Chera) Chola) Pandya: Using Archaeological

Evidence to Identify the Tamil Kingdoms of
Early Historic South India
SHINU A. ABRAHAM
DESPITE THE AMOUNT OF ARCHAEOLOGICAL RESEARCH that has been carried
out in South India, our understanding of the South Indian protohistoric period
is still, in many ways, in its infancy. It is only recently that scholars have begun
to move away from traditional culture-historical approaches and to draw on the
archaeological data to answer specific questions about the processes and condi-
tions surrounding emerging social complexity. Like many other regions of the
world that possess a literary tradition, South Asia's past has been largely built on
a foundation of epigraphic and textual evidence (e.g., Kulke and Rothermund
1986; Sastri 1966); one could argue that, for the time period spanning the transi-
tion from prehistory to history, it has been South Asia's documentary record that
has chiefly determined current interpretations of its early history. In the same
way, the onset of written records in South India made it possible to move away
from the tendency to treat the entire peninsular region as a whole and to isolate
smaller regions whose texts indicated distinct historical trajectories. In the south-
ern portion of the peninsula-the region that corresponds roughly to the present-
day states of Kerala and Tamil Nadu-the existence of a large documentary cor-
pus, both indigenous and foreign, and the occurrence of inscribed coins and cave
inscriptions, have given rise to the idea of a separate ethnic and linguistic region
known as "Tamilakam" (Fig. 1).
By contrast, the role of archaeology in the consideration of early Tamil identity
has been more or less secondary. The common tendency is for South Indian his-
torians to appropriate the archaeological data as a source of correlates for infor-
mation gleaned from the texts (e.g., Champakalakshmi 1996; Gurukkal 1989;
Thapar 1992)-in other words, to use the material record to search out "known"
historical patterns, events, or places (Morrison and Lycett 1997: 216). Archae-
ologists are equally culpable; it has become customary for South Indian archae-
ologists to label sites and objects in Kerala and Tamil Nadu as "Tamil," without
considering whether signifiers exist in the material record that substantiate or
Shiml A. Abraham is visiting assistant professor in the Department of Anthropology, S1. Lawrence
University, Canton, New York .
.-1.\;,11/ PnSpl'Oil'l'.';, Vol. -J.2. No.2 l(l 2003 by University of Haw:li'j Pn..'ss.
208 ASIAN PERSPECTIVES . 42(2) . FALL 2003
TAMILNADU
o
r
N
I
100 200 km
Fig. 1. States of South India.
refute this notion of cultural separateness. The underlying assumption continues
to be that the documentary record serves as the best and most reliable source for
knowledge about past identity. As will be demonstrated here, the archaeological
data from protohistoric Kerala and Tamil Nadu is not so clear-cut and, in fact,
appears to challenge the very notion of a separate culture region. By carefully
separating and analyzing the written and material records for Tamilakam (Fig. 2),
this study will examine what, if anything, it meant to have a Tamil identity in
Early Historic South India.
ARCHAEOLOGY, HISTORY, AND IDENTITY
The relationship between archaeology and history has become a mounting con-
cern among archaeologists, and research in other regions highlights some of the
ABRAHAM . TAMIL KINGDOMS OF HISTORIC SOUTH INDIA 209
600 km
t
N
o
INDIA
Fig. 2. Location ofTamilakam (c. 300 B.C.-A.D. 300).
issues that are relevant to this study of Kerala and Tamil Nadu. Many archae-
ologists agree, for instance, that the overwhelming tendency is to privilege the
information contained in texts at the expense of data derived from the archae-
ological record (e.g., Feinman 1997; Morrison and Lycett 1997). Conflicts also
emerge when historians focus on the texts at the expense of other materials, while
archaeologists limit their efforts to single-site evaluations (Thurston 1997: 260).
Also, archaeologists themselves can be prejudiced in their selective use of written
records to support their findings (Feinman 1997: 372). Nevertheless, the rationale
for conducting historically informed archaeology is indisputable; combining the
210 ASIAN PERSPECTIVES . 42(2) . FALL 2003
historical and archaeological data sets must lead to richer perspectives on the com-
plexities of the past.
Recognizing that history and archaeology are both media by which the past
can be accessed, the most basic way of linking the two is to use one record as a
ITleanS of identifying or corroborating information recovered by the other. Leone
and Potter (1988) argue that Binford's middle-range theory may be adapted from
ethnoarchaeological studies to the documentary record-in other words, that
textual evidence should be used to set up "descriptive grids" (Leone and Potter
1988: 13), based on the documentary record, against which the archaeological
data can be analyzed. Essentially, the documents would provide testable questions
that the material remains may be able to answer, and the answers in turn will offer
insights into both the documentary and material records and their relationship
(Leone and Crosby 1987: 14). Kosso likens the relationship between texts and
artifacts to the use of testimony and physical evidence in a court trial (Kosso
1995 : 182): each line of evidence can be used potentially to evaluate the other.
Although the two sources are seldom about the same event, peoples, or objects,
they are often about related things (Kosso 1995: 183), and hence may be used in
conjunction with one another to recreate the past. Kosso is careful to acknowl-
edge the lack of complete independence between historical and archaeological
forms of evidence but proposed that, with a systematic enumeration of the dis-
tinctions, it may be possible to compare and contrast the claims of each line
of evidence. Wylie (1985) espouses a similar approach in her discussion about
the use of analogy in archaeology. One kind of link often established between
archaeological and textual records (particularly in South Indian archaeology) is
the one rooted in analogical reasoning, i.e., the use of knowledge from one rec-
ord to make inferences about the second record, based often on assumptions of
uniformitarian principles. Although Wylie is referring to archaeological and eth-
nographic sources in her essay, the same debates apply to archaeological and his-
torical sources.
The interplay of material data and historical texts especially resonates during
the quest for past social identities. In the case of people who have left documen-
tary records, the historical references to specific past groups lead inevitably to the
labeling of sites and objects (Jones 1997: 27). At first glance, trying to locate cul-
tural or ethnic identity may look suspiciously like an outdated attempt to super-
impose old culture-historical taxonomies on early South India. The definition of
ethnicity itself is beset by difficulties. Shennan (1989: 14) states, "ethnicity must
be distinguished from mere spatial variation and should refer to self-conscious
identification with a particular social group at least partly based on a specific
locality or origin." One cannot simply search for artifact patterning in order to
locate ethnicity, any more than one can locate archaeological "cultures," since
this variation arises from the result of an enormous range of different processes.
Conversely, a particular material form may remain the same, but its meaning will
alter in different contexts; it will be used in different ways according to different
historical traditions and social contexts (Shanks and Tilley 1987).
It is commonly assumed that studies of ethnicity require access to people's
self-conscious reflections on identity (Jones 1999: 220), and that written sources
provide the ITlOst reliable accounts of such reflections. Once such issues are
understood through an analysis of texts, the archaeologist may try to search for
material markers of ethnicity in the archaeological record. But the archaeolog-
ABRAHAM . TAMIL KINGDOMS OF HISTORIC SOUTH INDIA 211
ical pursuit of groups named in historical texts gives rise to a new set of issues.
Names recorded in texts are often deliberate records of information about rulers
and bureaucrats, rather than the self-identifications of the general population
(Emberling 1997:313). The boundaries defined in the texts may not match
boundaries found in the material record, and the texts may also obscure ongoing
processes of identity formation and re-formation. The situation is compounded
by the fact that the material record of complex societies may have been manipu-
lated in their turn and may in fact obscure actual relations of power (Morrison
and Lycett 1994: 327). The texts do not provide, "self-evident meaning and can-
not, in themselves, solve archaeology's central methodological challenges" (Mor-
rison and Lycett 1997: 233).
Despite the theoretical concerns and methodological constraints, the influence
of text-based knowledge on the evaluation of archaeological data is undeniable
and even inevitable, at least in the case of South India. It is not possible at this
stage to confidently propose a solution, but perhaps by considering the early
Tamil written and material records separately, one may begin to understand the
informational gap between the two data sets and suggest possible ways to narrow
that gap.
"TAMILAKAM" AS A HISTORICAL ENTITY
The evidence for the identification of Tamilakam as a distinct culture region is
culled from a variety of text-based sources. The most important of these is a body
of prose poetry referred to as the "Sangam Anthology"-indigenous texts dating
to the first few centuries A.D. that comprise the earliest extant examples of Tamil
literature. The works that have survived include eight anthologies and ten idylls, a
work on grammar, and 18 minor works. There are altogether 2381 poems by 473
poets and 102 poems by anonymous authors, besides the grammar and 18 minor
works (Hart 1975: 7). Scholars are divided about the chronology of these works,
and propose dates ranging from the first century B.C. up to the sixth century
A.D. (Hart 1975; Ray 1995; Zvelebil 1992). Appreciating the Sangam corpus as a
source of historical data requires understanding their poetic and bardic nature: a
bulk of the poetry is concerned with extolling the exploits of rulers, warriors, and
patrons (Champakalakshmi 1996: 175)-a fact that has not prevented historians
and archaeologists alike from using the information contained in the texts as
referents for the archaeological record. The texts describe the, "cool land of the
Tamils" (Hart and Heifetz 1999: 28) and many lines are taken up with glorifying
the military exploits and bravery of the leaders of the three principal Tamil poli-
ties, the Chera, Chola, and Pandya. The term occurs often enough in the Sangam
literature, such that to speak of "Tamilakam," a "Tamil" land, and a "Tamil"
