Signs of Life: Engraved Stone Artefacts from Neolithic South India

Abstract

While exceedingly rare on any given archaeological site, engraved stone artefacts have nonetheless been reported from sites covering a range of periods and regions across the world. Attempts to interpret such engravings have often focused on potential representational or communicative functions, including their role in notational systems, symbolic depiction, and the development of early forms of writing. Contextual and microscopic investigation of a number of engraved artefacts discovered in a large assemblage of dolerite artefacts excavated from a Neolithic hilltop habitation and stone-tool production site in south India suggests, however, that an alternative interpretation of engraved stone artefacts is possible. Drawing on ethnographic evidence concerning the perception of stone, and particularly natural markings on stone, this article argues that the stone pieces on which the marks were engraved were more than just passive surfaces for the creation of unrelated signs. Instead, engravings appear to draw on natural features within and upon the surface of the dolerite, and to suggest an appreciation for the patterns of nature, as well as a lack of distinction between anthropogenic and natural markings. It is argued that the engravings may have been a response to a perceived ‘life-force’ within the dolerite. The fact that they were produced and then broken apart by knapping suggests that they may have been made to accentuate or attenuate a power that was perceived as either somehow beneficial or in need of careful control.

Full text

Page 1

1
Engraved Stone Artefacts from Neolithic South India
Signs of Life: Engraved Stone Artefacts
from Neolithic South India
dolerite flake and non-flake debitage, a biface and a
fractured cobble that were unremarkable in all regards
except that each of them bore a series of engraved
lines or grooves on their remnant cortical surfaces.
The engraved lines appeared in parallel vertical and
horizontal series that on some specimens produced a
sort of grid-like or cross-hatching effect. Comparison
to naturally occurring dolerite cortical surfaces and
examination under high-powered magnification indi-
cated that the lines were artificial, and likely produced
by a sharp-edged stone flake. At first, as only one or
two in an assemblage of thousands of artefacts, the
pieces appeared as mere oddities. However, as more
of them appeared, they began to demand a�ention.
While never more than rare occurrences — only 24
engraved specimens were found in an assemblage
Cambridge Archaeological Journal 16:2, ??–?? © 2006 McDonald Institute for Archaeological Research
doi:10.1017/S09597743060000?? Printed in the United Kingdom.
Adam Brumm, Nicole Boivin & Richard Fullagar
While exceedingly rare on any given archaeological site, engraved stone artefacts have none-
theless been reported from sites covering a range of periods and regions across the world.
A�empts to interpret such engravings have o�en focused on potential representational or
communicative functions, including their role in notational systems, symbolic depiction,
and the development of early forms of writing. Contextual and microscopic investigation
of a number of engraved artefacts discovered in a large assemblage of dolerite artefacts
excavated from a Neolithic hilltop habitation and stone-tool production site in south India
suggests that an alternative interpretation of engraved stone artefacts is possible, however.
Drawing on ethnographic evidence concerning the perception of stone, and particularly
natural markings on stone, the article argues that the stone pieces on which the marks
were engraved were more than just passive surfaces for the creation of unrelated signs.
Instead, engravings appear to draw on natural features within and upon the surface of
the dolerite, and to suggest an appreciation for the pa�erns of nature, as well as a lack of
distinction between anthropogenic and natural markings. It is argued that the engravings
may have been a response to a perceived ‘life-force’ within the dolerite. The fact that they
were produced and then broken apart by knapping suggests that they may have been made
to accentuate or a�enuate a power that was perceived as either somehow beneficial or in
need of careful control.
We are dealing here with worlds where everything
is meaningful, where anything may constitute a
sign, or is liable to say something, on the state of the
relationships between humans, and again between
them and surrounding ‘others’ within a sentient
landscape. The form of a cloud, the song of a bird, the
direction of the wind, the lines on a stone, a tickling
sensation on one’s body, or a dream are capable of
saying something to whoever is willing to decipher
it within a framework of cultural idioms … (Poirier
2003, 121)
In south India, during the technological analysis
of an assemblage of Neolithic stone artefacts from
the Sanganakallu-Kupgal complex of sites in mid-
eastern Karnataka, a number of unusual specimens
were encountered. These artefacts were all apparent
by-products of stone knapping, and consisted of

