Hunter-Gatherer Adaptations in Madurai Region,
Tamil Nadu, India: From c. 10,000
The Pleistocene-Holocene transition witnessed several signiﬁcant modiﬁ-
cations in human adaptation patterns, including a shift to microlithic implements,
small-game hunting and gathering, and a highly mobile lifestyle in several parts
of the world. The term Mesolithic is widely used to denote this cultural period,
which succeeded the Palaeolithic and preceded the emergence of an agropas-
toral way of life (the Neolithic). The transformation from hunting-gathering (the
Mesolithic) to agropastoralism (the Neolithic) that took place in certain territories
during the early Holocene was not a universal phenomenon, but a localized one.
It is a gradual process that has been going on for several millennia worldwide.
Evidence for Mesolithic and post-Mesolithic hunter-gatherers occurs exten-
sively in di¤erent parts of world including South Asia. In India, from the 1880s
when A.C.L. Carlleyle discovered the ﬁrst microliths in the Vindhyan hills (Misra
1965 : 58), several microlithic sites of the Mesolithic period have been brought
to light (Allchin and Allchin 1983 : 62–96; Misra 1965, 1989). The survival of
microlith-using groups, primarily subsisting on hunting and gathering and their
interactions with agropastoral communities of the protohistoric and historical
periods have been recognized in archaeological context in several areas like
Rajasthan, Gujarat, Central India, and Andhra Pradesh (Allchin 1977; Allchin and
Allchin 1974; Hooja 1988; Jacobson 1970, 1975; Misra 1973, 1976; Murty 1989;
Recent investigations in the hitherto archaeologically unexplored region of
southern Tamil Nadu have exposed a wide spectrum of evidence for human
occupation during the Mesolithic and the Iron Age–Early Historic periods (Selva-
kumar 1996, 1997, 2000). Based on these ﬁndings, this paper discusses the adap-
tation patterns of the hunter-gatherer groups, which occupied the Gundar Basin
(Madurai district of Tamil Nadu) from the early Holocene to the end of the Early
Historic period, i.e., up to c. a.d. 500.
The term Iron Age here refers to the period that falls between the introduction
V. Selvakumar is a post-doctoral fellow in the Department of Archaeology, Deccan College, Pune,
Asian Perspectives, Vol. 41, No. 1 (2002 by University of Hawai‘i Press.
of iron and the beginning of the Early Historic period. In South India, the Iron
Age is placed within a tentative time bracket of c. 1000 to 300 b.c., and the Early
Historic period, between c. 300 b.c.andc.a.d. 500 (Allchin and Allchin 1983).
the study area: environmental background
This study focuses on an area of about 400 km2(between 77350to 77550E
and 9400to 955 0N), in the upper Gundar Basin, which occupies the western
part of the Madurai district of Tamil Nadu (Figs. 1–2). The Varushanad-Andipatti
hills, an o¤shoot of the Western Ghats, and its branches, the Chaturagiri and
the Kudiraimalai (or the Vasimalai), enclose the area respectively, on the western,
southern, and northern sides. The basin has a gneissic base intruded by Char-
nockite and Khondalite groups of rocks (HBG 1976) and is drained by the
seasonal river Gundar and its tributaries, the Varattar and the Uppar. It can be
divided into two physiographic zones, the plains with a varying altitude of 150 to
250 m above mean sea level (hereafter, amsl) scattered with a few isolated hil-
locks, and the hilly tracts, which cover the area above 250 m amsl.Theclimate
of the area falls between semi-arid and sub-humid types with an annual average
precipitation of around 900 mm. Most of the rainfall occurs during the northeast
monsoon season (October to December) and a small amount in summer (April
to May). The ombrothermic diagram for the meteorological station at Madurai,
situated 50 km northeast of the basin, shows two dry seasons separated by wet
seasons (Gaussen et al. 1964).
The plain area of the basin is inhabited by agropastoral communities and is
largely under cultivation. It has no forest cover and hence today there are very
few wild animals except hare, jackal, and fox. The hilly tract has mostly denuded
scrub forest and a few pockets of dry deciduous forest, one of which harbors a
small settlement of Paliyans, a food-gathering group (P. M. Gardner 1972, 1988).
It also has numerous wild animals including di¤erent deer species, elephants, and
previous work and objectives of the present research
The southern part of Tamil Nadu has not been subjected to intensive archaeo-
logical research because of its marginalized location and the lack of any signiﬁcant
ﬁndings in the sporadic surveys conducted before the 1980s. Previous research on
the prehistoric front concentrated on the microlith-bearing Teris (red sand-dune
formations) situated along the southeast coast of India (Aiyappan 1944; Foote
1883, 1916; R.A.M. Gardner 1986, 1995; R.A.M. Gardner and Martingell 1990;
Zeuner and Allchin 1956).
Systematic archaeological research commenced in this basin mainly in the late
1950s when Raman (1970) carried out surface surveys. Before the 1950s, R. B.
Foote came across a few sites with lithic artifacts in the area, which lie to the east
and southeast of the basin, during his geological surveys in the late nineteenth
century (Foote 1883, 1916). However, Raman’s survey is the ﬁrst problem-
oriented investigation to be conducted in southern Tamil Nadu (Indian Archaeol-
ogy: A Review 1957– 1958 : 38, 1958– 1959 : 31–32, 1960 –1961 : 18– 19; Raman
1970). His surveys undertaken with the objective of reconstructing the culture-
asian perspectives .41(1) .spring 200272
history of the region, exposed the presence of the Mesolithic and the Iron Age–
Early Historic (or the Megalithic) periods and the absence of the Neolithic and
the Chalcolithic periods in the Gundar Basin. His studies also revealed two sites,
which he ascribed to the Middle Palaeolithic (Series II).
Despite Raman’s initial work, which established a preliminary cultural se-
quence, several aspects, especially the chronology and subsistence and settlement
patterns of the ancient cultures of the Gundar Basin in particular and southern
Tamil Nadu in general, have been largely ignored. Against this background, the
research under discussion was initiated. Furthermore, the studies conducted in the
1980s and 1990s on the Teri sites (R.A.M. Gardner 1986, 1995; R.A.M. Gardner
and Martingell 1990; Zeuner and Allchin 1956), as well as the recent research in
Sri Lanka (Deraniyagala 1992 : 118), also necessitated investigations in the interior
southern Tami Nadu for fresh evidence.
The investigations involved intensive surface surveys for six seasons and test
diggings at four sites. Due to lack of resources, systematic survey techniques (see
Ammerman 1981; Dunnell and Dancey 1983; Schi¤er et al. 1978) could not be
used in this study. Hence, a di¤erent strategy known as ‘‘purposive technique’’
(Schi¤er et al. 1978: 5), in which the areas around critical resources are surveyed,
was used. As a part of this strategy the neighborhood of the critical resources such
as plant and animal foods and water sources, and also the vicinities of caves and
rockshelters, hills and rock outcrops were intensively surveyed for detecting
archaeological remains. The critical resources were identiﬁed based on map study
Fig. 1. Areas surveyed in the upper Gundar Basin.
selvakumar .hunter-gatherer adaptations in tamil nadu 73
and enquiries made to the local people. A 2-km radius around the resources was
around the resources, which have dense scrub forest cover and hilly terrain, could
not be surveyed because of poor visibility and accessibility. An area of 1 km from
the edge of hillocks and rock outcrops was surveyed. In order to avoid any bias
that normally emerges in this type of survey, the black-soil plain lying in the
middle of the basin, which is devoid of any rock outcrops and critical resources,
was also surveyed at three locales. Though systematic and intensive survey meth-
ods could not be adopted, the signiﬁcance of recording the distribution of artifacts
on the landscape rather than the few points generally known as ‘‘sites’’ was taken
into consideration and e¤orts were made to identify and record the low-density
scatters and the isolated cultural materials, which are also known as ‘‘o¤-sites’’
(Foley 1981; Rossignol and Wandsnider 1992). In addition, the sections exposed
in the streams and wells were also inspected to identify buried artifact horizons.
Density of artifacts and the size of the area were the criteria used to distinguish
‘‘sites’’ and artifact scatters.
Surface surveys have revealed 109 microlithic sites in this basin. These microlithic
sites have been broadly divided into three groups, Category 1, Category 2, and
Category 3, based on their size and artifact density. On the fringes of the basin
at several localities, where artifacts are found continuously in varying concen-
trations, it was di‰cult to demarcate the site limits. These localities have been
demarcated and are shown as dotted areas in Figure 2.
Category 1: Sites with stray ﬁnds of a solitary or a few artifacts are grouped into
this category. Thirty-four such o¤-sites have been located. However, there may
be many more sites of this category buried and thereby remain undiscovered. The
majority of these sites are located away from the critical resources and are found
on the fringes of the basin and a few in the black-soil tract, which lies in the
middle of the basin (Fig. 2).
Category 2: Surface scatters, with a low artifact density and covering a speciﬁc
area but without any thick sediments, fall under this group. There are 66 such
sites, which vary in size from 6 m2to 9600 m2. These sites are also found mostly
on the fringes of the basin. However some of the sites occur in the vicinity of
critical resources, while others are away from such resources.
Category 3: Sites with thick deposits or a dense scatter of artifacts covering a
large area, which resulted due to multiple episodes of occupation, come under
this group. There are nine sites in this category: three are rockshelters and the rest
are open-air sites. Their size varies between about 8 m2and 7000 m2.Sitesof
this category are situated near resources such as water and rockshelters or in the
vicinity of strategic locations like elevated ground and the hillocks, which provide
a commanding view of the surrounding area.