people, implies some sense of cultural or ethnic identity-or at least some con-
sciousness of independence or separateness from neighbors to the north and
south.
Also important in the identification of Tamilakam is inscriptional evidence.
Within Tamil Nadu, for example, about 80 to 90 rock inscriptions have been
discovered in natural caverns. Along with fragmentary epigraphs on potsherds
from around 25 sites in southern India (Zvelebil 1992: 123) and outside South
Asia (Mahadevan 1993), they form the strongest linguistic evidence for sepa-
rating Tamilakam from the rest of South India. The inscriptions range from third
212 ASIAN PERSPECTIVES . 42(2) . FALL 2003
or second century B.C. to second or third century A.D., and are written in the
Brahm! script, which was common throughout the peninsula at the time, but
the language is an early form of Tamil. The Tamil inscriptional data varies from
the Sangam texts, since they have a different source and were recorded for differ-
ent purposes. These inscriptions were written in what are believed to be Buddhist
or Jain ascetic caves, and their purpose seems to be to remind travelers of the
bounty of various merchants and kings and their support for these sects (Kennedy
1976: 6). The inscriptions mostly contain personal and occupational names of
donors who endowed the Buddhist and Jain monks with stone beds in caverns
(Gurukkal 1989: 160), and one of their greatest benefits is that they confirm
certain king and place names that are mentioned in the earliest Sangam texts
(Zvelebil 1992: 124).
Non-Tamil South Asian inscriptions also include references to Tamilakam.
Among the most pivotal are the rock edicts of the North Indian Mauryan
emperor Asoka, dated to third century B.C. One of the edicts refers to five inde-
pendent states that presumably existed beyond the southern border of his empire:
the Choda (Chola), Pandya, Satiyaputra, Keralaputras (Chera), and Tamraparni
(Sri Lanka) (Zvelebil 1992: 110)-an indication that the polities of Kerala, Tamil
Nadu, and Sri Lanka were not incorporated into the Mauryan realm (Fig. 3). Also
relevant is the central Indian Hathigumpha inscription (possibly from the second
half of the second century B.C.), which discusses the destruction of a "confederacy
of Tamil powers" (Zvelebil 1992: 103).
Outside of South Asia, it is the Greco-Roman writings that provide the most
detailed historical information about Tamilakam. India is in fact frequently men-
tioned in the Western classical literature (McCrindle 1971 : xxi). The protohistoric
period in South India coincides with a phase of South Asian participation in the
flourishing maritime networks of the Indian Ocean. The South Asian subconti-
nent is well known for having had long-standing, varied, and complex forms of
interaction with the external world, and South India is no exception. Historical
and archaeological reconstructions of South India during its Early Historic period
have placed a great deal of emphasis on the long-distance maritime trade net-
works to which South India belonged-particularly its links with the Roman
Empire (Begley 1996; Ray 1989, 1994, 1995). The network incorporated a num-
ber of regions along the Indian Ocean littoral-including the Red Sea coast (Salo-
man 1991; Sidebotham 1986), the Arabian coast (Whitehouse and Williamson
1973), East Africa (Munro-Hay 1996), Southeast Asia (Ray 1994; Smith 1999),
Sri Lanka (Munro-Hay 1996), and China (Ray 1994)-and South India acted as
a major node in the interregional transmittal of goods during Hellenistic and
Roman times (Charlesworth 1926). Of all the historical sources available for
Tamilakam, the Greco-Roman references to South India are particularly useful
since they are to a large degree datable and help, therefore, to fix the centuries
during which overseas trade flourished. However, most of these texts refer not to
"Tamilakam," but to specific trade centers and ports in peninsular India.
THE ARCHAEOLOGY OF PROTOHISTORIC KERALA AND TAMIL NADU
Although extensive portions of Kerala and Tamil Nadu have been explored and a
number of sites excavated (general overviews of the region's archaeology can be
found in Brubaker 2001; Gurukkal and Varrier 1999; Gurumurthy 1991; Leshnik
ABRAHAM . TAMIL KINGDOMS OF HISTORIC SOUTH INDIA 213
SAKA KSHATRAPA
o
KALINGA
N
100 200 km
Fig. 3. Polities in early South India (c. 200 B.C.-A.D. 300).
1974; Moorti 1994; and Ramachandran 1980), the archaeological record for
Tamilakam is far from satisfactory. Relatively few radiocarbon dates are available,
and there are, unfortunately, no clear sequences or patterns in artifact assemblages
that have permitted the development of internal relative dating sequences. Docu-
mentation of sites has been inconsistent at best-most published data on Iron
Age burials take the form of brief notices, for example, and when excavations
reports are available, they tend to rely on existing ill-defined artifact classification
schen"les. Another issue is the fragmentary nature of archaeological research in
South India. Most synthetic studies are delimited by state, making it difficult to
understand ancient regional patterns that crossed modern political boundaries.
Also, scholars have tended to compartmentalize areas of research, focusing on
certain site types, such as Iron Age burials or Early Historic settlements, or on
21
4
ASIAN PERSPECTIVES . 42(2) . FALL 2003
specific material assemblages, such as cave inscriptions or coins. Although some
South Indian historians have understood the need to treat all these elements as
interrelated parts of a single past social formation (see, for example, Gurukkal
1995: 239-240), it is only recently that archaeologists are coming to the same
realization.
A general overview of the archaeological record of Tamilakam will be pre-
sented here in order to evaluate the extent to which issues of identity can be
addressed. Archaeologically, the Tamil Sangam era corresponds roughly to the
late Iron Age-Early Historic period (c. 300 B.C. to A.D. 300), which represents
a key stage in the development of South Indian material culture. In addition to
Iron Age burial/commemorative monuments (Fig. 4), ceramic types, and metal
Fig. 4. Excavated Iron Age burials in Tamilakam.
ABRAHAM . TAMIL KINGDOMS OF HISTORIC SOUTH INDIA 2 15
and stone artifacts, the onset of the Early Historic phase witnessed the appearance
of coastal and inland settlements, new ceramic categories, numismatic and inscrip-
tional finds, and evidence for a burgeoning overseas trade.
Excavated sites in Kerala and Tamil Nadu can be broadly categorized as burials
or settlement sites (Fig. 5), or a combination of the two. The South Indian Iron
Age is associated most closely with the hundreds of burial/commemorative struc-
tures that span the peninsula. Popularly and collectively referred to as "megaliths,"
the distribution of these features is vast and covers numerous culture regions in
South Asia, including Sri Lanka. Of the locally made ceramics traditionally asso-
ciated with the South Indian Iron Age, a type referred to as Black-and-Red Ware
is the earliest and most well known, and one of the most widely distributed
chronologically and geographically. It is found in Iron Age settlements, Iron Age
Fig. 5. Excavated settlements in Tamilabm.
216 ASIAN PERSPECTIVES . 42(2) . FALL 2003
burials, and Early Historic sites in South India. Megaliths and craft objects allied
with them-Black-and-Red Ware, beads, bangles, terracotta forms, metal imple-
ments, etc.-appear throughout the peninsula, extending northward as well as
south into Sri Lanka, and they have been treated, by archaeologists and historians
alike, largely as a singular phenomenon. The general uniformity in architectural
elements and material culture has led most scholars to believe that the megaliths
belonged to the same cultural complex-a long-standing yet untenable conclu-
sion, given the relatively few monuments that have been excavated systemati-
cally and the paucity of radiocarbon dates (Moorti 1994: 4). Moreover, very few
studies specifically examine variability within and among megalith complexes
(e.g., Leshnik 1974; Mclntosh 1985; Moorti 1994), and they disagree on the ex-
tent of identifiable social differentiation.
The Early Historic phase in Tamilakam coincides with a number of new
material developments in Peninsular India. One of the most important is the
beginning of writing in the form of inscriptions on pottery and on cavern walls.
Another important change is the introduction of a range of artifacts of foreign-
mostly Mediterranean-manufacture. Roman coins, amphorae, and ceramics are
commonly associated with the Early Historic occupation at Tamil sites and
accepted as tangible evidence of the overseas trade in which South India engaged
at the time. The Early Historic period is also associated with the beginning of
urban living in Tan"lilakanl (the sites assigned to this period are exclusively settle-
ments), as evidenced by the number of sites, both coastal and inland, containing
architectural features that include brick structures, ring wells, pits with drains, and
possible industrial items like soakage jars, dyeing vats, and terracotta ovens (e.g.,
Begley 1996; Mahalingam 1970; Nagaswamy 1995; Rajan 1994; Raman 1988;
Soundara Rajan 1994). The same craft industries associated with the late Iron Age
phase continue into the Early Historic phase and were perhaps supplemented by
other activities such as gold working and weaving. And, along with foreign coins,
the occurrence of local coins indicates at least the beginnings of a monetary sys-
tem of exchange.
The locally produced artifacts associated with early Tamil sites include a few
pottery types, the most important of which, in addition to the aforementioned
Black-and-Red Ware, include Rouletted Ware and Russet-Coated Painted Ware.
Again, issues such as variation and dating of these ceramic types are critical to the
understanding of Tamil social organization, but aside from Gogte's (1997) argu-
ment for a Bengal-based origin for Rouletted Ware, few studies question the
apparent geographical and temporal uniformity of these ceramics in South India.
Local coins are occasionally found in or near Early Historic sites. Iron is pervasive,
in the form of agricultural tools and hunting and war implements; other typical
artifacts include copper and bronze objects, beads, terracotta figurines, and shell
and glass bangles.
The link between social boundaries and material culture is undeniably com-
plex, and the very concept of social boundaries and its anthropological and
archaeological application is still being debated within the scholarly community
(e.g., Hegmon 1998). Whatever the specific medium of material culture and the
particular past that is being studied, a major goal in the study of formal variation
across space is, "to identify social groups, whose boundaries are marked by dis-
tinctive patterns in the archaeological record" (Stark 1998: 1). Although material
ABRAHAM . TAMIL KINGDOMS OF HISTORIC SOUTH INDIA 217
culture often plays an active role in social interaction, especially in the expression