Page 2

2
Brumm et al.
that exceeded well over 100,000 analyzed artefacts,
and some 500,000 additional sorted artefacts — they
nonetheless formed a clear pa�ern that could not be
ignored. The question was, though, how to interpret
them? They had not been reported from any other
south Indian Neolithic site, and served no obvious
practical purpose.
Given the rarity and enigmatic nature of the en-
graved stone artefacts, a predictable response might
have been to bury them as a brief note in the lithic-
analysis section of the Sanganakallu-Kupgal Project’s
planned site monograph. However, other occasional
archaeological examples of engraved and sometimes
subsequently flaked stone artefacts in South and
Southwest Asia and beyond indicated that we were
dealing with more than an obscure and one-off ar-
chaeological phenomenon. In addition, ethnographic
accounts of the perception and use
of stone in non-industrial societies
(addressed, for example, in Boivin
2004a; Brumm 2004; Taçon 1991), as
well as other observations in the area
from which the debitage derived,
suggested to us that the engravings
might well have significance in terms
of our understanding of stone-axe
production at Sanganakallu-Kupgal
during the Neolithic period. We
therefore decided to present some
of our ideas about what they might
mean, and more generally to high-
light their existence so as to encour-
age documentation and publication
of similar finds, both in south India
and elsewhere.
The context of the engraved
artefacts
Sanganakallu-Kupgal refers to a clus-
ter of granite hills that are straddled
by the villages of Sanganakallu and
Kupgal in the Bellary District of Kar-
nataka, south India (see Fig. 1). The
hills and the immediately surround-
ing plains are home to a significant
concentration of archaeological sites
that have recently been the focus of
detailed archaeological investigation
as part of the Sanganakallu-Kupgal
Project (see Korise�ar et al. mss.).
This work has led to renewed study
of a hill first identified by Robert
Bruce Foote as an axe production site in the late 1800s
(Foote 1887), but largely overlooked by subsequent
researchers. The hill is known locally as Hiregudda
(‘Big Hill’), but is commonly referred to in the ar-
chaeological literature as Kupgal or Peacock Hill (see
also Boivin 2004b; Figs. 2–3). Investigations by the
Sanganakallu-Kupgal Project led to the discovery of
various lithic production-related localities, including
an axe-manufacturing area on a medium-sized plateau
in the southeast part of the large and topographically
complex hill (Korise�ar et al. mss.). This noteworthy
locality has been labelled Area A according to the area
designation system employed by the Sanganakallu-
Kupgal Project. Area A features a particularly heavy
surface sca�er of dolerite bifacial axe manufacturing
debris, including axe blanks and axes. It has been
extensively modified by recent commercial granite
Figure 1. Map of the Sanganakallu-Kupgal study area showing the location
of Hiregudda, or Kupgal, Hill.