Test excavations were undertaken at the microlithic rockshelter sites of Kanavay-
patti, E. Gopalapuram, and Virappatti, and at the open-air site near S. Pappi-
asian perspectives .41(1) .spring 200274
nayakkanpatti in order to understand their chronological and cultural contexts
and also the nature of occupation.
Kanavaypatti (long. 775202000 E; lat. 95507200 N).Atrenchmeasuring1.5-
by-1.5 m was excavated in the northern part of the shelter formed by a massive
monolithic rock (approximate height, 10.4 m) lying 2 km northwest of the town
posit exposed in this trench was strongly cemented with calcrete and consisted of
several occupation ﬂoors. Apart from the artifacts, the sediment also contained
several rock fragments derived from the exfoliation of the monolith and sev-
eral gastropod shells of nonhuman origin.
The nature of the evidence found in
this trench suggests that this shelter witnessed numerous episodes of occupation
spanning several centuries. Despite the absence of any major variation in stratig-
Period I, the Mesolithic, has yielded a large number of microlithic artifacts,
principally made of quartz, hammerstones, wild animal bones, charcoal fragments,
and burned clay pieces. Domestic animal bones were absent, apart from a sheep/
goat bone, which may be considered a later intrusion.
Period II, the Iron Age–Early Historic, brought to light 32 pottery fragments
of the Iron Age and Historic periods, several domestic and wild animal bones, and
a few dressed stone slabs and grinding stones, besides the artifacts and features
E. Gopalapuram (long. 774102900 E; lat. 954 01100 N). E. Gopalapuram lies
2 km northwest of Elumalai, which is 18 km from Usilampatti by road. Two
Fig. 2. Microlithic sites in the upper Gundar Basin. Site Categories: Category 1 ¼C, Category
2¼U, Category 3 ¼g.
selvakumar .hunter-gatherer adaptations in tamil nadu 75
kilometers northeast of E. Gopalapuram at the foot of the Kudiraimalai hill, there
is a mushroom-shaped monolithic rock, known as Rajankal, with two overhangs
on its eastern and western faces (Fig. 3). Two test trenches, each measuring 1
m2—one in the western shelter and the other in the eastern—were excavated.
These trenches produced microliths along with a few fragments of Iron Age and
Historical period pottery from the lower-most levels. The trench excavated in the
western shelter (Fig. 4) had a very limited number of artifacts in the lower levels,
whereas its upper levels had a relatively higher quantity of artifacts.
Virappatti (long. 77530900 E; lat. 952 056 00 N). Virappatti is located 2 km
east of Mangalarevu, which is 17 km south of Usilampatti on the Peraiyur road. A
few microlithic scatters are found on the northwestern side of the village near a
hillock (197 amsl). There is also an Iron Age–Early Historic (Megalithic) burial
(cairn-circle) near the hillock. One hundred meters further southeast of the hill-
ock is a massive monolithic rock, known as ‘Nayakkanparai,’ which provides
shelter on the northwestern side. Very few microliths and Iron Age–Early His-
toric period pottery are found on the surface of the shelter. With a view to
ﬁnding out the relationship between pottery and lithic artifacts, a trench measur-
ing 2-by-1 m was dug in this shelter. This trench exposed a 0.7-m-thick deposit
belonging to the Iron Age–Early Historic period. It had a large amount of micro-
lithic artifacts, a small quantity of Iron Age–Early Historic pottery and animal
bones, traces of iron smelting activities such as iron slag, and a few fragmentary
terracotta pipes and iron objects. Fig. 5 shows the test trench with an iron object
S. Pappinayakkanpatti (long. 774201100 E; lat. 94805200 N). S. Pappinayak-
kanpatti lies 3 km southeast of Elumalai. A trench measuring 2-by-2 m was exca-
vated in the Early Historic agropastoral settlement situated about 1 km southwest
of the village near a hillock (Fig. 6). It revealed an aceramic microlithic horizon,
1 to 1.80 m below surface level, underneath an Early Historic deposit. The dis-
Fig. 3. View of the rockshelter (Trench I) at E. Gopalapuram. Such rockshelters abound in the
upper Gundar Basin. The ancient hunter-gatherers used the shelters, which are located close to
asian perspectives .41(1) .spring 200276
Fig. 4. The test trench excavated on the western side of the rockshelter at E. Gopalapuram. This
trench yielded quartz artifacts mixed with historical pottery.
Fig. 5. The trench at Virap-
situ. Several groups at di¤er-
ent points in time perhaps
used this shelter.
covery of the microlithic horizon was unexpected, as no clues to the buried
microliths were on the surface. The microliths occur in a red loam formed due
to in situ weathering of granite rocks. The top level of the Mesolithic occu-
pation was heavily disturbed up to a depth of 1.30 m by the pits dug during
the Early Historic period. This Mesolithic level yielded mostly microliths and a
few small teeth fragments of some unidentiﬁable animal species; but no manu-
ports or heavy-duty tools (large tools, more than 5 cm in size) were found. At a
depth of about 1 m below surface level, fragmentary remains of articulated human
femur bones were uncovered in a highly decayed condition. The Early Historic
settlers subsequently disturbed this burial, which seems to belong to the Meso-
nature of the archaeological record and site contexts
The location of the Gundar Basin in the rain shadow of the southwest monsoon
pattern and the smaller size of its catchment and drainage, which are some of
the characteristics of a low-energy environment, have minimized the impact of
natural agencies on the archaeological record. The activities of the natural forces
acting upon the basin at present provide valuable clues to understand the natural
processes that could have contributed to the present-day conﬁguration of the
archaeological record. Observations made in the basin show that erosional activ-
ities are intensive near the rills and the streams, and the depositional activities are
active in the low-lying areas. Hence, a relatively higher degree of modiﬁcation,
like displacement and burial of artifacts, speciﬁcally of the low-density scatters
(Category 1), might have taken place mostly in the neighborhood of the rills,
streams, and rivers, and more so in the eastern part of the basin, below the 140 m
contour as noticed at T. Pudupatti. At this site both rolled and fresh artifacts,
Fig. 6. The trench at S. Pappinayakkanpatti. Here the Mesolithic deposit underlies the Early His-
asian perspectives .41(1) .spring 200278
perhaps derived from two di¤erent contexts, occur together. Other sites which
are in the rockshelters and in the uplands, away from the intensive erosional range
of the streams and rivers, appear to have remained largely una¤ected. Supporting
evidence for this observation is the nature of the artifacts including their freshness
and also the microartifacts recovered from the excavated sites. Among the animal
bones found in the sites, only one bone fragment from Kanavaypatti showed
gnaw marks. Despite these minor disturbances, it can be concluded that the ma-
jority of the sites discovered so far in this basin, except for T. Pudupatti, were
probably a¤ected by natural agencies only to a limited extent.
The impact of cultural activities of succeeding periods is also an important
factor contributing to the formation of the archaeological record of a particular
period. Some of the determinant factors of settlement patterns such as preference
to establish the settlements adjacent to water sources, which are similar for both
the prehistoric and modern hunter-gatherer communities, cause the frequent
reoccupation and reuse of ancient sites located near critical resources.
decreases the visibility of the earlier portions of the archaeological record as
observed at S. Pappinayakkanpatti, where no clues for the Mesolithic horizon
covered by the Early Historic deposit were found on the surface. Similar buried
deposits can also be expected beneath a few of the Iron Age–Early Historic and
Medieval habitation sites of this basin. The Iron Age–Early Historic people have
also chosen some of the pre–Iron Age microlithic sites for burying the dead, and
as a result modiﬁed the earlier remains considerably.
Cultivation and related activities, which have been going on in this basin
for several centuries, too, have considerably a¤ected the archaeological sites. The
larger artifacts (more than 4 cm across) that hinder agricultural activities are gen-
erally collected by the farmers and discarded along the boundary of the ﬁelds,
while the smaller ones are left untouched. Plowing buries the microliths and
a¤ects their visibility, more in the case of the low-density scatters. Some of the
uncultivated areas near the hill fringes, which had a few microlithic scatters (site
Categories 1 and 2) in 1992–1993, became completely invisible in 1996 after they
were brought under cultivation. Although plowing does not seem to cause large-
scale displacement of the microliths, the removal of vegetation cover, which pro-
tects the microliths from erosional activities, indirectly exposes them to displace-
ment in the areas that have a steep gradient.
One of the main issues of the present research concerns the cooccurrence of
artifacts such as pottery, domestic animal bones, lithic artifacts, and iron objects at
the rockshelter sites. It is necessary to consider the formation processes under cir-
cumstances such as physical association, which can arise as a result of diverse fac-
tors, does not always warrant an inference of contemporaneity. Here the role of
natural agencies in causing this association seems to be very meager, and one or
more of the following reasons could have been responsible for generating such an
1. The material evidence in the rockshelters resulted only due to activities like
occasional camping by the (iron using) agropastoral groups.
2. Intrusion of pottery and domestic animal bones into the pure Mesolithic
record during a later period.
3. Occupation of these sites by hunter-gatherers, agropastoral groups, and
other groups on di¤erent occasions.
selvakumar .hunter-gatherer adaptations in tamil nadu 79
4. Interactions between hunter-gatherers and agropastoralists.
5. Microlith-using agropastoralists (impoverished, without access to iron).
The ﬁrst two propositions assume that the hunter-gatherers had no role in the
formation of these deposits and they were not contemporaneous with agropas-
toral groups. However, in archaeological context there is clear evidence for the
presence of hunter-gatherers during the Iron Age–Early Historic period. The
stratigraphic evidence from the excavated sites also shows that these materials
were not later intrusions into the Mesolithic record.