of social status and identity, it is not easy to articulate a straightforward relation-
ship between artifact and identity (Janusek 2002: 37). It has been argued that,
rather than relying on a single category of material culture, the researcher must
employ a careful, multivariate approach to interpret the identity and the links
between material data and past social processes (e.g., Dieder and Herbich 1998;
Janusek 2002). Emberling suggests a series of steps in the process of identifying
markers of identity (Emberling 1997: 311). First, one must identify a potentially
distinctive group. Then, the social and geographical boundaries of the alleged
group must be established by comparing patterns of artifact classes with those of
neighboring groups. Through the careful study of production and use, it might
then be possible to characterize the kind of group associated with those artifact
classes. Finally, these patterns should be compared with other categories of evi-
dence to substantiate claims of ethnic difference.
Even this cursory overview of the sites and material culture from Kerala and
Tamil Nadu highlights a key point: despite the common tendency for South
Indian historians and archaeologists to speak of "Tamil" material culture, the
archaeological evidence that sets Tamilakam as a region apart from the rest of
South India has never been clearly identified. If one evaluates Tamil cultural
identity using Elnberling's guidelines, then the claims of ethnic difference appear
to falter, since nearly all the material culture found in Tamilakam-that is, in
Kerala and Tamil Nadu-can be found elsewhere in peninsular India. Similarly
configured urban centers and habitation sites are located throughout South India
and Sri Lanka. And, although the majority of Iron Age burials are situated in
South India, they are widely distributed throughout the Indian subcontinent, and
only one or two types are unique to Tamilakam. In the same way, the distri-
bution of ceramics, iron, and other artifacts are dispersed across the alleged past
cultural-ethnic, linguistic, and geographic boundaries of South India.
Indeed, what seems to distinguish Tamilakam material cultural formations dur-
ing the late Iron Age-Early Historic period is the lack of some kind of evidence
one finds in neighboring regions; the Deccan region immediately to the north,
for example, is notable for its large number of both simple and elaborate Buddhist
sites, as well as for the wide array of numismatic finds-examples of locally
minted coins that have helped to reconstruct the political dynasties of the early
Deccan. Archaeological, inscriptional, and numismatic data indicate that the Dec-
can followed a different historical trajectory-it was part of the Mauryan realm
until its decline in the third century B.C., after which a cluster of later rulers
claimed the territory, the most important being the Satavahanas. So, too, is there
a separate story for Sri Lanka, whose Early Historic period is said to have begun
with northern Indian merchants settling on the island, followed by the introduc-
tion of Buddhism by an envoy of the Mauryan king Asoka in the third century
B.C. (Bopearachchi 1997: xvi).
"MATERIALIZING" THE TAMIL IDENTITY
Given the nature and distribution of sites and artifacts across the Tamil landscape,
it has been difficult to discern a pattern of style or symbol or process that stands
out, making simplistic normative notions of a Tamil identity appear misplaced.
218 ASIAN PERSPECTIVES . 42(2) . FALL 2003
In fact, except for the linguistic evidence, it would be challenging to identify a
"Tamil" artifact or site on the basis of the physical or stylistic traits alone. Perhaps
the only identifiably "Tamil" artifacts are the cave and pottery inscriptions, but
they have not been studied for the express purpose of establishing evidence about
the identity of Tamilakam. It seems plausible to suggest that our scholarly con-
sciousness of a Tamil land and the subsequent attribution of "Tamilness" to
material culture are based on the poorly articulated relationship between texts and
artifacts in early South India. The tendency of historians and archaeologists to
assign a text-based "Tamil identity" to objects and sites found in the Tamil region
may have obscured the possibility that more than one dialogue was in progress. In
the case of Tan'lilakam, a central concern must be how people used their material
culture to define themselves and to express their affiliations. In his study of later
prehistoric Europe, Wells (1998) mentions three basic ideas that have opened up
the archaeological study of identity: first is the view that .material culture serves as
an active agent for constructing meanings and communicating information with
others; second, identity is not a static attribute, but instead a dynamic, flexible,
and contingent property; third is the application of individual agency to the study
of archaeological contexts (Wells 1998: 240-241).
Furthermore, textual sources must not be taken at face value, but rather, "con-
sidered in terms of the social and political contexts in which they were produced,
and positions and interests of the authors and the audiences, and the active role
which texts may have played in the construction and negotiation of cultural
identity" (Jones 1999: 223-224). It is perhaps not a coincidence, in this formative
stage of Tamil social history, that both the Sangam texts and external references
to Tamilakam concentrate on military and political endeavor. As mentioned, the
Sangam poems are essentially heroic court poetry in praise of the military prowess
of various Tamil rulers as they competed for territory and the spoils of plunder.
The external inscriptions are by polities that are supposed to have attempted, or
profited by, hostile incursions into the region. In addition to the rock edicts of
the Mauryan ruler Asoka, there is a reference in a Sangam poem to a Mauryan
invasion of the south. The Hathigumpha inscription describes a league of Tamil
states strong enough to constitute a threat to the safety of the northern Kalinga
polity and claims that a Tamil league or confederacy had been in existence 113
years before (Kant 2000: 39). If group distinctiveness functioned as a strategy to
develop, maintain, and justify elite activities, then the early Tamil texts may signal
the efforts of petty rulers to promote a nascent self-awareness based on linguistic
and historical commonalities. This would be consistent in a context where the
development and maintenance of a supposed regionwide identity was necessary in
order to muster support and resources in the face of internal and external threats.
With this approach, such expressions of identity lead, not to an archaeological
reiflcation of such groups, but instead to an appreciation that identity may vary
depending on social domain and with relation to different scales of social interac-
tion (Jones 1997: 129).
The material culture of early Tamil South India, on the other hand, does not
obviously expose to us markers of a regional Tamil ethnicity or identity, and one
partial explanation may be that it reflects not the self-consciousness regionalism of
the documents, but a pattern of social and cultural processes that were primarily
small-scale, self-organized, or at least not controlled to a great degree by the soi-
ABRAHAM . TAMIL KINGDOMS OF HISTORIC SOUTH INDIA 2I9
disant rulers of Tamilakam. In this case, the archaeological data represent comple-
mental)' (and perhaps competing) perspectives of identity compared to those of
the South Indian texts. The activities and concerns represented by Tamil artifacts
are less apparently political than they are economic and social-aspects of lifestyle
in which an ethnic symbolism was not recorded in the conspicuous manner found
in the texts. From this perspective, it may not be a surprise to find archaeologists
agreeing that the activities best illustrated in the Tamil material record have to do
with trade, hunting, agriculture, and craft industries, and it may be in such activ-
ities that identity is situated. Tamil identity may be sought in those assemblages
that direct attention to regional distinctions in trade strategies, for example, such
as the patterns in Roman artifact distribution in Tamilakam that contrast strik-
ingly with the neighboring Deccan region. Although both regions participated in
long-distance Indian Ocean trade during the Early Historical period, the Deccan
contains a large number of Roman artifacts, while the bulk of Roman material in
Tamilakam is in the form of Roman silver and gold coin hoards. Romila Thapar
(1992) has suggested that this variability indicates unique patterns of trade organi-
zation; if so, material signatures of identity may reside not in static interpretations
of the archaeological data, but in understanding the distinct socio-economic pro-
cesses in which each region was engaged.
If the extent to which the larger populace embraced a common Tamil identity
is debatable, the answer may lie in a re-evaluation of the scale of analysis (Fein-
man 1997: 375). If daily Tamil life as represented by material culture was most
likely local rather than regional, chances are that the material assemblages here
reveal systems of identity that did not extend beyond the village, settlement, or
community level. As rich as the Sangam textual corpus is, for example, the bulk
of the information contained within it focuses narrowly on the activities of the
elite rulers and warriors, much like the texts originating in many early states and
civilizations (Adams 2001: 347). Such a focus overshadows the data on the daily
lifestyles and activities of the majority of the populace. In the search for alterna-
tive models, however, a re-examination of the Sangam texts reveals clues regard-
ing a system of local identities that might have greater relevance for the South
Indian archaeologist. In contrast to the better-known themes of kingship, warfare,
and plunder, the aintinai theme of the Sangam anthologies concerns the division
of Tamilakam into a series of microecological zones-a fivefold physiographic
division of the landscape-that comprised coastal regions, hilly backwoods, pasto-
ral tracts, riverine wetlands, and arid-parched zones (Champakalakshmi 1996: 96).
Associated with these microzones were different forms of production and social
formation, ranging from hunting-gathering to animal husbandl)', shifting agricul-
ture, wetland agriculture, fishing, and salt manufacture. To date, almost no con-
sideration has been given by South Indian archaeologists to this representation of
ancient Tamil life, yet it is an opportunity to appreciate the multiple and multi-
scalar dialogues present in the Sangam texts and address their implications for
interpreting identity in the South Indian material record. Perhaps ethnic identity
was not a salient social feature, as has been assumed for protohistoric Kerala and
Tamil Nadu. Instead of accepting received notions of a region-based "Tamilness,"