Page 3

3
Engraved Stone Artefacts from Neolithic South India
quarrying and sediment extraction
activities, which have destroyed part
of the area and exposed strata bear-
ing thick deposits of similarly flaked
dolerite material.
Four of the engraved artefacts
described in this article come from
several small trenches excavated
with the aim of learning more about
these exposed strata. The 20 remain-
ing specimens, however, come from
a circular stone structure (Feature 1;
see Figs. 4–5) several metres away
that was visible on the surface, and
contained a dense surface sca�er
of flaked dolerite artefacts. The
structure was excavated down to
bedrock, revealing stratified deposits
with a maximum depth of around
60 cm (Fig. 6), and covering a period
from approximately 1750 to 1250
BC. Archaeological remains recov-
ered from Feature 1, and extensive
grinding features associated with it,
suggest that during the final phases
of occupation (around 1400–1250
BC) the structure functioned as a
lithic-production ‘workshop’ for the
manufacture of bifacial edge-ground
stone axes (Brumm et al. mss.). All
stages of bifacial axe manufacture
seem to have occurred within Fea-
ture 1, from the initial reduction of
unmodified stone blocks and the
bifacial thinning and contouring of
large bifacial ‘rough-outs’ transported
from nearby dolerite quarries, to the
final trimming, pecking and grinding
of finished axe blanks. Massive piles
of dolerite waste accumulated in a
lithic dumping area located a short
distance to the south of Feature 1,
where a small trench (Trench 1) was
also excavated (Fig. 7).
The engraved stone specimens
discussed in this article (see Figs. 8–13) come from
three different excavated localities in Area A: 1) strati-
fied deposits within Feature 1; 2) stratified deposits in
the lithic dumping area (Trench 1) adjacent to Feature
1; and 3) deposits to the southwest of Trench 1 (Trench
6). The la�er area consists primarily of artefacts rede-
posited by slope-wash from the lithic dumping area.
Engraved dolerite artefacts were found in both the
early and late chronological phases of occupation in
Feature 1 (see Tables 1 & 2).
Recent radiocarbon dating evidence (Fuller et al.
mss.) suggests that Feature 1 was the focus of domestic
habitation during the earliest phase of occupation of
the structure (c. 1750–1500 BC). Relatively thin lay-
ers of pale brown-grey ashy silt and compact brown
silt lying atop granite bedrock represent the early
Figure 2. Hiregudda, as viewed from the east.
Figure 3. Map of Hiregudda, showing the location of various key
archaeological areas. Area A contains the richest archaeological deposits, and
appears to have been a major focus for the production of edge-ground axes.
Dolerite for production of the axes was procured from local quarries, such as
those found in Areas B and J. Shaded areas denote dolerite trap dykes.

Page 4

4
Brumm et al.
01 m 5 0 cm
Grinding features
1
2
3
4
1
4
2
3
Feature 1
05
metres
Figure 4. Circular stone structure
(Feature 1) in Area A, Hiregudda.
Figure 5. Plan view showing Feature
1. Also depicted are rock surfaces
with axe-grinding grooves (1–4) and
cupule-like grinding hollows (2–3),
petroglyphs (4) and pecked and ground
quartz veins (2, 4; quartz veins are
represented by grey shaded lines)
associated with axe-grinding grooves.

Page 5

5
Engraved Stone Artefacts from Neolithic South India
occupation phase (Fig. 6). Artefact
densities are comparatively low in
these bo�om layers, suggesting the
floor of the dwelling may have been
kept clean of refuse during this oc-
cupation phase. Dolerite stone axes
were manufactured inside Feature
1 during this early phase, but on a
much less intensive scale than in suc-
ceeding layers.
The early habitation phase of
the structure probably terminated
at around 1500 BC, when permanent
Neolithic se�lement at Hiregudda
may have been discontinued. Feature
1 was reoccupied at around 1400 BC,
at which point it became the focus
of intensive stone axe production.
Considerably thicker stratified layers
of rich clayey silt and pale brown silt
admixed with dense lithic deposits
represent the late occupation phase.
Very substantial quantities of dolerite
debitage and other refuse (i.e. pot-
sherds and animal-bone fragments) accumulated in
the floor of the dwelling during this phase, with most
of the lithic debris accruing in the northwest quadrant
towards the rear of the structure. The degree of patina-
Figure 6. Section drawing showing stratified occupation deposits within Feature 1. The northwest deposits lie within
the circular stone structure, those to the southeast lie outside. Inside the structure, the basal stratigraphic contexts 3029
and 3030 comprised a dark brown clayey silt with a relatively small amount of lithics, and a dark brown compact grussy
silt respectively. These layers correspond to the early occupation phase of the structure (c. 1750 to 1500 BC). Contexts
3026, 3027, 3028, 3031, 3058 and 3065 comprised of rich clayey silts and pale brown silts. Dolerite axe manufacturing
debitage was extremely dense in these upper layers, which correspond to the late phase of the structure (c. 1400–1250
BC). Context 3032 consisted of a hollow fill containing very dense dolerite debitage.
Table 1. Trenches/features and particular
stratigraphic contexts in Area A in which engraved
artefacts have been discovered to date.
Trench/FeatureContexts in which
engraved artefacts
were found
Feature 13034, 3035, 3038, 3058, 3059,
3060, 3061, 3151, 3153, 3167
Trench 1
3015, 3016
Trench 6
3073, 3078
Table 2. Radiocarbon ages (from wood charcoal) and preliminary chronological phasing for
Feature 1 contexts in which engraved artefacts have been discovered. Radiocarbon dates are
a�er Fuller et al. (mss.).
Feature 1 contexts
in which engraved
artefacts were found
Contexts with
radiocarbon
ages
Chronology Phase
3034, 3035, 3038,
3058, 3059, 3060, 3061
3034 (3042±30)c. 1400–1250 ��late phase
(Neolithic to
Megalithic transition)
3151, 3153, 3167 3151 (3314±30)c. 1750–1500 ��early phase
(Late Neolithic)
Figure 7. Thickly stratified lithic deposits exposed by excavations in Area A,
Hiregudda.
tion on most lithics and the weathered state of some
recovered bone fragments suggests that artefacts prob-
ably lay exposed on the ground surface for extended
periods during the late occupation phase. This could