It appears that there were at least two distinct groups, the agropastoralists and
hunter-gatherers, living in the basin during the Iron Age–Early Historic period.
Though identiﬁcation of (‘ethnic’) groups in archaeology is generally considered
highly risky (Atherton 1983; DeCorse 1989; Hassan 1987 : 3; Shennan 1989), in
this context there are valid reasons for distinguishing the archaeological sites of
the agropastoralists and hunter-gatherers. First, the agropastoral groups, who had
iron technology, need not have engaged in such large-scale quartz knapping as
seen in the rockshelter sites, though they used a limited range of lithic (mostly
made of chert) and bone tools. Second, there is a temporal continuity in the oc-
cupation of the rockshelter at Kanavaypatti from the Mesolithic to the Iron Age–
Early Historic period and technological similarity of lithic industries, which point
to the continuing existence of hunter-gatherers in the post-Mesolithic period.
The characterization of agropastoral communities in the Early Historic Sangam
Tamil literature suggests that the possibility of these sites being created by
microlithic-using, impoverished pastoralists is low (Singaravelu 1969; Zvelebil
1975). The microlithic sites in the basin can be grouped into three contexts
periodization and chronology
The Mesolithic phase is the earliest cultural phase that has been clearly repre-
sented in this region and no clear-cut evidence for the Palaeolithic has been dis-
covered so far. It has to be mentioned here that Raman (1970) assigned the
heavy-duty artifacts he found at T. Pudupatti and Sivarakottai to the Middle
Palaeolithic (Series II). During the recent investigations, a limited number of sim-
ilar heavy-duty artifacts were noticed at Perumalkoilpatti and Vittilpatti in surface
contexts together with microliths. It appears that these artifacts could have formed
a part of the tool kit of the Mesolithic groups rather than representing a di¤erent
techno-tradition, as evidence for the association of microliths and heavy-duty
tools are available from the eastern parts of India (Mohanty 1993; Ota 1986).
Table 1.Chronological Context of the Microlithic Sites from
the Upper Gundar Basin
context name of the sites
Mesolithic or Pre–Iron Age Kanavaypatti Period I and S. Pappinayakkanpatti Period I
Iron Age and Historical E. Gopalapuram, Virappatti, and Kanavaypatti Period II
Mixed or context uncertain All the surface scatters (Categories 1 and 2) and the
unexcavated high-density sites (Category 3)
asian perspectives .41(1) .spring 200280
However, more investigation is necessary in this regard. Some of the workers
have dated all the microlithic sites found in this area to the Mesolithic period
(Raman 1970; Vedachalam 1985). Earlier, Zeuner and Allchin (1956: 20) had
postulated a sequence of three microlithic industries for the Tamil Nadu coast.
Raman (1970) disagreed with them regarding the presence of the third micro-
lithic industry, associated with the Neolithic culture in southern Tamil Nadu.
Citing the absence of evidence for the Neolithic sites there, he contended that all
the microlithic sites that were brought to light during his surveys in this region
belonged to the Mesolithic period. From the stratigraphic and artifactual evidence
recovered during the recent investigations, it is clear that the microlithic sites of
this basin can be ascribed to two chronological contexts, the Mesolithic and the
Iron Age–Early Historic periods (Selvakumar 2001).
Evidence for the Mesolithic period was found in the test excavations at S.
Pappinayakkanpatti and Kanavaypatti. While at S. Pappinayakkanpatti the Meso-
lithic deposit lies below the Early Historic remains of agropastoral groups, at
Kanavaypatti it is overlain by a deposit comprising microliths, animal bones, and
pottery, which was mostly a result of the occupation of the Iron Age–Early His-
In the Iron Age–Early Historic period, microliths occur in two distinct cultural
contexts. A limited quantity of microliths found along with a large amount of
pottery, and other artifacts in the village-type habitation sites are identiﬁed as the
settlements of agropastoral groups (not included on the map here, see Selvakumar
2000: Fig. 2). The rockshelter sites, which have a large quantity of microliths and
a few pottery fragments and located mainly in the marginal areas of the basin,
were mostly occupied by hunter-gatherers, and agropastoral groups could have
periodically used these sites as well. Evidence for the latter context is found at
Virappatti, E. Gopalapuram, and Kanavaypatti Period II.
There are no radiocarbon dates available for the Mesolithic and post-
Mesolithic hunter-gatherer sites of the upper Gundar Basin. The two charcoal
samples from the test trench at Virappatti, assayed by the Physical Research Lab-
oratory, Ahmedabad, turned out to be modern (PRL 1800 and 1801).
absence of radiocarbon dates, the Mesolithic remains can only be tentatively dated
on the basis of the dates given to other Indian Mesolithic sites. In India, with the
support of several radiocarbon dates, the Mesolithic period is generally placed be-
tween c. 10,000 b.p. and c. 4000 b.p. (for a detailed discussion, see Deraniyagala
1992 : 110–112; Kennedy et al. 1992; Possehl 1994). The ﬂuorine-phosphate
analyses of bones from Mesolithic sites of the upper Gundar Basin also show
values similar to those from other Indian Mesolithic sites (Kshirsagar 1993; Selva-
kumar 1997: Table 60). Therefore, taking these facts into consideration, the be-
ginning of the Mesolithic in the Gundar Basin can be tentatively dated to the
early Holocene period. However, it has to be noted that the neighboring Sri
Lankan microlithic sites have much older dates, i.e., 28,500 b.p. (Deraniyagala
1992 : 118, 472–479). Unlike the Teri sites of southeast India, where microliths
occur on the surface of the red sand dunes, the Sri Lankan sites yield microliths
below the sand dunes. In India, very few Mesolithic sites have given early Holo-
cene dates. Most of the radiocarbon dates are in the post-mid-Holocene time
bracket. Hence, there is no substantial evidence at present to date the Indian
microlithic sites beyond the Holocene.
selvakumar .hunter-gatherer adaptations in tamil nadu 81
The presence of the Iron Age and Early Historic pottery at a few microlithic
sites is a clear pointer to the continuing survival of hunter-gatherers up to the
Early Historic period. In spite of the absence of radiocarbon dates from the Gun-
dar Basin, the stratigraphic and ceramic evidence suggests that the Iron Age must
have begun before the mid-ﬁrst millennium b.c. For dating the Early Historic
period, there is ample data, including a radiocarbon date from the Early Historic
agropastoral settlement at S. Pappinayakkanpatti (PRL NO. 1965: 2040 G150
Based on this evidence, the Iron Age–Early Historic period can be dated
to c. 500 b.c.toa.d. 500. It is uncertain whether the microliths were used by
hunter-gatherers as late as the Medieval period. It is interesting to note that the
Paliyans, who live in the Varushanad-Andipatti hill range, were using quartz and
steel for making ﬁre until a few decades ago (P. M. Gardner 1972 : 417). More
investigation on the sites occupied by the Paliyans during the Medieval and re-
cent periods would reveal whether they used the lithic artifacts during the post–
Early Historic period.
Ethnographic data on the Paliyans and other groups living in the Gundar Basin
have been used in this study to understand hunter-gatherer adaptations during
the Mesolithic and Iron Age–Early Historic periods. The Paliyans are a hunter-
gatherer group living in the Varushanad-Andipatti hills and on its fringes and
in other hill areas of southern Tamil Nadu (Census of India 1969; Gardner 1972;
Thurston and Rangachari 1909). P. M. Gardner’s work on the Paliyans focused
on their socio-economic aspects (Gardner 1965, 1966, 1969, 1972, 1982, 1984).
The Paliyans used to disperse into small groups in the summer months because of
a reduction in their food supply. The contemporary Paliyans do not strictly fol-
low a hunting-gathering way of life and very few of them live in the hills, which
is their original habitat. Most of the group lives on the hill fringes in concrete
houses built by the state government of Tamil Nadu and collect forest products
and sell them in the local market. They exploit their original habitat in groups of
two to ﬁve individuals and then move back to the settlements at the end of the
day. The ethnographic study undertaken by the author focused on the subsistence
and settlement patterns of the Paliyans and their interactions with the agropastoral
groups. Information regarding the ancient mobility pattern of the Paliyans was
collected on the basis of the enquiries made to elderly Paliyans.
the lithic industry
The lithic industry of hunter-gatherers of the upper Gundar Basin is dominated
by the microlithic elements consisting of both geometric and nongeometric tools,
mostly fashioned on ﬂakes (Figs. 7, 8). This industry also has a limited amount of
heavy-duty tools, largely made on cherty quartzite. A total of 13,260 artifacts
consisting microliths and heavy-duty tools were collected from the basin by three
di¤erent means (Table 2).
Since the uncontrolled (random) collection under-
taken in the ﬁrst season (in 1992) could not be considered a representative sample
of the lithic industry of hunter-gatherers, a systematic collection of artifacts was
made at four sites from a total of 38 grids (the grid size was 1 m2)fromthree
asian perspectives .41(1) .spring 200282
Category 3 sites, and four grids from a Category 2 site. However, in the case of
low-density scatters, the random surface collection can be treated as representative
in view of the fact that at several of the low-density scatters a majority of the
artifacts were collected.