it may be more useful to the archaeologist to consider smaller, resource-based
distinctions as the more appropriate scale of analysis when examining the archae-
ological record of Kerala and Tamil Nadu.
220 ASIAN PERSPECTIVES . 42(2) . FALL 2003
CONCLUSION
By uncritically subsuming the study of artifacts under the study of texts, Tamil
scholars are in danger of missing important social strategies and practices con-
tained in both the textual and material data, and impoverishing our understanding
about the con"lplexities of the Tamil past. The discussion presented here suggests
one method by which the documentary and material records for Kerala and Tamil
Nadu may be integrated, while at the same time guarding against some of the
broad tendencies described by Feinman (1997): the preference for documents at
the expense of archaeological data; the unsystematic and selective use of textual
records by archaeologists; and the lack of attention paid to spatial and chronolog-
ical scales of analysis. Consciously and critically addressing these issues may allow
South Indian scholars to determine what, if anything, distinguishes, "the cool land
of the Tamils" from the rest of peninsular South India.
REFERENCES CITED
ADAMS, R. McC.
2001 Complexity in archaic states. journal of Anthropological Archaeology 20: 345-360.
BEGLEY, V.
1992 Ceramic evidence for pre-Periphls trade on the Indian coasts, In ROlne and India: The
Ancient Sea Trade: 157-196, ed. V. Begley and R. D. de Puma. Madison: University of
Wisconsin Press.
1996 Allcient Port of Arikamedu: New Excavations and Researches 1989-1992, vol. 1. Pondichery:
Ecole Franc,:aise d'Extreme Orient.
BOPEARACHCHI, O.
1997 Foreword, in Tabropane: Ancient Sri Lanka as Known to Greeks and ROlnans: ix-xxii, ed.
D.P.M. Weerakoddy. Turnhout: Brepols.
BRUBAKER, R.
2001 Aspects of mortuary variability in the South Indian Iron Age. Bnlletill of the DeceO/I College
Postgradnate and Research hlstitnte: 60-61 (2000-2001): 253-302.
CHAMPAKALAKSHMI, R.
1996 Trade, Ideology and Urbanizatioll: Sonth India 300 S.c. to A.D. 1300. Delhi: Oxford Uni-
versity Press.
CHARLESWORTH, M. P.
1926 Trade-Rontes and Commerce the RomOlI Empire. Chicago: Ares Publishers, Inc.
DIETLER, M., AND I. HERBlCH
1998 Habitus, techniques, and style: An integrated approach to the social understanding of
material culture and boundaries, in The Archaeological of Social Bourldaries: 232-263, ed. M.
Stark. Washington, D.C., and London: Smithsonian Institution Press.
EMBERLlNG, G.
1997 Ethnicity in complex societies: Archaeological perspectives. jonrt1al of
Research 5(4): 295-344.
FEINMAN, G. M.
1997 Thoughts on new approaches to combining the archaeological and historical records.
journal of Archaeological Method OIld Theor)' 4(3-4): 367-377.
GOGTE, V. D.
1997 The Chandraketugahr-Tamluk region of Bengal: Source of the Early Historic Rouletted
Ware from India and Southeast Asia. MOIl and Environment 22(1): 69-85.
GURUKKAL, R.
1989 Forms of production and forces of change in ancient Tamil society. Stndies in His!or)'
5(2): 159-175.
1995 Beginnings of the Historic Period: The Tamil south, in Ramt Perspcrri/Jes of Earl)' Sonth
Indian History: 237-265, ed. R. Thapar. Bombay: Popular Prakashan.
ABRAHAM . TAMIL KINGDOMS OF HISTORIC SOUTH INDIA 221
GURUKKAL, R., AND M.R.R. VARRIER
1999 CI/ltl/ral History Kerala, vol. 1. Thiruvananthapuram, Kerala: Department of Cultural
Publications, Government of Kerala.
GURUMURTHY, S.
1991 Cerall/ic Traditiolls ill SOllth II/dia, dOl III/ to 300 A.D. Madras: University of Madras.
HART, G. L.
1975 The Poell/s of AI/ciel/! Tall/il: Their Miliell al/d Their Sallshit COl/llterparts. Berkeley: Univer-
sity of California Press.
HART, G. L., AND H. HEIFETZ, TRANS. AND EDS.
1999 The FOl/r HI/I/dred SOl/gs of l;fIar al/d Wisdoll/: AI/ Alltholog)' of Poell/s frOIl/ Classical TO/I/il-
The PI/rtll/tllllim. New York: Columbia University Press.
HEGMON, M.
1998 Technology, style, and social practices: Archaeological practices, in The Archaeology of
Social BOl/lldaries: 264-270, ed. M. Stark. Washington and London: Smithsonian Institu-
tion Press.
jANUSEK, J. W.
2002 Out of many, one: Style and social boundaries in Tiwanaku. Latin Americall Alltiql/ity
13(1): 35-61.
JONES, S.
1997
1999
The Archaeology of Etllllicity. London and New York: Routledge.
Historical categories and the praxis of identity: The interpretation of ethnicity in histori-
cal archaeology, in Historical Archaeology: Back froll/ the Edge: 219-232, ed. P.P.A. Funari,
M. Hall, and S. jones. London and New York: Routledge.
KANT, S.
2000
Kosso, P.
1995
Hathigl/lUpha II/scription of KharalJl'la alld the Bhabrll Edict of Asoka: A Critical Stl/dy. Delhi:
D. K. Printworld.
KENNEDY, R. S.
1976 The king in early South India, as chieftain and emperor. lI/di01l Historical Review
3(1):1-15.
Epistemic independence between textual and material evidence, in Methods ill the Medi-
terralleall: Historical alld Arc/lOeolojiical Views 01/ Texts and Archaeolojiy: 177-196, ed. D. 13.
Small. Leiden: E. J. Brill.
KULKE, H., AND D. ROTHERMUND
1986 A History of India. Calcutta: Rupa and Co.
LEONE, M. P., AND C. CROSBY
1987 Middle-range theory in historical archaeology, in COllslIlI/er Choice ill Historical Archaeol-
ogy: 397-410, ed. S. Spencer-Wood. New York: Plenum Press.
LEONE, M. P., AND P. B. PorrER
1988 Introduction: Issues in historical archaeology, in The Recover)' of ill Historical
Archaeolojiy itl the Eastem United States: 1-22, ed. M. P. Leone and P. 13. Potter. Wash-
ington, D.C.: Smithsonian Institution Press.
LESHNIK, L. S.
1974 SOItth II/dial/ 'Megalithic' BI/rials: The Palldl/kal COII/plex. Wiesbaden: F. Steiner.
MAHADEVAN, I.
1993 Ancient Tamil contacts abroad: Recent epigraphical evidence. JOllmal of the I//Stitl/te of
SOl/th Asiall Stlldies 12(1): 136-155.
MAHALINGAM, T. V.
1970 Report on the Excallatiolls ill the LOllier Kaveri Valley (Til'llkkall/pl/li)'1IJ" O/Id Alajiarai). Madras:
University of Madras.
MCCRINDLE,J. W.
1971 [1901J AllriCllt Illdia as Described iI/ Classical Literatl/re. New Dehli: Munshiram Manoharlal.
McINTOSH, J.
1985 Dating the South Indian megaliths, in SOllth Asiall Archaeology 1983: 467-493, ed. J.
Schotsmans and M. Taddei. Naples: Istituto Universitario Orientale Dipartimento Di
Studi Asiatici.
222 ASIAN PERSPECTIVES . 42(2) . FALL 2003
MOORTl, U. S.
1994 Megalithic Cllltllre of SOllth Illdia: Socio-EcollolI/ic Perspectives. Varanasi: Ganga Kaveri Pub-
lishing House.
MORRISON, K. D., AND M. LYCETT
1994 Centralized power, centralized authority' Ideological claims and archaeological patterns.
Asiall Perspectives 32: 312-353.
1997 Inscriptions as artifacts: Precolonial South India and the analysis of texts. jOllmal of Archae-
ological Method alld Theory 4(3-4): 215-237.
MUNRo-HAY, S.C.H.
1996 Aksumite interests overseas, in Illdiall Oceall ill Alltiqllity: 403-416, ed. J. Reade. London
and New York: Kegan Paul International and The British Museum.
NAGASWAMY, R.
1995 Romall Kamr. Madras: Brahad Prakashan.
RAJAN, K.
1994 Archaeology of Tall/il Nadll (Kongu COlllltry). Delhi: Book Indian Publishing Co.
RAMACHANDRAN,K. S.
1980 Archaeology of South India: Tamil Nadu. Delhi: Sundeep.
RAMAN, K. V.
1988 Excavations at Uraiyllr (Tiruchirapal/i), 1965-69. Madras: University of Madras.
RAY, H. P.
1989 Early historical settlement in the Deccan: An ecological perspective. Man and Environment
14(1): 103-108.
1994 The Winds of Change: Bllddhism and the Mariti/lle Links of Early South India. Delhi: Oxford
University Press.
1995 Trade and contacts, in RecCllt Perspectives 011 Early II/diall History: 142-167, ed. R. Thapar.
Bombay: Popular Prakashan.
SALOMAN, R.
1991 Epigraphic remains of Indian traders in Egypt. jOl/rnal of the American Orielltal Society
111(4) :731-736.
SASTRI, N.
1966 A History cif SOllth [lldia. Madras: Oxford University Press.
SHANKS, M., AND C. TILLEY
1987 Recollstmcting ArcllOeology: Theory alld Practice. Cambridge: Cambridge University Press.
SHENNAN, S.
1989 Introduction: Archaeological approaches to cultural identity, in Archaeological Approaches to
Cllltllral Identity: 1-32, ed. S. J. Shennan. London and New York: Routledge.
SIDEBOTHAM, S. E.
1986 ROI/lall Ecollomic Policy ill the Erythra Thalassa, 30 B. C. -A.D. 217. Leiden: E. J. Brill.
SOUNDARA RAJAN, K. V.
1994 KaveripattillOm Excavations 1963-73 (A Port City 011 the Tall/illladll Coast). New Delhi:
Archaeological Survey of India.
SMITH, M. L.
1999 'Indianization' from the Indian point of view: Trade and cultural contacts with Southeast
Asia in the early first millennium CEo jOllmal of the Ecollolllic alld Social History of the Oriellt
42(1): 1-26.
STAIlK, M.
1998 Technical choices and social boundaries in material culture patterning: An introduc-
tion, in The Archaeology of Social BOlllldaries: 1-11, ed. M. Stark. Washington, D.C., and
London: Smithsonian Institution Press.
THAPAR, R.
1992 Black gold: South Asia and the Roman maritime trade. SOllth Asia 15(2): 1-27.
THUIlSTON, T.
1997 Historians, prehistorians, and the tyranny of the historical record: Danish state formation
through documents and archaeological data. jOllmal of Archaeological Method alld Theory
4(3-4): 239-263.
ABRAHAM . TAMIL KINGDOMS OF HISTORIC SOUTH INDIA 223
WELLS, P. S.
1998 Identity and material culture in the later prehistory of Central Europe. JOllrJIal or Archaeo·
logical Research 6(3): 239-298.
WHITEHOUSE, D., AND A. WILLIAMSON
1973 Sasanian maritime trade. IRAN 11.
WYLIE, A.
1985 The reaction against analogy. Advallces ill A rchaeological Method alld Theory 8: 63-111.
ZVELEBIL, K.
1992 COII/pal/iol/ Stl/dies to the History of Til/I/il Literatl/re. Leiden: E. J. Brill.
ABSTRACT
For the southern Indian states of Kerala and Tamil Nadu, the most important doc-
umentary source for information on early South lndian culture is a body of prose
poetry known as the Sangam anthology. These indigenous texts date to the first few
centuries A.D. and comprise the earliest extant examples of Tamil literature. Not
surprisingly, this is also the period to which can be traced the first indications of the
concept of a "Tamil" identity in South India. Archaeologically, the Tamil Sangam
era corresponds roughly to the late Iron Age-Early Historic period (c. 300 B.C. to
A.D. 300), which represents a key stage in the development of South Indian material
culture. Prevailing analyses of early Tamil society have relied heavily on the histori-
cal texts, often at the expense of critically examining the material culture from Ker-
ala and Tamil Nadu. This study examines the relationship between South Indian
archaeology and history and argues that any framework for interpreting early Tamil
identity must acknowledge the important qualitative differences in the ways that
texts and arti£1cts construct and reflect ethnic identity, and that archaeologists and
historians must analyze their respective data sets within the larger social, political,
and economic practices of early Tal11ilakal11. KEYWORDS: South Asia, South India,
Tal11ilakam, Kerala, Tamil Nadu, history and archaeology, cultural-ethnic identity.