Page 6

6
Brumm et al.
Figure 9. Dolerite biface thinning flake engraved with
grid-like pa�ern. Scale bar is in 10-mm increments.
(Photo M. Moore.)
Figure 10. Engraved dolerite flake. Scale bar is in 10-mm
increments. (Photo M. Moore.)
Figure 11. Bifacially reduced tabular dolerite cobble,
marked with a carefully pecked grid-like pa�ern on
one side. Scale bar is in 10-mm increments. (Photo M.
Moore.)
Figure 12. Fractured dolerite cobble engraved with
complex grid-like pa�erns on three adjoining sides. Scale
bar is in 10-mm increments. (Photo M. Moore.)
Figure 8. Engraved dolerite weathering spall. Scale bar
is in 10-mm increments. (Photo: M. Moore.)

Page 7

7
Engraved Stone Artefacts from Neolithic South India
Figure 13. A–F: Various engraved dolerite debitage pieces recovered from excavations at Area A. Scale bar is in 10-mm
increments.

Page 8

8
Brumm et al.
imply periodic abandonment of the structure.
Engraved dolerite artefacts are found both in the
earliest phase of domestic habitation in Feature 1 and
during the later ‘industrial’ phase (see Table 2). They
are, however, mostly found in contexts associated with
the late axe-production phase. Based on recent chrono-
metric evidence and stratigraphical correlation with
nearby sites, this late phase of occupation in Feature 1
and adjacent areas in Area A on Hiregudda has been
a�ributed to the Neolithic–Megalithic Mesolithic??
transition (Fuller et al. mss.).
Further analysis of the extremely large as-
semblage from Area A will likely yield additional
discoveries of engraved artefacts, as may analysis of
lithic assemblages from other areas excavated by the
Sanganakallu-Kupgal Project (reports of two surface
finds of similar engraved dolerite artefacts in Area A
and within the vicinity of Hiregudda have already
come to light since work began on this manuscript). So
far our work has sampled only a very small proportion
of visible archaeological localities. What is clear from
analyses carried out to date, however, is that these
artefacts concentrate within Area A, and particularly
within Feature 1. They also cover a significant time
period (at least several hundred years), and are not
the product of a limited number of events. This is
indicated not just by their persistence through mul-
tiple phases of activity at the site, but also by the fact
that none of the engraved artefacts discovered so far
could be refi�ed, or appear to derive from common
parent cores.
The engraved dolerite artefacts