Because of the mixed character of assemblages it was di‰cult to decide on any
speciﬁc reasonable unit for analyzing the lithic artifacts. At one level, the entire
Fig. 7. a–d: retouched blade; e: backed blade; f–g: triangles; h: lunate; i: penknife; j–k: burin; l–m:
point; n: borer.
selvakumar .hunter-gatherer adaptations in tamil nadu 83
Fig. 8. a, c: chopper; b: knife; d: hammerstone; e: grinding stone; f: ringstone; g–h: blade core.
collection, which can be taken to represent the lithic industry of hunter-gatherer
groups of the Mesolithic and the Iron Age–Early Historic periods, was treated as
one unit and analyzed. At another level, the assemblages from the excavated sites,
which can be chronologically segregated into the Mesolithic and Iron Age–Early
Historic periods are analyzed separately and the results are presented at the end of
this section (Table 3 and Table 4). Only a preliminary analysis of the artifacts is
presented here. A modiﬁed version of the classiﬁcation schemes of Clark and
Kleindienst (1974 : 71–105) and Sankalia (1964) is employed for the analysis of
the artifacts. The term ‘heavy-duty’ is used to refer to artifacts that are larger than
5 cm across. The term ‘chip’ means tiny ﬂakes, which are by-products of sec-
ondary working and generally smaller than 3 cm. The term ‘chunk’ refers to
amorphous quartz blocks with ﬂake scars. Because microwear (use-wear) analysis
was not carried out, function is assigned to the artifacts on stylistic and morpho-
The artifacts have been broadly classiﬁed into ﬁnished implements, modiﬁed
pieces, and debitage.
Finished implements include artifacts that have speciﬁc design or shape and
bear some amount of secondary working. Under this class fall blades, crescents,
triangles, penknives, burins, points, borers, scrapers, choppers, and ringstones,
which form 2.66 percent (354) of the 13,260 artifacts collected. Out of the total
199 microlithic implements other than the blades, only 17 are made on blades and
the rest are on ﬂakes.
Modiﬁed pieces encompasses blades and ﬂakes possessing traces of modiﬁcation
and use, and hammerstones, grinding stones, and querns, representing 4.01 per-
cent (532) of the collection.
Simple ﬂakes/blanks, chips, chunks, nodules, crystals, core rejuvenation ﬂakes,
and blade-core preparation ﬂakes are grouped in the debitage category. They
constitute a major proportion of the lithic material, constituting 93.31 percent
(12,373) of the collection.
Choice of Raw Material
Quartz is the principal raw material used by the hunter-gatherers in the Gundar
Basin (Fig. 9). It appears that the paucity of easily workable raw materials, such as
chalcedony and chert, compelled hunter-gatherers to opt for quartz, which occurs
profusely in this area. Irrespective of its numerical dominance, quartz is compara-
tively less represented among the ﬁnished implements than any other material.
Due to its intractable nature, it yields a large quantity of debitage. A similar
Table 2.Sources of the Lithic Collection
sources of the artifacts number of artifacts collected
Uncontrolled surface collection 1271
Systematic surface collection 923
From the test trenches at Kanavaypatti, Virappatti,
E. Gopalapuram, and S. Pappinayakkanpatti
selvakumar .hunter-gatherer adaptations in tamil nadu 85
Table 3.Frequency Distribution of Lithic Artifacts from Mesolithic Hunter-Gatherer Sites of the Upper Gundar Basin
quartzite chert jasper chalcedony gneiss total
Blade 11 2 — — — — 13
Burin 4 1 — — — — 5
Lunate 2 — — — — — 2
Triangle 10 1 — 1 — — 12
Borer 5 — — — 1 — 6
Point 32 — 1 — 1 — 34
Scraper 6 1 — — — — 7
Chopper 3 1 — — — — 4
Total 75 6 1 1 2 — 85
Modiﬁed ﬂake 56 1 — 2 — — 59
Modiﬁed blade 6 3 — — — — 9
Hammerstone 4 — — — — 1 5
Total 66 4 — 2 — 1 73
Blade core 5 — — — — — 5
Flake core 30 2 — — — — 32
Simple blade 22 3 5 1 1 — 32
Simple ﬂake 32 1 1 — — — 34
Core rejuvenation ﬂake 4 — — — — — 4
Blade core 20 4 — — 1 — 25
Chip 2740 104 18 21 4 — 2887
Raw material 60 3 — — — — 63
Chunk 90 1 — 1 — — 92
Shatter 113 — — — — — 113
Total 3116 118 24 23 6 — 3287
Grand Total 3257 128 25 26 8 1 3445
Table 4.Frequency Distribution of Lithic Artifacts from the Iron Age–Early Historic Hunter-Gatherer Sites of
the Upper Gundar Basin
quartzite chert jasper chalcedony gneiss quartzite total
Lunate 13 — — — — — — 13
Triangle 42————— 6
Borer 31————— 4
Chopper 4 — — — — — — 4
Modiﬁed blade 8 — — — — — — 8
Hammerstone 17 — — — — 8 1 26
Quern — — — — — 2 — 2
Total 97 1 — — — 10 1 109
Flake core 52 1 — 1 — — — 54
Simple blade 47 — 4 2 — — — 53
Simple ﬂake 52 1 1 1 — — — 55
Core rejuvenation ﬂake — 10 — 1 — — — 11
Chip 5754 154 14 16 9 — — 5947
Raw material 518 2 — — — — — 520
Chunk 221 3 1 — — — — 225
Shatter 409 — — — — — — 409
Total 7118 161 25 20 9 — — 7333
Grand Total 7307 166 26 20 9 10 1 7539
situation is common in the quartz-dominated assemblages from sub-Saharan Africa,
Sri Lanka, and the region south of Mysore in India, where other suitable raw
materials are not freely available (Bisson 1990 : 104; Deraniyagala 1992; Hivernel
Quartz has many variants, such as milky quartz, smoky quartz, and clear quartz.
It is available in the basin in the form of veins in granite formations and also as
angular chunks at the base of the hills. Quartz crystals and pebbles also occur in
the basin, but in very small quantities. Milky quartz was used for manufacturing
heavy-duty artifacts, while clear quartz and smoky quartz were preferred for the
manufacture of microliths. Evidence from stratiﬁed contexts suggests that due to
the sharpness of the edges, some of the naturally occurring angular quartz blocks
were also probably used without ﬂaking. This can, however, only be ascertained
through microwear (use-wear) analysis.
Chert, cherty quartzite, and jasper have been used in small quantities and they
form only 10.31 percent of the total collection. Among these materials, chert has
a higher percentage of ﬁnished implements because of its easily ﬂakeable nature.
The source of chert is not known; it probably occurs in association with the
Dharwar rock group or it could be of nonlocal origin. The Vaippar Basin (Viru-
dunagar district), lying immediately south of the Gundar Basin, is reported to
have chert-bearing limestone deposits (Balachandran, pers. comm. 1992), which
could have been exploited. Cherty quartzite, which is available in the basin itself,
was also utilized for making heavy-duty tools. One of the main sources of this raw
material is T. Krishnapuram, where it was noticed in the exposed section of a well.
Pattern of Artifact Distribution
The analysis of the lithic artifacts recovered from the di¤erent categories of sites
revealed the following results.
whereas that of the ﬁnished implements is low. The sites of other categories have
a very small amount of debitage and a relatively higher percentage of ﬁnished
implements (Fig. 10).
2. Category 3 sites are dominated by quartz artifacts, whereas Category 1 and
2 sites have a high percentage of chert artifacts (Fig. 11).
Fig. 9. Frequency distribution of raw materials used in the microlithic industries.
asian perspectives .41(1) .spring 200288
3. The artifact types from Category 3 sites are varied; in contrast, Category 1
and 2 sites have only a few varieties of artifacts. Permanent features like querns
and grinding stones, which are not easily transportable, are conﬁned to the sites of
4. Among the ﬁnished implements, choppers are more frequent (69.69 per-
cent) in the sites of Categories 1 and 2.
5. Out of a total of 72 blade-core preparation ﬂakes, 70 are from Category 3
6. No major pattern is seen in the distribution of blade cores. They are almost
equally distributed; 54.83 percent of blade cores are from Category 3 sites and the
rest are from sites of other categories.
7. In the distribution of simple blades, there is a clear pattern. Of the 111 sim-
ple blades, 103 are from Category 3 sites.
Lithic Industries of the Mesolithic and the Iron Age–Early Historic Hunter-Gatherers
Determination of chronological context was possible only in the case of the lithic
assemblages recovered from the test trenches. Out of 11,066 artifacts from the test
trenches, 3445 artifacts are from the Mesolithic context, 7539 are from the Iron
Age–Early Historic context, and the rest (82) are unstratiﬁed. Furthermore, in the
frequency of di¤erent artifact types, no signiﬁcant pattern was noticed between
Fig. 10. Frequency distribution of artifact categories in di¤erent site categories.
Fig. 11. Distribution of raw materials by site categories.
selvakumar .hunter-gatherer adaptations in tamil nadu 89
the lithic industries of the Mesolithic and the Iron Age–Early Historic periods,
except in the case of querns and grinding stones. At Kanavaypatti querns and
grinding stones are more common in the Iron Age–Early Historic period (two
quern fragments and eight grinding stones) than the Mesolithic, which had only
one grinding stone. In the distribution of various artifact categories, no remark-
able di¤erence was observed: the proportions of the ﬁnished implements, modi-
ﬁed pieces, and debitage in the Mesolithic assemblage are 2.46 percent, 2.11
percent, and 95.42 percent, respectively, and in the Iron Age–Early Historic
contexts, 1.28 percent, 1.44 percent, and 97.26 percent.