As will be demonstrated here. By carefully separating and analyzing the written and material records for Tamilakam (Fig.208 ASIAN PERSPECTIVES . and research in other regions highlights some of the . this study will examine what. The underlying assumption continues to be that the documentary record serves as the best and most reliable source for knowledge about past identity. 2). AND IDENTITY The relationship between archaeology and history has become a mounting concern among archaeologists. HISTORY. in fact. 1. if anything. appears to challenge the very notion of a separate culture region. States of South India. the archaeological data from protohistoric Kerala and Tamil Nadu is not so clear-cut and. refute this notion of cultural separateness. 42(2) . FALL 2003 r N I o TAMILNADU 100 200 km Fig. it meant to have a Tamil identity in Early Historic South India. ARCHAEOLOGY.

TAMIL KINGDOMS OF HISTORIC SOUTH INDIA 209 INDIA t N o 600 km Fig. Location ofTamilakam (c. that the overwhelming tendency is to privilege the information contained in texts at the expense of data derived from the archaeological record (e.D. for instance.g. issues that are relevant to this study of Kerala and Tamil Nadu. Many archaeologists agree. while archaeologists limit their efforts to single-site evaluations (Thurston 1997: 260). archaeologists themselves can be prejudiced in their selective use of written records to support their findings (Feinman 1997: 372).C. Also. the rationale for conducting historically informed archaeology is indisputable. Conflicts also emerge when historians focus on the texts at the expense of other materials. 300).. Feinman 1997. Nevertheless.-A.ABRAHAM . combining the . Morrison and Lycett 1997). 300 B. 2.

but its meaning will alter in different contexts. a particular material form may remain the same. At first glance. the use of knowledge from one record to make inferences about the second record. Wylie (1985) espouses a similar approach in her discussion about the use of analogy in archaeology. Leone and Potter (1988) argue that Binford's middle-range theory may be adapted from ethnoarchaeological studies to the documentary record-in other words. 42(2) . Shennan (1989: 14) states. the documents would provide testable questions that the material remains may be able to answer. the historical references to specific past groups lead inevitably to the labeling of sites and objects (Jones 1997: 27). One kind of link often established between archaeological and textual records (particularly in South Indian archaeology) is the one rooted in analogical reasoning. Although the two sources are seldom about the same event. In the case of people who have left documentary records. Essentially. they are often about related things (Kosso 1995: 183). The interplay of material data and historical texts especially resonates during the quest for past social identities. or objects. the most basic way of linking the two is to use one record as a ITleanS of identifying or corroborating information recovered by the other. it may be possible to compare and contrast the claims of each line of evidence. i. Kosso likens the relationship between texts and artifacts to the use of testimony and physical evidence in a court trial (Kosso 1995 : 182): each line of evidence can be used potentially to evaluate the other. with a systematic enumeration of the distinctions.. Conversely." since this variation arises from the result of an enormous range of different processes. Kosso is careful to acknowledge the lack of complete independence between historical and archaeological forms of evidence but proposed that.e. that textual evidence should be used to set up "descriptive grids" (Leone and Potter 1988: 13). the archaeologist may try to search for material markers of ethnicity in the archaeological record. against which the archaeological data can be analyzed." One cannot simply search for artifact patterning in order to locate ethnicity. the same debates apply to archaeological and historical sources. Although Wylie is referring to archaeological and ethnographic sources in her essay. any more than one can locate archaeological "cultures. it will be used in different ways according to different historical traditions and social contexts (Shanks and Tilley 1987). But the archaeolog- . and hence may be used in conjunction with one another to recreate the past. based on the documentary record. and the answers in turn will offer insights into both the documentary and material records and their relationship (Leone and Crosby 1987: 14). FALL 2003 historical and archaeological data sets must lead to richer perspectives on the complexities of the past. based often on assumptions of uniformitarian principles. trying to locate cultural or ethnic identity may look suspiciously like an outdated attempt to superimpose old culture-historical taxonomies on early South India. Once such issues are understood through an analysis of texts. The definition of ethnicity itself is beset by difficulties. It is commonly assumed that studies of ethnicity require access to people's self-conscious reflections on identity (Jones 1999: 220). peoples.210 ASIAN PERSPECTIVES . and that written sources provide the ITlOst reliable accounts of such reflections. Recognizing that history and archaeology are both media by which the past can be accessed. "ethnicity must be distinguished from mere spatial variation and should refer to self-conscious identification with a particular social group at least partly based on a specific locality or origin.

TAMIL KINGDOMS OF HISTORIC SOUTH INDIA 211 ical pursuit of groups named in historical texts gives rise to a new set of issues. such that to speak of "Tamilakam. and propose dates ranging from the first century B. "self-evident meaning and cannot. and the texts may also obscure ongoing processes of identity formation and re-formation. There are altogether 2381 poems by 473 poets and 102 poems by anonymous authors. Names recorded in texts are often deliberate records of information about rulers and bureaucrats. The inscriptions range from third . at least in the case of South India. The works that have survived include eight anthologies and ten idylls. Chola. about 80 to 90 rock inscriptions have been discovered in natural caverns. Scholars are divided about the chronology of these works. The boundaries defined in the texts may not match boundaries found in the material record. Appreciating the Sangam corpus as a source of historical data requires understanding their poetic and bardic nature: a bulk of the poetry is concerned with extolling the exploits of rulers. Despite the theoretical concerns and methodological constraints. It is not possible at this stage to confidently propose a solution. Within Tamil Nadu.C. for example. The most important of these is a body of prose poetry referred to as the "Sangam Anthology"-indigenous texts dating to the first few centuries A. and Pandya. Zvelebil 1992). The texts do not provide. they form the strongest linguistic evidence for separating Tamilakam from the rest of South India. rather than the self-identifications of the general population (Emberling 1997:313)." a "Tamil" land.D. a work on grammar. in themselves. besides the grammar and 18 minor works (Hart 1975: 7).D. solve archaeology's central methodological challenges" (Morrison and Lycett 1997: 233). "TAMILAKAM" AS A HISTORICAL ENTITY The evidence for the identification of Tamilakam as a distinct culture region is culled from a variety of text-based sources. (Hart 1975. The term occurs often enough in the Sangam literature. that comprise the earliest extant examples of Tamil literature. up to the sixth century A. the influence of text-based knowledge on the evaluation of archaeological data is undeniable and even inevitable. "cool land of the Tamils" (Hart and Heifetz 1999: 28) and many lines are taken up with glorifying the military exploits and bravery of the leaders of the three principal Tamil polities. warriors. one may begin to understand the informational gap between the two data sets and suggest possible ways to narrow that gap. the Chera.ABRAHAM . Also important in the identification of Tamilakam is inscriptional evidence. implies some sense of cultural or ethnic identity-or at least some consciousness of independence or separateness from neighbors to the north and south. Ray 1995. and 18 minor works. but perhaps by considering the early Tamil written and material records separately. The situation is compounded by the fact that the material record of complex societies may have been manipulated in their turn and may in fact obscure actual relations of power (Morrison and Lycett 1994: 327). and patrons (Champakalakshmi 1996: 175)-a fact that has not prevented historians and archaeologists alike from using the information contained in the texts as referents for the archaeological record. and a "Tamil" people. The texts describe the. Along with fragmentary epigraphs on potsherds from around 25 sites in southern India (Zvelebil 1992: 123) and outside South Asia (Mahadevan 1993).