Preliminary microscopic analysis conducted on the
engraved artefacts confirmed our initial interpretation
(based on examination with a hand-lens) that all of the
pieces under question had been purposely engraved.
Each artefact was examined under a stereo-zoom mi-
croscope (Zeiss Stemi) with an oblique external light
source, at magnifications up to ×60. Surfaces were
further examined under a metallographic microscope
(Ziess Axio 100 and Olympus BX60) with vertical
incident illumination (brightfield and darkfield) and
polarizing filters, at various magnifications: ×50, ×100,
×200, ×500 and ×1000.
Under magnification, parallel alignments and
overlaps of smoothing, linear striations and in some
cases even gouges can be clearly observed in many
of the grooves. These striations are consistent with
the production of the grooves by a tool-edge (see
Alvarez et al. 2001; Fritz 1999), probably sharp-edged
flakes of dolerite used in repeated cu�ing strokes. The
engraved grooves featured both angular to sharp ‘V’-
and concave ‘U’-shaped cross-sections that may have
been produced by tool-edges with different degrees
of wear and slightly different shapes (Figs. 14 & 15).
The absence of smoothing and linear striations inside
some of the grooves helped us to distinguish between
anthropogenic markings in the dolerite and other
groove-like features formed by cracking and natural
weathering processes (see below). In the case of one
artefact, a bifacially reduced tabular cobble (see Fig.
11), however, impact marks and the absence of incised
grooves with linear striations suggested that the mark-
ings were made using a different technique, probably
Figure 14. Photomicrograph of engraved lines on the
fractured dolerite cobble (see Fig. 12). Width of field:
20 mm.
Figure 15. Photomicrograph of engraved lines on the
dolerite weathering spall (see Fig. 8). Width of field:
20 mm.

Page 9

9
Engraved Stone Artefacts from Neolithic South India
involving careful pecking.
Importantly, microscopic analysis of the en-
graved objects confirmed that the so� cortical surfaces
of all of the incomplete pieces had been engraved prior
to their reduction by knapping. In the laboratory, the
criteria used to assess whether or not the engraved
pieces were marked prior to their reduction was simi-
lar to those used in the analysis of flake scar overlap
on flaked implements. For example, high-powered
magnification revealed that flake scars and/or flake
initiations or terminations had cleanly truncated many
of the engraved grooves. This observation was consist-
ent with engraving before reduction. It was also clear
that where an engraved groove extended to the very
edge of a flake or fragment, the engraving marks did
not continue around onto the ventral surface of the
artefact. This is what one would expect if the pieces
of debitage had been engraved a�er, rather than prior
to, reduction.
Table 3 lists the range of technological catego-
ries represented by the 24 engraved stone artefacts.
While one of the artefacts consists of a weathering
spall produced through natural processes, the rest of
the pieces bear unambiguous evidence of production
by direct freehand percussion. Despite the relative
coarseness and patination of the dolerite material, a
range of diagnostic features of percussive stone flak-
ing, such as striking platforms, dorsal and ventral
surfaces, bulbs of force, dorsal scarring, and ‘lipped’
platforms resulting from bending-initiated fractures
(see Andrefsky 1998; Co�erell & Kamminga 1987) are
readily identifiable on most of the debitage pieces.
Significantly, one of the flakes consists of a biface
thinning flake produced when knapping a biface
(Fig. 9), another a redirecting flake produced when
the platform surface of a core becomes re-oriented by
a percussion blow (Fig. 13a), and another a probable
‘contact removal flake’ (a flake removed from a flake
blank core, preserving the former point of impact on
the flake). These three artefacts, and the degree of tech-
nological variability represented by the others, sug-
gest that the engraved pieces were produced during
complex core reduction strategies, especially bifacial
axe manufacture (see Brumm et al. mss.).
The engravings were all non-figurative in nature.
They consisted mostly of carefully executed sets of
parallel and sub-parallel horizontal and vertical (in
some cases diagonal) lines, sometimes converging to
form grid- or la�ice-like pa�erns. Very li�le morpho-
logical variability can be observed in the imagery. It
seems that more or less the same pa�ern was depicted
on all of the stones. While at least two of the engraved
objects seem to have been used as tools for grinding
other stones or hard surfaces (see below), none of the
markings appear to have any conceivable utilitarian
function. In their microscopic features and morphol-
ogy, the grooves and lines are consistent with the types
of markings produced during rock-art engraving (see
Alvarez et al. 2001).
Other archaeological reports of engraved stones
As we have already suggested, the engraved dolerite
artefacts from Hiregudda are not entirely unique.
Indeed, part of the reason the artefacts deserve fur-
ther consideration is that they are representative of
a wider phenomenon. While we are unaware of any
other reports of engraved stones in south India, such
finds have been made in the northern part of the South
Asian subcontinent, in Southwest Asia, and beyond.
Though never very common in most areas, they date
to a wide range of periods, from the Palaeolithic
through to the historic era, and occur on both modified
and unmodified stones Aren’t all stones they occur
on modified by definition?- ed.. Archaeological exam-
ples of engraved stones from a range of different time
periods and geographical locations in Southwest Asia
are summarized in Table 4. We limit the following dis-
cussion to those examples that are of direct relevance
to the south Indian study area and the nature of the
engraved stone assemblage. Examples discussed in
the paper are not summarised in Table 4.
Only a few examples of engraved stone artefacts
have previously been reported for South Asia. Kenoy-
er notes the recovery of an oval pebble engraved with
markings (interpreted to represent eyes, nose and a
mouth) from the Upper Palaeolithic cave site of Gar-I-
Asp in northern Afghanistan (Kenoyer 1993, 239). This
object has been dated to between 20,000 and 15,000 BC.
Sonawane has also noted the discovery of a Mesolithic
Table 3. Technological categories represented by the engraved dolerite
artefacts from Hiregudda.
TypeNumber
Flakes9
Flake sha�er6
Other non-flake debitage2
Multiple platform core1
Other core1
Biface1
Split cobble 1
Weathering spall1
Other2
Total24