A few other patterns were also noticed in the distribution of lithic artifacts: (a)
the Mesolithic assemblage from the trench at S. Pappinayakkanpatti had an ex-
ceptionally high percentage of points, and (b) overall, the Mesolithic assemblages
had a lower proportion of burins, lunates, and scrapers, and a higher proportion
of points and triangles than the Iron Age–Early Historic assemblages. These pat-
terns, which are conditioned by several factors, such as the nature of the manu-
facturing and extractive activities undertaken at the sites on di¤erent occasions,
and the mode of use, reuse, and disposal of the implements, emerging from the
analysis of such a small collection from limited excavations, cannot be taken to
represent the general characteristics of the Mesolithic and Iron Age–Early Historic
industries. A much closer examination of the lithic assemblages might shed more
light on the similarities and di¤erences in the technological and morphological
aspects of these industries.
Several wild and domestic animal bones were recovered from the test trenches
excavated at Kanavaypatti and Virappatti and also in surface contexts at Tach-
chappatti, Andipatti, and Tullukuttinayakkanur (Table 5).
Kanavaypatti: The test trench at Kanavaypatti revealed 163 identiﬁable bones
forming 14.3 percent of the collection, out of which only 79 specimens could be
identiﬁed speciﬁc to species. This collection had both wild and domestic animal
species; Period I, the Mesolithic, had only wild animal bones, except for one
sheep/goat bone, whereas Period II, the Iron Age–Early Historic, had both.
Traces of butchering and charring were noticed on a few bones, while two of
them possessed evidence for the use of sharp metal tools for butchering. No bone
tools were found, except one, which is ground.
Virappatti: Among the animal bones recovered from the test trench dug at this
site, only 11 were identiﬁable, and the rest of the bones were highly fragmentary
splinters. A total of eight bones including one bone each of Bos indicus and Her-
mitragus hylocrius from Tachchapatti, and 25 bones of Bos/Bubalus, small ruminants,
large mammals, and other animals from Tullukuttinayakkanur were collected in
The test trenches at Kanavaypatti, E. Gopalapuram, and Virappatti produced 32,
214, and 911 pottery fragments, respectively, while at Andipatti 374 fragments
were collected from the surface. At Virappatti a few fragments of terracotta pipes
asian perspectives .41(1) .spring 200290
Table 5.Animal Species Represented in the Faunal Collection from the Hunter-Gatherer Sites of the Upper Gundar Basin
sl. no. species common name kanavaypatti virappatti tachchappatti*
1. Antilope cervicapra Black buck 29 3 1?
2. Axis axis Spotted deer/Chital 1 — —
3. Boselaphas tragocamelus Blue Bull/Nilgai 6 — —
4. Cervus unicolor Deer/Sambar 4 — 1?
5. Tetracerus quadricornis Four-horned antelope/Chowsinga 1 — —
6. Gazella bennetti Barasinga 3 — —
7. Hermitragas hylocrius Nilgiri Ibex/Thar — — 1
8. Hystrix indica Porcupine 1 — —
9. Lepus nigricollis Hare — 1 —
10. Vulpes sp. Jackal/fox 1? — —
11. Bos indicus Cattle 14 1 1
12. Bubalus bubalis Bu¤alo 10 1 —
13. Ovis/Capra Sheep/Goat 3 2 1
14. Herpestes edwardsi Mongoose 1 — —
15. Rattus rattus Rat 1 — —
16. Equids (smaller) Horse/Ass 1 — —
17. Varanus sp. Monitor lizard 3 — —
18. Bandicota indica Mongoose 1 — —
19. Lissemys punctata Turtle 1 — —
20. Lammellidens Freshwater mussel 3 — —
*The animal bones from Tachchappatti are surface collection; the rest are from test trenches. The ﬁgure indicates number of identiﬁable specimens (NISP).
and crucibles used in iron-smelting activities were also recovered. Though a few
of the sherds found on the surface and in the upper levels of the test trenches
might belong to the Medieval or modern periods, the fabric and diagnostic forms,
such as bowls, lids, and jars, clearly reveal that most of the pottery belongs to the
Iron Age–Early Historic period.
Artifacts such as a fragmentary bead of amethyst and a double-segmented bead
(modern?) at Kanavaypatti; a tubular glass bead waster (cf. Wheeler et al. 1946 :
99, Fig. 40), a terracotta spindle whorl with the remains of a 2-mm-thick iron rod
inserted through the perforation (cf. Rajan 1994 : 105, pl. 22) from the surface,
and an iron object weighing 1.2 kg from the test trench at Virappatti; and a disc
E. Gopalapuram can be assigned to the Iron Age–Early Historic period. The
Mesolithic level had only a few mica ﬂakes and hematite pieces retrieved from the
test trench at Kanavaypatti. No evidence for hunter-gatherer territoriality such as
rock art is found in this basin. The rock art reported so far in the basin appears to
belong to the historical period.
There is very limited evidence for the reconstruction of the palaeoenvironmental
conditions in the Gundar Basin. Hence, data from the neighboring regions, the
faunal remains from the Mesolithic, the Iron Age and the historical sites, and the
present-day environment are used for this purpose. Evidence from the Teri sites,
which lie about 200 km southeast of the basin, suggests that humid conditions
prevailed in that area around the beginning of the Holocene and the present-day
semi-arid conditions came into being during the middle Holocene (R.A.M.
Gardner 1995). Similar evidence has also been found in the Nilgiri Mountains,
which lie approximately 200 km northwest of the study area (Gupta 1990; Gupta
and Prasad 1985). Several studies on the palaeoclimate show the monsoonal cli-
mate, which prevails in the Indian subcontinent, was intense during the early
Holocene (Roberts and Wright 1993; Singh et al. 1990; Thompson et al. 1993;
VanCampo 1986: 386; Wasson 1995; Wasson et al. 1984). Hence it can safely be
assumed that the present-day rainfall pattern would have persisted during the
early Holocene on an extensive scale. Thus, the pattern of resource distribution
observable in spatial and temporal (annual) scales at present can be projected for
the Mesolithic phase.
To facilitate the study of the settlement system of Mesolithic hunter-gatherers,
a brief account of the resource distribution in space and time observed in the
basin at present is given here. The rainy season (between October and December)
creates bountiful water sources and supports the growth of vegetation throughout
the area. However, during the dry season there is a sharp decline in the available
water and vegetation. In the area east of the upper Gundar Basin (east of long.
77500E) the water sources completely dry up, but the upper Gundar Basin has a
few perennial water sources. The Andipatti hill ranges and its spurs, and also the
isolated hillocks in the upper Gundar Basin, which trap the monsoonal rains and
asian perspectives .41(1) .spring 200292
release the water through several seep springs, provide adequate water supply during
the dry summer months. The summer rains, which are almost regular, replenish
have been present in a higher frequency in the eastern parts of the basin.
Based on the present-day animal migration pattern, it can be suggested that
during wet seasons the availability of abundant food supply could have attracted
the wild animals into the area east of the basin. The wild animals, which do not
go beyond the fringes of the basin, perhaps moved deep into the plains during the
pre–Iron Age period when there was no dense human occupation. Similar sea-
sonal animal migration patterns are common in several undisturbed wildlife pre-
serves (Eisenberg and Lockhart 1972; Ramachandran et al. 1986).
settlement system of the mesolithic hunter-gatherers
The settlement system approach has emerged as an important methodological tool
in archaeological research and it is widely employed to investigate the human
interrelationship with their environment (Bettinger 1980; Binford 1972, 1980;
Chang 1968, 1972; Jochim 1976; Paddayya 1982; Trigger 1978; Wiley 1953).
Various constraints, such as the lack of a chronological resolution at most of the
sites, absence of data for reconstructing the palaeoenvironment, the limited scope
of the present study, and other factors preclude a complete analysis of the settle-
ment system of Mesolithic hunter-gatherers. Hence, a preliminary attempt has
been made to understand some aspects of the Mesolithic settlement system.
The ethnoarchaeological study of contemporary hunter-gatherers furnishes
valuable data for developing methods to better understand the settlement systems
of the ancient hunter-gatherers (Binford 1980; O’Connell 1995; Yellen 1977).
The organizing principles of the hunter-gatherer settlement system is determined
by a combination of several factors, such as resource availability, technological
adaptation, and certain cultural preferences (Binford 1980; Brantingham 1998;
Kelly 1983, 1992; Murdock 1967; Yellen 1977). Taking the organizational aspects
involved into consideration, Binford has categorized two kinds of mobility: logis-
tic and residential, among contemporary hunter-gatherers and has attempted to
correlate these with the archaeological traces they produce. This model can be
adopted to understand the mobility of the Mesolithic groups in the Gundar Basin.
The locational context, size, and artifact density of the microlithic sites suggest
that Mesolithic hunter-gatherers used at least two broad and distinct functional
locales: base camps and temporary camps/speciﬁc activity areas. Several Category
3 site characteristics, such as (a) the abundance of artifacts in general and waste
materials, especially quartz, which requires extra time and energy for knapping,
(b) a higher proportion of debitage in relation to ﬁnished implements, (c) a lim-
ited number of sites, and (d) the location close to resources, indicate a high level
of redundancy in use and a long duration of occupation. The sites of Categories 1
and 2 were relatively less frequently occupied or were occupied for a short dura-
tion because of (a) the small amount of artifacts found at these sites, (b) their
higher frequency, and (c) a relatively low proportion of debitage when compared
to ﬁnished and modiﬁed artifacts. Their location in the areas far from resources
like rockshelters and water facilities and the distribution in most cases around
Category 3 sites also support this claim. Therefore it is quite likely that most of
selvakumar .hunter-gatherer adaptations in tamil nadu 93
the Category 3 sites largely functioned as base camps, forming ‘‘hubs of subsis-
tence activities,’’ while sites in Categories 1 and 2 served as temporary camps
where extractive and other activities would have taken place. Recurrent short-
term occupations can also result in Category 3 sites and hence, some of the Cate-
gory 3 sites, e.g., E. Gopalapuram, served as temporary camps. An important
point to be noted here is that all the sites in Category 3 need not have functioned
as base camps throughout their occupational history. In a particular time period, a
base camp could have been used as a special purpose–temporary camp as well.