which was common throughout the peninsula at the time.C. and complex forms of interaction with the external world. 1995). but the language is an early form of Tamil. Of all the historical sources available for Tamilakam. since they have a different source and were recorded for different purposes. most of these texts refer not to "Tamilakam. Smith 1999). Sri Lanka (Munro-Hay 1996). dated to third century B. Outside of South Asia. The protohistoric period in South India coincides with a phase of South Asian participation in the flourishing maritime networks of the Indian Ocean. Non-Tamil South Asian inscriptions also include references to Tamilakam. 3). the Greco-Roman references to South India are particularly useful since they are to a large degree datable and help. India is in fact frequently mentioned in the Western classical literature (McCrindle 1971 : xxi). and their purpose seems to be to remind travelers of the bounty of various merchants and kings and their support for these sects (Kennedy 1976: 6).). Pandya. 42(2) . and South India is no exception. Also relevant is the central Indian Hathigumpha inscription (possibly from the second half of the second century B.C. and are written in the Brahm! script. The network incorporated a number of regions along the Indian Ocean littoral-including the Red Sea coast (Saloman 1991. Southeast Asia (Ray 1994. East Africa (Munro-Hay 1996). However. Keralaputras (Chera). Satiyaputra. Gurukkal and Varrier 1999. which discusses the destruction of a "confederacy of Tamil powers" (Zvelebil 1992: 103). Leshnik . Ray 1989. the Arabian coast (Whitehouse and Williamson 1973). Among the most pivotal are the rock edicts of the North Indian Mauryan emperor Asoka. The Tamil inscriptional data varies from the Sangam texts. The inscriptions mostly contain personal and occupational names of donors who endowed the Buddhist and Jain monks with stone beds in caverns (Gurukkal 1989: 160). Historical and archaeological reconstructions of South India during its Early Historic period have placed a great deal of emphasis on the long-distance maritime trade networks to which South India belonged-particularly its links with the Roman Empire (Begley 1996. 1994. therefore. These inscriptions were written in what are believed to be Buddhist or Jain ascetic caves. FALL 2003 or second century B. varied. Gurumurthy 1991.C. Tamil Nadu. One of the edicts refers to five independent states that presumably existed beyond the southern border of his empire: the Choda (Chola). it is the Greco-Roman writings that provide the most detailed historical information about Tamilakam. to fix the centuries during which overseas trade flourished. The South Asian subcontinent is well known for having had long-standing. Sidebotham 1986).D.. and China (Ray 1994)-and South India acted as a major node in the interregional transmittal of goods during Hellenistic and Roman times (Charlesworth 1926).212 ASIAN PERSPECTIVES . THE ARCHAEOLOGY OF PROTOHISTORIC KERALA AND TAMIL NADU Although extensive portions of Kerala and Tamil Nadu have been explored and a number of sites excavated (general overviews of the region's archaeology can be found in Brubaker 2001. and one of their greatest benefits is that they confirm certain king and place names that are mentioned in the earliest Sangam texts (Zvelebil 1992: 124). and Sri Lanka were not incorporated into the Mauryan realm (Fig. to second or third century A. and Tamraparni (Sri Lanka) (Zvelebil 1992: 110)-an indication that the polities of Kerala." but to specific trade centers and ports in peninsular India.

Relatively few radiocarbon dates are available. 300). Another issue is the fragmentary nature of archaeological research in South India. and Ramachandran 1980). and there are. no clear sequences or patterns in artifact assemblages that have permitted the development of internal relative dating sequences. scholars have tended to compartmentalize areas of research. Most synthetic studies are delimited by state. Also. unfortunately. 3. and when excavations reports are available. Moorti 1994. they tend to rely on existing ill-defined artifact classification schen"les. for example. or on . 200 B.-A.ABRAHAM . Polities in early South India (c. making it difficult to understand ancient regional patterns that crossed modern political boundaries. focusing on certain site types.C. TAMIL KINGDOMS OF HISTORIC SOUTH INDIA 213 KALINGA SAKA KSHATRAPA N o 100 200 km Fig.D. 1974. Documentation of sites has been inconsistent at best-most published data on Iron Age burials take the form of brief notices. such as Iron Age burials or Early Historic settlements. the archaeological record for Tamilakam is far from satisfactory.

21 4 ASIAN PERSPECTIVES . and metal Fig. Although some South Indian historians have understood the need to treat all these elements as interrelated parts of a single past social formation (see. 300). Archaeologically. the Tamil Sangam era corresponds roughly to the late Iron Age-Early Historic period (c. 4). Excavated Iron Age burials in Tamilakam. which represents a key stage in the development of South Indian material culture. In addition to Iron Age burial/commemorative monuments (Fig. A general overview of the archaeological record of Tamilakam will be presented here in order to evaluate the extent to which issues of identity can be addressed. ceramic types.C. . it is only recently that archaeologists are coming to the same realization. 300 B. FALL 2003 specific material assemblages. such as cave inscriptions or coins. 42(2) .D. 4. for example. to A. Gurukkal 1995: 239-240).

. Excavated sites in Kerala and Tamil Nadu can be broadly categorized as burials or settlement sites (Fig. new ceramic categories. numismatic and inscriptional finds. and evidence for a burgeoning overseas trade.ABRAHAM . including Sri Lanka." the distribution of these features is vast and covers numerous culture regions in South Asia. or a combination of the two. a type referred to as Black-and-Red Ware is the earliest and most well known. The South Indian Iron Age is associated most closely with the hundreds of burial/commemorative structures that span the peninsula. Popularly and collectively referred to as "megaliths. Iron Age Fig. 5). and one of the most widely distributed chronologically and geographically. TAMIL KINGDOMS OF HISTORIC SOUTH INDIA 2 15 and stone artifacts. It is found in Iron Age settlements. the onset of the Early Historic phase witnessed the appearance of coastal and inland settlements. Excavated settlements in Tamilabm. 5. Of the locally made ceramics traditionally associated with the South Indian Iron Age.

Leshnik 1974. as evidenced by the number of sites. very few studies specifically examine variability within and among megalith complexes (e. along with foreign coins. in the form of agricultural tools and hunting and war implements.. Hegmon 1998). The general uniformity in architectural elements and material culture has led most scholars to believe that the megaliths belonged to the same cultural complex-a long-standing yet untenable conclusion. other typical artifacts include copper and bronze objects. beads. terracotta forms. Roman coins. beads. the occurrence of local coins indicates at least the beginnings of a monetary system of exchange. Raman 1988. And.g. bangles. etc. in addition to the aforementioned Black-and-Red Ware. by archaeologists and historians alike. One of the most important is the beginning of writing in the form of inscriptions on pottery and on cavern walls. and the very concept of social boundaries and its anthropological and archaeological application is still being debated within the scholarly community (e. The link between social boundaries and material culture is undeniably complex. metal implements. issues such as variation and dating of these ceramic types are critical to the understanding of Tamil social organization. Mclntosh 1985. the most important of which. and Early Historic sites in South India. dyeing vats. Megaliths and craft objects allied with them-Black-and-Red Ware. and ceramics are commonly associated with the Early Historic occupation at Tamil sites and accepted as tangible evidence of the overseas trade in which South India engaged at the time. extending northward as well as south into Sri Lanka. 42(2) . FALL 2003 burials. and they have been treated. Moreover. The Early Historic phase in Tamilakam coincides with a number of new material developments in Peninsular India. Again. containing architectural features that include brick structures. but aside from Gogte's (1997) argument for a Bengal-based origin for Rouletted Ware. both coastal and inland.g. The Early Historic period is also associated with the beginning of urban living in Tan"lilakanl (the sites assigned to this period are exclusively settlements). Nagaswamy 1995. Whatever the specific medium of material culture and the particular past that is being studied. Begley 1996. and terra cotta ovens (e. Iron is pervasive. whose boundaries are marked by distinctive patterns in the archaeological record" (Stark 1998: 1). Soundara Rajan 1994). a major goal in the study of formal variation across space is. Although material . "to identify social groups. and shell and glass bangles..-appear throughout the peninsula.g. largely as a singular phenomenon. terracotta figurines. ring wells. include Rouletted Ware and Russet-Coated Painted Ware. and they disagree on the extent of identifiable social differentiation. Rajan 1994.216 ASIAN PERSPECTIVES . Another important change is the introduction of a range of artifacts of foreignmostly Mediterranean-manufacture. amphorae. Mahalingam 1970.. Moorti 1994). given the relatively few monuments that have been excavated systematically and the paucity of radiocarbon dates (Moorti 1994: 4). pits with drains. Local coins are occasionally found in or near Early Historic sites. The same craft industries associated with the late Iron Age phase continue into the Early Historic phase and were perhaps supplemented by other activities such as gold working and weaving. few studies question the apparent geographical and temporal uniformity of these ceramics in South India. and possible industrial items like soakage jars. The locally produced artifacts associated with early Tamil sites include a few pottery types.

Indeed. Similarly configured urban centers and habitation sites are located throughout South India and Sri Lanka. If one evaluates Tamil cultural identity using Elnberling's guidelines. Through the careful study of production and use. in Kerala and Tamil Nadu-can be found elsewhere in peninsular India. "MATERIALIZING" THE TAMIL IDENTITY Given the nature and distribution of sites and artifacts across the Tamil landscape. since nearly all the material culture found in Tamilakam-that is. followed by the introduction of Buddhism by an envoy of the Mauryan king Asoka in the third century B.g. Dieder and Herbich 1998. Then. the archaeological evidence that sets Tamilakam as a region apart from the rest of South India has never been clearly identified. as well as for the wide array of numismatic finds-examples of locally minted coins that have helped to reconstruct the political dynasties of the early Deccan. iron.. the social and geographical boundaries of the alleged group must be established by comparing patterns of artifact classes with those of neighboring groups.C. In the same way. Emberling suggests a series of steps in the process of identifying markers of identity (Emberling 1997: 311). Janusek 2002).ABRAHAM . after which a cluster of later rulers claimed the territory. . First. inscriptional. is there a separate story for Sri Lanka. (Bopearachchi 1997: xvi). then the claims of ethnic difference appear to falter. it has been difficult to discern a pattern of style or symbol or process that stands out. it is not easy to articulate a straightforward relationship between artifact and identity (Janusek 2002: 37). So. the researcher must employ a careful. although the majority of Iron Age burials are situated in South India. whose Early Historic period is said to have begun with northern Indian merchants settling on the island. the Deccan region immediately to the north. rather than relying on a single category of material culture. the most important being the Satavahanas. they are widely distributed throughout the Indian subcontinent. Even this cursory overview of the sites and material culture from Kerala and Tamil Nadu highlights a key point: despite the common tendency for South Indian historians and archaeologists to speak of "Tamil" material culture. TAMIL KINGDOMS OF HISTORIC SOUTH INDIA 217 culture often plays an active role in social interaction.C. one must identify a potentially distinctive group. and geographic boundaries of South India. is notable for its large number of both simple and elaborate Buddhist sites. what seems to distinguish Tamilakam material cultural formations during the late Iron Age-Early Historic period is the lack of some kind of evidence one finds in neighboring regions. and numismatic data indicate that the Deccan followed a different historical trajectory-it was part of the Mauryan realm until its decline in the third century B. and only one or two types are unique to Tamilakam. especially in the expression of social status and identity. Finally. making simplistic normative notions of a Tamil identity appear misplaced. for example. It has been argued that. these patterns should be compared with other categories of evidence to substantiate claims of ethnic difference. And. linguistic. multivariate approach to interpret the identity and the links between material data and past social processes (e. too. Archaeological. the distribution of ceramics.. it might then be possible to characterize the kind of group associated with those artifact classes. and other artifacts are dispersed across the alleged past cultural-ethnic.