Page 10

10
Brumm et al.

Page 11

11
Engraved Stone Artefacts from Neolithic South India
engraved chert microblade core in northwest India,
near the town of Chandravati in Rajasthan (Sona-
wane 1987; see also Kenoyer 1993, 241; Fig. 16E). The
engraved pa�ern on the core consists of a complex ar-
rangement of cross-hatched geometric forms. A�er the
pa�ern was engraved on the core, a few microblades
were removed from it, destroying part of the imagery
(Sonawane 1987, 54). Meanwhile, in Baluchistan, a
stone ‘spatula’ incised with non-figurative geometric
imagery has been reported from the earliest se�lement
layer at Mehrgarh (Jarrige & Lechevallier 1979, 470). In
addition, at the Neolithic site of Burzahom in northern
India, excavations revealed two flat plaque-like stone
slabs, each of which featured engraved figurative and
non-figurative imagery (Sant 1991, 163–4, pls. 35 &
36). Several small flat rectangular stone ‘harvesters’
with perforated holes and incised geometric motifs
were also discovered at the Neolithic site of Gu�ral,
in Kashmir (Sant 1991, pl. 38). According to Sankalia,
these objects were probably neck ornaments or
pendants (Sankalia 1974, 303).
Important examples of engraved stone artefacts
from the Near East archaeological record include some
incised limestone and basalt pebbles from the late
Neolithic Yarmukian culture in the Southern Levant
(Gopher & Orrelle 1996; Fig. 16G; see also Eirikh-Rose
2004 and Stewart & Rupp 2004 for comprehensive
lists of engraved stones from throughout prehistoric
Southwest Asia and the Mediterranean). The Yarmu-
kian incised pebbles, which date to around the mid-
sixth millennium bp, assume a variety of forms. Most
consist of elongated or sub-spherical natural pebbles
marked with a fairly consistent range of imagery.
Incised motifs include horizontal slits, deep linear
engravings, parallel linear markings, and grid or net
pa�erns. More complex arrangements of incisions
supposedly representing anthropomorphic features
are also found on some stones. Interestingly, several
of the 39 incised pebbles illustrated by Gopher & Or-
relle appear to have been worked as cores; flake scars
with radiating compression rings are clearly evident
in the drawings (e.g. Gopher & Orrelle’s (1996), figs.
1:1, 1:4, 1:5, fig. 3:3 & fig. 5:1). These authors make no
mention of this, however, and it is difficult to deter-
mine from the illustrated specimens alone whether
these pebbles were incised before or a�er they were
possibly flaked.
Engraved lithics have also been reported from
more recent archaeological sites in the Levantine
region. Rosen provides a brief overview of incised
stone tools found at several different sites in the post-
Neolithic Levant (Rosen 1997). These objects consist
mostly of ‘tabular scrapers’ with a repetitive corpus
of incised imagery (see Fig. 16h). Incised motifs on
these artefacts are dominated by sets of horizontal,
vertical and diagonal lines, sometimes converging to
form grids or other complex abstract pa�erns. As with
the Yarmukian pebbles, some illustrated examples
of tabular scrapers seem to indicate that secondary
retouch removed portions of the engraved imagery.
Rosen, however, suggests that the images were incised
on the stone tools a�er they were manufactured at
quarry sites (Rosen 1997, 75).
Various engraved stone examples have been re-
ported from sites beyond South and Southwest Asia,
and span a wide range of time periods. In particular,
many small ‘portable’ stones, rocks and pebbles en-
graved with figurative and non-figurative imagery
have been documented at Upper Palaeolithic sites in
Europe, as have tens of thousands of engraved stone
plaque�es (Bahn & Vertut 1997, 89–92). At some Up-
per Palaeolithic sites in France (e.g. Labastide and
Enléne), a few of these plaque�es bear evidence of
percussion, suggesting intentional breakage (Bahn &
Vertut 1997, 90).
Moving beyond the Palaeolithic, a number of
‘carved stone balls’, some featuring incised decora-
tions, have been reported from the Aberdeenshire
region of Neolithic Scotland (MacGregor 1999).
Also, at the Graig Lwyd axe quarry in North Wales,
a roughly oval stone plaque engraved with finely
‘scratched’ geometric motifs was located (Piggo� 1954,
290). Outside prehistoric Europe, some of the earliest
examples of engraved stones are the 77,000-year-old
pieces of ochre engraved with geometric pa�erns from
Blombos Cave in Southern Africa (Henshilwood et al.
2002). Other prominent examples include the ovate
engraved pebbles discovered in the rockshelter site
of Kamikuroiwa in Japan (Aikens & Higuchi 1982,
106–7). The la�er stones were obtained from deposits
radiocarbon dated to 12,165 bp, and feature arrange-
ments of parallel, crosshatch and curvilinear lines. In
North America, at least 30 plaque-like stones engraved
Figure 16. (on le�) Engraved stones and stone tools from
various archaeological contexts: A–D) engraved pressure-
flaked obsidian projectile points from Can Hasan III,
Neolithic Turkey (a�er Ataman 1988); E) engraved chert
microblade core from Chandravarti in northern India,
probably Mesolithic (a�er Sonawane 1987); F) engraved
slate arrow-point from a Neolithic site in Scandinavia
(a�er Goldhahn 2002); G) engraved pebble from the
Yarmukian period, Levant (a�er Gopher & Orrelle
1996); H) engraved ‘tabular scraper’ from post-Neolithic
Levantine region (a�er Rosen 1997). Drawings are not to
scale.