These base camps and temporary camps can be equated, respectively, with Bin-
ford’s ‘‘residential base’’ and ‘‘location’’ associated with the residential mobility
pattern (1980 : 9).
Most sites in Categories 1 and 2 are located on the slope of the basin fringe.
The distribution pattern of sites indicates that this area was perhaps mostly ex-
ploited through short-duration camping. The present resource distribution also
indicates that this area could have had abundant plant resources, which would
have been exploited by the Mesolithic people. The sites in Category 3 are located
close to areas that have good water sources. The occurrence of a limited num-
ber of Category 3 sites near the basin fringes could be due to the presence of
dense jungle cover in those areas. Generally, there is not much variation in the
nature of artifacts occurring in the sites in various categories. Choppers are com-
mon in Category 2 sites and there are more ﬁnished implements than those of
Ethnographic information gathered from the Paliyans, who live in the study
area, reveals that they used to exploit the resources from such base camps and shift
their residences whenever the resources were exhausted in a particular area.
Hence, considering this evidence and also the pattern of site distribution, a ‘‘resi-
dential mobility’’ pattern can be suggested for Mesolithic hunter-gatherers of
the upper Gundar Basin. However, this should not be interpreted to mean that
Mesolithic groups had a typical ‘‘residential mobility’’ (i.e., as described by Binford
1980), since these two kinds of mobility are not mutually exclusive (Binford 1980).
The next issue that has to be investigated is the organization of mobility in a
particular annual cycle. As the scope of this study is limited, certain tentative
propositions on the mobility patterns of Mesolithic groups are put forth here,
based on the suggested variation in the distribution of resources and also with the
supporting evidence from ethnographic contexts. In order to understand the mo-
bility pattern of a particular hunter-gatherer group, ﬁrst, the ﬂuctuation in the
availability of resources, and second, the preferred segments among the available
resources have to be established. As elaborated earlier, during the monsoon and
post-monsoon seasons, the resources would have been available in the Gundar
Basin and its surroundings, while in the dry season the resources would have
mainly been conﬁned to the western part of the upper Gundar Basin. The west-
ern part of the upper Gundar Basin with its forested hills, which is home to many
species of dioscorea yams (Dioscorea sp.) above 500 m amsl and several vegetables
like Acacia intisia caesia (mimosa), which are exploited even by the people from
the plains during famines (P. M. Gardner 1972: 413), would have o¤ered boun-
teous food supply in the dry months. The area east of the basin (east of longitude
77500E) would have had various freshwater, animal, and plant resources as well
in the wet seasons.
asian perspectives .41(1) .spring 200294
With regard to the food preference, it is obvious from the faunal evidence that
both big and small animals were hunted (Table 5). Even though no plant remains
have been recovered so far from any of the Mesolithic sites, the diverse species of
edible plants available in the basin (Selvakumar 1997 : 178–182) indicate that plant
food, such as roots, fruits, and leaves, also would have formed a major source of
subsistence for Mesolithic groups. However, the extent of preference of vegetable
food is not clear owing to the lack of evidence.
Taking the variation in resources suggested above and also the limited infor-
mation available on food preference into consideration, it can be proposed that
during the post-monsoon seasons, Mesolithic groups subsisted on the resources
available in the area to the east of the upper Gundar Basin and during the dry
seasons depended on the resources available in the western part of the basin. The
proposed animal migration and the abundance of plant resources and freshwater
fauna during the monsoon and post-monsoon seasons would have enabled their
shortage of such resources in this area, the groups were probably forced to move
into the upper Gundar Basin.
Although it is quite likely that Mesolithic groups would have had a similar
mobility pattern to that proposed above, resolving the issues as to ‘‘aggregation-
dispersal’’ (fusion-ﬁssion), demographic aspects, and home range is problematic
due to the limited nature of the present investigations. The living hunter-
gatherers have di¤erent levels and extents of mobility varying from a sedentary
one exploiting a small territory to a highly mobile one having several hundreds
of square kilometers of home range with di¤ering patterns such as dry-season
aggregation and wet season dispersal and vice-versa (Paddayya 1982 : 89–90).
Hunter-gatherers living in the same or identical environment have di¤erent types
of mobility patterns, depending upon their technological and social organization
and other cultural choices (Turnbull 1968 :135–137; Yellen 1977 :53). In India, a
number of hunter-gatherers, such as the Chenchus (Furer-Haimendorf 1943) and
the Birhors (Williams 1969), have dispersal and congregation patterns conditioned
by economic factors. Until a few decades ago, the Paliyans used to disperse into
small groups in the summer months due to scarcity of food resources (P. M.
Gardner 1969 : 158, 1984 : 414). The mobility pattern observed among these con-
temporary hunter-gatherers, who have been inﬂuenced by their agropastoral
neighbors, cannot be directly applied to the Mesolithic context. To determine the
dispersion-aggregation pattern of Mesolithic hunter-gatherers, further detailed
research including surface surveys of the area to the east of the upper Gundar
Basin (east of longitude 77500) and collecting data on the seasonality of the sites
hunter-gatherers of the iron age–early historic period
Gundar Basin has plenty of evidence for the Iron Age agropastoral settlements (for
details, see Selvakumar 2000). The Iron Age began in the upper Gundar Basin
before the middle of the ﬁrst millennium b.c. (Selvakumar 2000; Soundararajan
1973). There is no evidence for the indigenous development of agropastoral
groups in the Gundar Basin. The ceramic evidence from the Gundar Basin indi-
cates that there was migration of agropastoral groups from the northern part of
selvakumar .hunter-gatherer adaptations in tamil nadu 95
peninsular India. It is possible that these agropastoral groups, who practiced small-
scale farming and cattle and sheep-goat pastoralism and had iron technology most
probably upset the ecological equilibrium. The hunter-gatherers and agropastoral
groups perhaps had to compete with each other for available resources. The
number of Iron Age settlements appearing in the middle of the upper Gundar
Basin, which only had hunter-gatherer settlements during the Mesolithic period,
suggests the possible pressure. Though the pressure on the hunter-gatherers might
not have been high during the initial stages of the Iron Age when the density of
agropastoral settlements was less, in the Early Historic period it might have in-
creased, as the number of agropastoral settlements increased considerably (Selva-
It is not known whether the relationship between the agropastoralists and
hunter-gatherers was one involving conﬂict or cooperation. It is most likely that
with the appearance of the agropastoral groups in the ﬁrst millennium b.c.,
hunter-gatherers were gradually displaced from the core areas of the basin and
driven into the peripheral territories. The reoccupation of the Mesolithic site at
S. Pappinayakkanpatti, which is in the middle of the basin, by the agropastoral
groups, and the evidence of hunter-gatherer occupation at the rockshelters in
Virappatti and E. Gopalapuram, which lie in the marginal areas of the basin and
were largely uninhabited during the Mesolithic period, support this claim. Never-
theless, the evidence, such as beads, pottery, and domestic animal bones from the
rockshelter sites, most probably points towards positive interactions between
hunter-gatherers and agropastoral groups. The presence of querns and grinding
stones in Kanavaypatti Period II and the absence of the same in Period I, can be
taken to denote that hunter-gatherers acquired edible grains from the agropastoral
groups. Some of the wild animal bones and lithic materials occurring in the
agropastoral sites were also, probably, a consequence of such interactions. Similar
interrelationships are mentioned in the early Tamil (Sangam) texts (Singaravelu
1969) and also take place even today among the Paliyans and the village settlers
in this basin (P. M. Gardner 1972) as well as among other groups in di¤erent parts
of India (Misra and Nagar 1997). Besides exploiting the forest and marginal areas
for their own subsistence, hunter-gatherers perhaps also collected forest products,
which were in demand among the agropastoral village communities. In this re-
gard the proposition that a few of the hunter-gatherer groups of India have
played a substantial role in the Early Historic trade and exchange in general and
Indian Ocean trade (Indo-Roman) in particular, by supplying several forest prod-
ucts (Fox 1969; Gupta 1997; Morris 1983; Stiles 1993), deserves mention here.
This study has shown that microlithic sites occur in the interior southern Tamil
Nadu in three di¤erent culture-chronological contexts. Many of the microlithic
sites, especially the surface scatters, seem to be palimpsestic in nature, having both
Mesolithic and Iron Age–Early Historic remains. The 109 microlithic sites dis-
covered in the basin have been divided into three categories on the basis of their
size and artifact density. The characteristics of various categories of sites and the
artifact distribution pattern indicate that Mesolithic groups probably had a resi-
dential mobility pattern. It has been proposed that Mesolithic groups would have
asian perspectives .41(1) .spring 200296
moved into the area east of the basin during the post-monsoon season and
ﬂuctuations in resource availability.
There is convincing evidence that the hunter-gatherers, who continued to ex-
ist in the Iron Age–Early Historic period, interacted with the agropastoral groups.
The relationship between these two groups seems to have been one involving
both conﬂict and cooperation; while some groups interacted with the agropastoral
groups, others appear to have been gradually driven into marginal areas.
Questions related to the proposed mobility pattern of Mesolithic groups, the
nature and dynamics of the interaction between hunter-gatherers and agropastoral
groups during the Iron Age–Early Historic period, the mobility pattern of the
hunter-gatherers during the post-Mesolithic phase, and other related issues have
to be focused on by future investigations.