In addition to the rock edicts of the Mauryan ruler Asoka. and justify elite activities.218 ASIAN PERSPECTIVES . maintain. that both the Sangam texts and external references to Tamilakam concentrate on military and political endeavor. The Hathigumpha inscription describes a league of Tamil states strong enough to constitute a threat to the safety of the northern Kalinga polity and claims that a Tamil league or confederacy had been in existence 113 years before (Kant 2000: 39). but they have not been studied for the express purpose of establishing evidence about the identity of Tamilakam. It seems plausible to suggest that our scholarly consciousness of a Tamil land and the subsequent attribution of "Tamilness" to material culture are based on the poorly articulated relationship between texts and artifacts in early South India. hostile incursions into the region. the Sangam poems are essentially heroic court poetry in praise of the military prowess of various Tamil rulers as they competed for territory and the spoils of plunder. With this approach. it would be challenging to identify a "Tamil" artifact or site on the basis of the physical or stylistic traits alone. If group distinctiveness functioned as a strategy to develop. 42(2) . self-organized. and the active role which texts may have played in the construction and negotiation of cultural identity" (Jones 1999: 223-224). or at least not controlled to a great degree by the so i- . a central concern must be how people used their material culture to define themselves and to express their affiliations. As mentioned. Furthermore. The material culture of early Tamil South India. but instead a dynamic. but rather. "considered in terms of the social and political contexts in which they were produced. and contingent property. does not obviously expose to us markers of a regional Tamil ethnicity or identity. and one partial explanation may be that it reflects not the self-consciousness regionalism of the documents. In his study of later prehistoric Europe. textual sources must not be taken at face value. but instead to an appreciation that identity may vary depending on social domain and with relation to different scales of social interaction (Jones 1997: 129). The external inscriptions are by polities that are supposed to have attempted. FALL 2003 In fact. then the early Tamil texts may signal the efforts of petty rulers to promote a nascent self-awareness based on linguistic and historical commonalities. It is perhaps not a coincidence. The tendency of historians and archaeologists to assign a text-based "Tamil identity" to objects and sites found in the Tamil region may have obscured the possibility that more than one dialogue was in progress. flexible. This would be consistent in a context where the development and maintenance of a supposed regionwide identity was necessary in order to muster support and resources in the face of internal and external threats. such expressions of identity lead. identity is not a static attribute. Wells (1998) mentions three basic ideas that have opened up the archaeological study of identity: first is the view that . in this formative stage of Tamil social history. or profited by. second.material culture serves as an active agent for constructing meanings and communicating information with others. but a pattern of social and cultural processes that were primarily small-scale. except for the linguistic evidence. In the case of Tan'lilakam. on the other hand. Perhaps the only identifiably "Tamil" artifacts are the cave and pottery inscriptions. not to an archaeological reiflcation of such groups. third is the application of individual agency to the study of archaeological contexts (Wells 1998: 240-241). there is a reference in a Sangam poem to a Mauryan invasion of the south. and positions and interests of the authors and the audiences.

and craft industries. such as the patterns in Roman artifact distribution in Tamilakam that contrast strikingly with the neighboring Deccan region. Tamil identity may be sought in those assemblages that direct attention to regional distinctions in trade strategies. ranging from hunting-gathering to animal husbandl)'. it may not be a surprise to find archaeologists agreeing that the activities best illustrated in the Tamil material record have to do with trade. or community level." it may be more useful to the archaeologist to consider smaller. pastoral tracts. Although both regions participated in long-distance Indian Ocean trade during the Early Historical period. resource-based distinctions as the more appropriate scale of analysis when examining the archaeological record of Kerala and Tamil Nadu. Perhaps ethnic identity was not a salient social feature. If daily Tamil life as represented by material culture was most likely local rather than regional. but in understanding the distinct socio-economic processes in which each region was engaged. riverine wetlands. To date. the archaeological data represent complemental)' (and perhaps competing) perspectives of identity compared to those of the South Indian texts. fishing. almost no consideration has been given by South Indian archaeologists to this representation of ancient Tamil life. hunting. while the bulk of Roman material in Tamilakam is in the form of Roman silver and gold coin hoards. the Deccan contains a large number of Roman artifacts. Such a focus overshadows the data on the daily lifestyles and activities of the majority of the populace. From this perspective. wetland agriculture. . a re-examination of the Sangam texts reveals clues regarding a system of local identities that might have greater relevance for the South Indian archaeologist. Instead of accepting received notions of a region-based "Tamilness. as has been assumed for protohistoric Kerala and Tamil Nadu. hilly backwoods. If the extent to which the larger populace embraced a common Tamil identity is debatable. shifting agriculture. for example. and it may be in such activities that identity is situated. In this case. the aintinai theme of the Sangam anthologies concerns the division of Tamilakam into a series of microecological zones-a fivefold physiographic division of the landscape-that comprised coastal regions. if so. As rich as the Sangam textual corpus is. material signatures of identity may reside not in static interpretations of the archaeological data. TAMIL KINGDOMS OF HISTORIC SOUTH INDIA 2I9 disant rulers of Tamilakam. agriculture. chances are that the material assemblages here reveal systems of identity that did not extend beyond the village. the answer may lie in a re-evaluation of the scale of analysis (Feinman 1997: 375). for example. and plunder. the bulk of the information contained within it focuses narrowly on the activities of the elite rulers and warriors. and salt manufacture. much like the texts originating in many early states and civilizations (Adams 2001: 347).ABRAHAM . Associated with these microzones were different forms of production and social formation. Romila Thapar (1992) has suggested that this variability indicates unique patterns of trade organization. and arid-parched zones (Champakalakshmi 1996: 96). warfare. settlement. yet it is an opportunity to appreciate the multiple and multiscalar dialogues present in the Sangam texts and address their implications for interpreting identity in the South Indian material record. In the search for alternative models. In contrast to the better-known themes of kingship. however. The activities and concerns represented by Tamil artifacts are less apparently political than they are economic and social-aspects of lifestyle in which an ethnic symbolism was not recorded in the conspicuous manner found in the texts.

Foreword. REFERENCES CITED ADAMS. R. . R. D. CHARLESWORTH. M. ed. distinguishes. V. Thapar. 1926 DIETLER. BOPEARACHCHI. V. In ROlne and India: The Ancient Sea Trade: 157-196. the unsystematic and selective use of textual records by archaeologists. "the cool land of the Tamils" from the rest of peninsular South India. ed.c. 1997 BRUBAKER. 1989 1995 Forms of production and forces of change in ancient Tamil society. and style: An integrated approach to the social understanding of material culture and boundaries. Bnlletill of the DeceO/I College Postgradnate and Research hlstitnte: 60-61 (2000-2001): 253-302. Allcient Port of Arikamedu: New Excavations and Researches 1989-1992. 1997 Ethnicity in complex societies: Archaeological perspectives. GURUKKAL.P. 2001 CHAMPAKALAKSHMI. in Tabropane: Ancient Sri Lanka as Known to Greeks and ROlnans: ix-xxii. vol. The discussion presented here suggests one method by which the documentary and material records for Kerala and Tamil Nadu may be integrated. EMBERLlNG. 1992 1996 Ceramic evidence for pre-Periphls trade on the Indian coasts. D.D. McC. R. de Puma. Stark. Trade. Aspects of mortuary variability in the South Indian Iron Age. D. Complexity in archaic states. 1997 The Chandraketugahr-Tamluk region of Bengal: Source of the Early Historic Rouletted Ware from India and Southeast Asia. in The Archaeological of Social Bourldaries: 232-263.:aise d'Extreme Orient. techniques. of Archaeolo. Pondichery: Ecole Franc. 1997 GOGTE.220 ASIAN PERSPECTIVES . Weerakoddy. if anything. ed. 42(2) . journal of Anthropological Archaeology 20: 345-360. and the lack of attention paid to spatial and chronological scales of analysis. D. M. O. while at the same time guarding against some of the broad tendencies described by Feinman (1997): the preference for documents at the expense of archaeological data. HERBlCH Habitus. and impoverishing our understanding about the con"lplexities of the Tamil past. Stndies in His!or)' 5(2): 159-175. Madison: University of Wisconsin Press. 1.. R. Tamil scholars are in danger of missing important social strategies and practices contained in both the textual and material data.. Consciously and critically addressing these issues may allow South Indian scholars to determine what. in Ramt Perspcrri/Jes of Earl)' Sonth Indian History: 237-265.M. MOIl and Environment 22(1): 69-85. Ideology and Urbanizatioll: Sonth India 300 S. Thoughts on new approaches to combining the archaeological and historical records. 2001 BEGLEY. Bombay: Popular Prakashan. P. ed. journal of Archaeological Method OIld Theor)' 4(3-4): 367-377.~i(al FEINMAN. Inc. V. M. FALL 2003 CONCLUSION By uncritically subsuming the study of artifacts under the study of texts. jonrt1al Research 5(4): 295-344. to A. Trade-Rontes and Commerce AND ~f the RomOlI Empire. Beginnings of the Historic Period: The Tamil south. 1300. Washington. 1998 I. and London: Smithsonian Institution Press. Begley and R. Delhi: Oxford University Press. M. Turnhout: Brepols.C. G. G. Chicago: Ares Publishers. 1996 R.

