
© The Fund fra Medilerraneaii Archaeoiogy/Equintix Piijlishiiig Ltd.. 2009