An earlier version of the paper was presented at the 25th annual conference of the
Indian Society for Prehistoric and Quaternary studies held at Sri Nagar (Garhwal),
U. P., India. I would like to thank the University Grants Commission of India for
ﬁnancial assistance, Prof. K. Paddayya for his support, Dr. P. K. Thomas and his
colleagues for identifying the animal bones, and Dr. Richa Jhaldiyal for her useful
comments on the draft and her help in preparing the maps.
1. Colonies of similar gastropods have been found in rockshelters and caves that have humid atmo-
spheres in the Gundar Basin.
2. In Selvakumar 1996 and 1997, the complexities involved in the formation of the archaeological
record were not taken into consideration, and as a result, all the surface scatters were treated as
belonging to the Mesolithic.
3. Letter (dated 22 August 1995) from Dr. Sheela Kusumgar, Physical Research Laboratory, Ah-
4. Letter (dated 5 March 1997) from Dr. Sheela Kusumgar, Physical Research Laboratory, Ah-
5. The 23 lithic artifacts from the Early Historic levels in the test trench at S. Pappinayakkanpatti
were included in the assemblages of Mesolithic hunter-gatherers in Selvakumar (1997) due to my
oversight. They have been excluded here.
6. Though chronology of the sites in Categories 1 and 2 is uncertain, some of them most likely
have both Mesolithic and Iron Age–Early Historic assemblages.
1944 Mesolithic artifacts from Sawyerpuram in Tinnevelly district, South India. Spolia Zeylanica
24(2) : 145– 154.
1977 Hunters, pastoralists and early agriculturalists in South Asia, in Hunters, Gatherers and First
Farmers Beyond Europe—An Archaeological Survey: 127–144, ed. J.V.S. Megaw. Leicester:
Leicester University Press.
Allchin, B., and F. R. Allchin
1974 The relationship of Neolithic and later settled communities with those of late stone age
hunters and gatherers in Peninsular India, in Indian Society, Historical Beginnings Essays in
Memory of D. D. Kosambi:45–66,ed.R.S.Sharma.NewDelhi:PeoplesPublishing
1983 The Rise of Civilization in India and Pakistan. New Delhi: Select Book Service Syndicate.
selvakumar .hunter-gatherer adaptations in tamil nadu 97
Ammerman, A. J.
1981 Surveys and Archaeological Research. Annual Review of Anthropology 10 : 63–88.
Atherton, J. H.
1983 Ethnoarchaeology in Africa. African Archaeological Review 1 : 75–104.
Bettinger, R. L.
1980 Explanatory/predictive models of hunter-gatherer adaptation, in Advances in Archaeological
Method and Theory 3: 189–255, ed. M. B. Schi¤er. New York: Academic Press.
Binford, L. R.
1972 An Archaeological Perspective. New York: Seminar Press.
1980 Willow smoke and dogs’ tails: Hunter-gatherer settlement systems and archaeological site
formation. American Antiquity 45 : 4– 20.
1990 Lithic reduction sequences as an aid to the analysis of late stone age quartz assemblages
from the Luano Spring Chingola, Zambia. The African Archaeological Review 8 : 103–138.
Brantingham, P. J.
1998 Mobility, competition, and Plio-Pleistocene hominid foraging groups. Journal of Archaeo-
logical Method and Theory 5(1) : 57–98.
Census of India
1969 Paliyans, in Census of India 1961. Ethnographic Notes on Scheduled Tribes, Vol. 9, pt.
5B(i) : 195– 248. (Published in 1969.)
Chang, K. C., ed.
1968 Settlement Archaeology. Pao Alto: National Press Books.
Chang, K. C.
1972 Settlement Patterns in Archaeology. Addison Wesley Module in Anthropology No. 24.
Clark, J. D., and M. R. Kleindienst
1974 The stone age cultural sequence: Terminology, typology and raw material, in Kalambo
Falls, Prehistoric Site, Later Prehistoric Cultures, vol. 2: 71 –105, ed. J. D. Clark. Cambridge:
Cambridge University Press.
1989 Material aspects of Limba, Yalunka and Kuranko ethnicity: Archaeological research in
northeastern Sierra Leone, in Archaeological Approaches to Cultural Identity: 125–138, ed.
Stephen Shennan. London: Unwin Hyman.
Deraniyagala, S. U.
1992 The Prehistory of Sri Lanka: An Ecological Perspective. Colombo: Department of Archaeo-
Dunnell, R. C., and W. S. Dancey
1983 The siteless survey: A regional scale data collection strategy, in Advances in Archaeological
Method and Theory 6: 266–288, ed. M. B. Schi¤er. New York: Academic Press.
Eisenberg, J. F., and M. Lockhart
1972 An Ecological Reconnaissance of Wilpattu National Park, Ceylon. Smithsonian Contribution
to Zoology 101. Washington D.C.: Smithsonian Institution Press.
1981 O¤-site archaeology: An alternative approach for the short sighted, in Patterns of the Past:
Studies in Honour of David Clarke: 157–183, ed. I. Hodder, G. Isaac, and N. Hammond.
Cambridge: Cambridge University Press.
Foote, R. B.
1883 On the geology of Madura and Tinnevelly districts. Memoirs of the Geological Survey of
India 20(2) : 1– 103.
1916 The Foot Collection of Prehistoric and Protohistoric Antiquities: Notes on Their Ages and Distri-
bution. Madras: Government Museum.
Fox, R. G.
1969 Professional primitives: Hunter-gatherers of nuclear South Asia. ManinIndia49(2) : 139–
1943 The Chenchus: Jungle Folk of the Deccan. London: Macmillan.
asian perspectives .41(1) .spring 200298
Gardner, P. M.
1965 Ecology and social structure in refugee populations: The Paliyans of South India.
Unpublished Ph.D. diss. University of Pennsylvania, Philadelphia, and Ann Arbor,
Michigan: University Microﬁlms.
1966 Symmetric respect and memorate knowledge: The structure and ecology of individualis-
tic culture. Southwestern Journal of Anthropology 22 : 389– 415.
1969 Paliyan social structure, in Contributions to Anthropology: Band Societies: 153–171, ed. D.
Damas. Proceedings of the Conference on Band Organization, Ottawa, August 30 to
September 2, 1965. National Museum of Canada, Bulletin 228. Ottawa: The Queens
1972 The Paliyans, in Hunters and Gatherers Today: 404–450, ed. M. G. Bicchieri. Illinois:
1982 Ascribed austerity: A tribal path to purity. Man 17 : 462– 469.
1984 Bicultural oscillation as a long-term adaptation to cultural frontiers: Cases and questions.
Human Ecology 13 : 411– 432.
1988 Pressures for Tamil propriety in Paliyan social organization, in Hunters and Gatherers,
vol. 1: History, Evolution and Social Change: 91–106, ed. T. Ingold, D. Riches, and
1986 Quaternary coastal sediments and stratigraphy, southeast India. Man and Environment
10 : 51– 72.
1995 Red dunes and Quaternary paleoenvironment in India and Sri Lanka, in Quaternary
Environments and Geoarchaeology of India. Essays in Honour of Prof. S. N. Rajaguru: 391–404,
ed. S. Wadia, R. Korisettar, and V. S. Kale. Bangalore: Geological Society of India
Memoir No. 32.
Gardner, R.A.M., and H. Martingell
1990 Microlithic sites and their palaeoenvironmental setting, southeast India: A reevaluation.
Geoarchaeology 5 : 1–13.
Gaussen, H., P. Legris,and M. Viart
1964 International Map of Vegetation and of Environmental Conditions at 1/1,000,000: Notes on the
sheet Cape Comarin. New Delhi: Indian Council of Agricultural Research.
Gupta, H. P.
1990 Sholas in the South Indian montane: Past, present and future. Palaeobotanist 38 : 394–
Gupta, H. P., and K. Prasad
1985 The vegetational development during 30,000 years b.p. at Colgrain, Ottacamund,
Nilgiris, S. India. Journal of Palynology 21 : 174.
1997 Roman Egypt to Peninsular India: Archaeological patterns of trade 1st century
.to 3rd century
. Ph.D. diss., University of Poona.
Hassan, F. A.
1987 Re-forming archaeology: A foreword to natural formation processes and the archaeolog-
ical record, in Natural Formation Processes and the Archaeological Record:1–9,ed.D.T.Nash
and M. D. Petraglia. Oxford: BAR International Series 352.
1976 Know Your District: Madurai. Handbook. Calcutta: Geological Survey of India on the
Occasion of 125th Anniversary Celebrations (1851–1976). Calcutta.
1986 Artifacts made of quartz, in La Cotte De St. Brelade 1961–78. Excavations by C.B.M.
Mcburny: 315–324, ed. P. Callow and J. M. Cornford. Norwich: Geo Books.
1988 The Ahar Culture and Beyond: Settlements and Frontiers of Mesolithic and Early Agricultural
Sites in Rajasthan. Oxford: BAR International Series 412.
Indian Archaeology: A Review
Volumes for 1957–1958, 1958 –1959, and 1960–1961. New Delhi: Archaeological Sur-
vey of India.
selvakumar .hunter-gatherer adaptations in tamil nadu 99
1970 Mesolithic contexts in the Vindyan hills of central India. Ph.D. diss., Columbia Uni-
1975 Static sites and peripatetic peoples: Notes on the archaeology population mobility in
eastern Malwa, in Pastoralists and Nomads in South Asia:68–91,ed.L.S.LeshnikandG.D.
Sontheimer. Wiesbaden: Otto Harrassowitz.