SOItth II/dial/ 'Megalithic' BI/rials: The Palldl/kal COII/plex. M. M. as chieftain and emperor. D.: Smithsonian Institution Press. AND M..~ ill Historical Archaeolojiy itl the Eastem United States: 1-22. Delhi: D. New York: Plenum Press. G. AND EDS. H. 1997 1999 Historical categories and the praxis of identity: The interpretation of ethnicity in historical archaeology. Washington and London: Smithsonian Institution Press. Cerall/ic Traditiolls ill SOllth II/dia. ed.A. ed. 2000 KENNEDY. AND 1988 P. Wiesbaden: F. Epistemic independence between textual and material evidence. Brill. ed. AND H. GURUMURTHY. in The Recover)' of Meanill. MCCRINDLE. M. TRANS. 1976 Kosso. 1999 CI/ltl/ral History t~f V ARRIER Kerala. dOl III/ to 300 A. 1993 Ancient Tamil contacts abroad: Recent epigraphical evidence. I.D. 1970 T. Berkeley: University of California Press. Latin Americall Alltiql/ity 13(1): 35-61. J. W. The Poell/s of AI/ciel/! Tall/il: Their Miliell al/d Their Sallshit COl/llterparts. S. Leone and P. Steiner. New Dehli: Munshiram Manoharlal.fIar al/d Wisdoll/: AI/ Alltholog)' of Poell/s frOIl/ Classical TO/I/ilThe PI/rtll/tllllim. The king in early South India. ROTHERMUND A History of India. J. K. V. MAHADEVAN. Dating the South Indian megaliths. S. and social practices: Archaeological practices. Washington. ed. . Government of Kerala. S. R. ed.. London and New York: Routledge. M. 1998 Technology. in SOllth Asiall Archaeology 1983: 467-493. in COllslIlI/er Choice ill Historical Archaeology: 397-410. of MAHALINGAM. G. 1975 HART. B. LESHNIK. S. Leiden: E. The Archaeology of Etllllicity. Small. in Methods ill the Medi1995 terralleall: Historical alld Arc/lOeolojiical Views 01/ Texts and Archaeolojiy: 177-196. D. vol. M. LEONE. 1999 HEGMON. The FOl/r HI/I/dred SOl/gs of l.R.. Printworld. S. Hathigl/lUpha II/scription of KharalJl'la alld the Bhabrll Edict of Asoka: A Critical Stl/dy.P. ed. Madras: University of Madras. Taddei.. 13. P. Spencer-Wood. J. L.C. Schotsmans and M. and S.. KULKE.R. PorrER Introduction: Issues in historical archaeology. Stark. KANT. lI/di01l Historical Review 3(1):1-15. 13. CROSBY Middle-range theory in historical archaeology.J. Funari. AND C. TAMIL KINGDOMS OF HISTORIC SOUTH INDIA 221 GURUKKAL. Naples: Istituto Universitario Orientale Dipartimento Di Studi Asiatici. Potter. P. J. London and New York: Routledge. 1. 1991 HART. JOllmal of the I//Stitl/te SOl/th Asiall Stlldies 12(1): 136-155.ABRAHAM . M. Kerala: Department of Cultural Publications. in The Archaeology of Social BOl/lldaries: 264-270. style. Calcutta: Rupa and Co. P. 2002 JONES. P. Thiruvananthapuram. in Historical Archaeology: Back froll/ the Edge: 219-232. jones. 1971 1985 [1901J AllriCllt Illdia as Described iI/ Classical Literatl/re. R. S. HEIFETZ. New York: Columbia University Press. McINTOSH. Out of many. jANUSEK. Madras: University of Madras. L. 1974 L. D. Report on the Excallatiolls ill the LOllier Kaveri Valley (Til'llkkall/pl/li)'1IJ" O/Id Alajiarai). AND 1986 LEONE. Hall. one: Style and social boundaries in Tiwanaku. W. 1987 P.

1999 'Indianization' from the Indian point of view: Trade and cultural contacts with Southeast Asia in the early first millennium CEo jOllmal of the Ecollolllic alld Social History of the Oriellt 42(1): 1-26. New Delhi: SMITH. MUNRo-HAY. Brill.D. K. Varanasi: Ganga Kaveri Publishing House. Archaeology of Tall/il Nadll (Kongu COlllltry). 1965-69. 217. L. S. and London: Smithsonian Institution Press. H. 011 the Tall/illladll Coast). Reade.H. -A. ed. and the tyranny of the historical record: Danish state formation through documents and archaeological data. 1997 Historians. Thapar. Excavations at Uraiyllr (Tiruchirapal/i). Inscriptions as artifacts: Precolonial South India and the analysis of texts.C. Delhi: Oxford University Press. Delhi: Sundeep.. Leiden: E.222 ASIAN PERSPECTIVES . prehistorians. U. S. V. 1989 Introduction: Archaeological approaches to cultural identity. 1986 1994 ROI/lall Ecollomic Policy ill the Erythra Thalassa. RAY. Romall Kamr. 1992 Black gold: South Asia and the Roman maritime trade. SIDEBOTHAM. Washington. SOUNDARA RAJAN. SHENNAN. London and New York: Kegan Paul International and The British Museum. AND M. Shennan. jOllmal of Archaeological Method alld Theory 4(3-4): 215-237.C. K. 1994 1980 1988 RAMACHANDRAN. Stark. 1995 RAJAN. S. London and New York: Routledge. SASTRI. The Winds of Change: Bllddhism and the Mariti/lle Links of Early South India. . SHANKS. THUIlSTON. K. J. R. M. NAGASWAMY. Cambridge: Cambridge University Press. Man and Environment 1994 1995 14(1): 103-108. M. C. N. S. S. R. ed. 1997 Asiall Perspectives 32: 312-353. 1994 Megalithic Cllltllre of SOllth Illdia: Socio-EcollolI/ic Perspectives. in RecCllt Perspectives 011 Early II/diall History: 142-167. in The Archaeology of Social BOlllldaries: 1-11. Delhi: Book Indian Publishing Co. jOl/rnal of the American Orielltal Society 111(4) :731-736. THAPAR. KaveripattillOm Excavations 1963-73 (A Port City Archaeological Survey of India. R. Madras: Oxford University Press. R. J. SALOMAN. FALL 2003 MOORTl. P. MORRISON. RAMAN. D. LYCETT 1994 Centralized power. Bombay: Popular Prakashan. centralized authority' Ideological claims and archaeological patterns. 1989 Early historical settlement in the Deccan: An ecological perspective. M. S. in Archaeological Approaches to Cllltllral Identity: 1-32. M. AND C. Madras: Brahad Prakashan.. D. Archaeology of South India: Tamil Nadu. E. 30 B. 1991 Epigraphic remains of Indian traders in Egypt. Trade and contacts. 1966 1987 A History cif SOllth [lldia. Madras: University of Madras. 1996 Aksumite interests overseas. STAIlK. TILLEY Recollstmcting ArcllOeology: Theory alld Practice. J. ed. SOllth Asia 15(2): 1-27. ed. 1998 Technical choices and social boundaries in material culture patterning: An introduction. K.. V.K. jOllmal of Archaeological Method alld Theory 4(3-4): 239-263. in Illdiall Oceall ill Alltiqllity: 403-416. 42(2) . T.

300). 300 B. Prevailing analyses of early Tamil society have relied heavily on the historical texts. KEYWORDS: South Asia. South India.. D. Identity and material culture in the later prehistory of Central Europe. Kerala. TAMIL KINGDOMS OF HISTORIC SOUTH INDIA 223 WELLS.C. the most important documentary source for information on early South lndian culture is a body of prose poetry known as the Sangam anthology. often at the expense of critically examining the material culture from Kerala and Tamil Nadu. The reaction against analogy. K. JOllrJIal or Archaeo· logical Research 6(3): 239-298. 1973 WYLIE.ABRAHAM . Not surprisingly.D. this is also the period to which can be traced the first indications of the concept of a "Tamil" identity in South India. IRAN 11. 1985 ZVELEBIL. to A. which represents a key stage in the development of South Indian material culture. history and archaeology. Leiden: E. This study examines the relationship between South Indian archaeology and history and argues that any framework for interpreting early Tamil identity must acknowledge the important qualitative differences in the ways that texts and arti£1cts construct and reflect ethnic identity. J. and comprise the earliest extant examples of Tamil literature. and that archaeologists and historians must analyze their respective data sets within the larger social. Tamil Nadu. A. COII/pal/iol/ Stl/dies to the History 1992 of Til/I/il Literatl/re. cultural-ethnic identity. Brill. WILLIAMSON Sasanian maritime trade. Archaeologically. political. Advallces ill A rchaeological Method alld Theory 8: 63-111. ABSTRACT For the southern Indian states of Kerala and Tamil Nadu. 1998 P.D. AND A. and economic practices of early Tal11ilakal11. Tal11ilakam. the Tamil Sangam era corresponds roughly to the late Iron Age-Early Historic period (c. . These indigenous texts date to the first few centuries A. WHITEHOUSE. S.

Sign up to vote on this title
UsefulNot useful