Jochim, M. A.
1976 Hunter-Gatherer Subsistence and Settlement. A Predictive Model.NewYork:AcademicPress.
Kelly, R. L.
1983 Hunter-gatherer mobility strategies. Journal of Anthropological Research 39 : 277 –306.
1992 Mobility/sedentism: Concepts, archaeological measures, and e¤ects. Annual Review of
Anthropology 21 : 43– 66.
Kennedy, K.A.R., J. R. Lukacs, R. F. Pastor, T. L. Jonston, N. C. Lovell, J. N. Pal, B. E.
Hemphill,and C. B. Burrow
1992 Human Skeletal Remains from Mahadaha: A Gangetic Mesolithic Site. South Asia Occasional
Papers and Theses No. 11. Ithaca: Cornell University.
1993 Role of ﬂuorine in the chronometric dating of Indian stone age cultures. Man and Envi-
ronment 18(2) : 23– 32.
Misra, V. N.
1965 Mesolithic phase in India, in Indian Prehistory:577–585,ed.V.N.MisraandM.S.Mate.
Poona: Deccan College Post-Graduate and Research Institute Building Centenary and
Silver Jubilee Series 32.
1973 Bagor: A Mesolithic settlement in northwest India. World Archaeology 5(1): 92 –110.
1976 Evidence of culture contact between terminal stone age hunter-gatherers and contem-
porary farmers, in Archaeological Congress and Seminar 1972: 115–121, ed. U. V. Singh.
Kurukshetra: B. N. Chakravarty University.
1989 Stone age India: An ecological perspective. Man and Environment 14 : 17–64.
Misra, V. N., and Malti Nagar
1997 From tribe to caste: An ethnoarchaeological perspective, in FromTribetoCaste: 136–166,
ed. D. Nathan. Shimla: Indian Institute of Advanced Study.
1993 Mesolithic hunter-gatherers of Keonjhar district, Orissa, India. Asian Perspectives 32(1) :
1967 Ethnographic atlas: A summary. Ethnology 6 : 109– 236.
1989 Pre–Iron Age agricultural settlements in South India: An ecological perspective. Man and
Environment 14(1) : 65– 81.
O’Connell, J. F.
1995 Ethnoarchaeology needs a general theory of behavior. Journal of Archaeological Research
3(3) : 205– 255.
Ota, S. B.
1986 Mesolithic culture of the Phulbani district, Orissa with special reference to the heavy tool
component. Bulletin of the Deccan College Post Graduate and Research Institute 45: 79 –87.
1982 The Acheulian Culture of the Hunsgi Valley, Peninsular India: A Settlement System Perspective.
Poona: Deccan College Post Graduate Research Institute.
Possehl, G. L., comp.
1994 Radiometric Dates for South Asian Archaeology. An Occasional Publication of the South Asia
Section. The University of Pennsylvania Museum, Philadelphia, Pennsylvania.
1994 Archaeology of Tamil Nadu: Kongu Country. New Delhi: Book India Publishing Co.
asian perspectives .41(1) .spring 2002100
Ramachandran, K. K., P. Vijayakumar,and P. S. Easa
1986 Ecology of larger mammals of Periyar Wild Life Sanctuary. Journal of Bombay Natural His-
tory Society 83(3) : 505–524.
Raman, K. V.
1970 Distribution pattern of cultural traits in the pre- and protohistoric times of Madurai re-
gion. Aaraichi 1 : 499– 509.
Roberts, N., and H. E. Wright Jr.
1993 Vegetational, lake-level, and climatic history of the Near East and southwest Asia, in
Global Climates since the Last Glacial Maximum: 194–220, ed. H. E. Wright Jr., J. E. Kutz-
bach, T. Webb III, W. F. Ruddiman, F. A. Street-Perrott, and P. J. Bartlein. Minneapo-
lis: University of Minnesota Press.
Rossingol, J., and.L.Wandsnider,eds.
1992 Space, Time and Archaeological Landscapes.NewYork:PlenumPress.
Sankalia, H. D.
1964 Stone Age Tools: Their Techniques, Names and Probable Functions. Deccan College Building
Centenary and Silver Jubilee Series No. 1. Poona: Deccan College Post Graduate and
Schiffer, M. B., A. P. Sullivan,and T. C. Klinger
1978 The design of archaeological surveys. World Archaeology 10 : 1–28.
1996 Archaeological investigations in the upper Gundar Basin. Man and Environment 22(2) : 27–
1997 Investigations into the prehistoric and protohistoric cultures of the upper Gundar Basin.
Ph.D. diss., University of Poona.
2000 Investigations into the Iron Age–Early Historic cultures of the upper Gundar Basin,
Madurai district, Tamil Nadu, India. South Asian Studies 16 : 119 –132.
2001 Culture and chronology of lithic industries from southern Tamil Nadu, in Kaveri: Studies
in Epigraphy, Archaeology and History—Professor Y. Subbarayalu Felicitation Volume: 256–
266, ed. S. Rajagopal. Chennai: Pannpattu Veliyiittakam.
1989 Introduction: Archaeological approaches to cultural identity, in Archaeological Approaches to
Cultural Identity: 1–32, ed. Stephen Shennan. London: Unwin Hyman.
1969 Social Life of Tamils. The Classical Period. Kuala Lumpur: University of Malaysia.
Singh, G.R.J. Wasson,andO.P.Agrawal
1990 Vegetational and seasonal climatic changes since the last full glacial in the Thar Desert,
northwestern India. Review of Palaeobotany and Palynology 64 : 351–35 8.
1993 Hunter-gatherer trade in wild forest products in the early centuries with the Port of
Broach India. Asian Perspectives 32(2) : 153 –167.
Thompson,Webb III, W. F. Ruddiman, F. A. Street-Perrott, V. Markgraf, J. E. Kutzbach,
P. J. Bartlein,and W. L. Prell
1993 Climatic changes during the past 18,000 years: Regional synthesis, mechanisms and
causes, in Global Climates since the Last Glacial Maximum: 514–535, ed. H. E. Wright Jr.,
J. E. Kutzbach, T. Webb III, W. F. Ruddiman, F. A. Street-Perrott, and P. J. Bartlein.
Minneapolis: University of Minnesota Press.
Thurston, E., and Rangachari, K.
1909 Castes and Tribes of Southern India, vol. 5. Madras: Government Press.
1978 Times and Traditions: Essays in Archaeological Interpretation. New York: Columbia Univer-
Turnbull, C. M.
1968 The importance of ﬂux in two hunting societies, in Man The Hunter: 132–137, ed.
R. B. Lee and I. DeVore. Chicago: Aldine.
Van Campo, E.
1986 Monsoon ﬂuctuations in two 20,000-yr b.p. oxygen-isotope/pollen records o¤ southwest
India. Quaternary Research 26 : 376– 388.
selvakumar .hunter-gatherer adaptations in tamil nadu 101
1985 Madurai Mavatta Paguthikalil Nunkarkala Kandupidipugal. Paper presented during the
seminar conducted by the Department of Archaeology, Tamil Nadu.
Wasson, R. J.
1995 The Asian monsoon during the late Quaternary: A test of orbital forcing and pa-
laeoanalogue forecasting, in Quaternary Environments and Geoarchaeology of India. Essays in
Honour of Prof. S. N. Rajaguru:22–35,ed.S.Wadia,R.Korisettar,andV.S.Kale.Ban-
galore: Geological Society of India Memoir No. 32.
Wasson, R. J., G. I. Smith,and D. P. Agrawal
1984 Late Quaternary sediment, mineral and infrared geochemical history of Didwana Lake,
Thar Desert, India. Palaeogeography, Palaeoclimatology, Paleoecology 46 : 345–372.
Wheeler, R.E.M., A. Ghosh,and Krishna Deva
1946 Arikamedu: An Indo-Roman trading-station on the east coast of India. Ancient India
2 : 17–124.
Wiley, G. R.
1953 Prehistoric Settlement Patter ns in the Viru Valley, Peru. Bureau of American Ethnology,
Bulletin 155. Washington, D.C.: Smithsonian Institution.
Williams, B. J.
1969 The Birhor of Hazaribagh, in Contributions to Anthropology: Band Societies: 142–152, ed. D.
Damas. Proceedings of the Conference on Band Organization, Ottawa, August 30 to
September 2, 1965. National Museum of Canada, Bulletin 228. Ottawa: The Queens
Yellen, J. E.
1977 Archaeological Approaches to the Present. Models for Reconstructing the Past.NewYork:Aca-
Zeuner, F. E., and B. Allchin
1956 The microlithic sites of Tinnevelly district, Madras State. Ancient India 12 : 4–20.
1975 Pastoralism as reﬂected in the classical theory of landscapes, in Pastoralists and Nomads in
South Asia: 30–39, ed. L. S. Leshnik and G. D. Sontheimer. Wiesbaden: Otto Harrasso-
Archaeological investigations undertaken in the upper Gundar Basin (Madurai dis-
trict, Tamil Nadu, India) between 1991 and 1996 have yielded evidence for the
Mesolithic and Iron Age–Early Historic periods. The research has revealed that
microlith-using hunter-gatherers continued to exist during the Iron Age and Early
Historic periods, and interacted with agropastoral groups. This paper discusses
the settlement system of Mesolithic hunter-gatherers and interactions between the
hunter-gatherers and the agropastoral groups of the Iron Age–Early Historic period.
Keywords: Madurai, Tamil Nadu, India, South Asia, microliths, Mesolithic, Iron
Age, Early Historic, hunter-gatherers.
asian perspectives .41(1) .spring 2002102