Page 12

12
Brumm et al.
with complex geometric pa�erns were recovered from
excavations at the 10,900- to 11,200-year-old Palaeoin-
dian Clovis site of Gault in central Texas (Collins &
Hester 2001). Of particular interest was a Clovis point
found ‘sandwiched’ between two engraved stone
cobbles at Site 41BL323 in southernmost Bell County,
central Texas (Collins et al. 1991, 13).
As the preceding discussion suggests, only a
small portion of reported engraved stones consist of
stone tools or knapping by-products. These include
the various examples mentioned above, and also
perhaps one of the earliest recorded specimens: a
stone pressure flaking tool engraved with geometric
imagery from an Upper Palaeolithic site in northern
Italy (Marshack 1972, 454). Another rare example of
an engraved stone tool from Palaeolithic contexts
‘Palaeo-Indian’ is not normally referred to as ‘Pal-
aeolithic’ – perhaps ‘Pleistocene’? – ed.. consists of a
chert flake engraved with a geometric pa�ern on the
exterior cortex, found at the Palaeoindian Wilson-
Leonard Site (41WM235) in central Texas (Collins et
al. 1991, 15). A so� limestone hammerstone incised
with parallel lines was also recovered from Natufian
deposits in Hayonim Cave in the Levant (Belfer-Co-
hen 1991, 576). In general, however, most examples
of engraved stone artefacts from around the world
seem to date to the Neolithic ‘Holocene’ is be�er as
‘Neolithic’ is not used in much of the world – ed.
period. For instance, incised stones used as bark cloth
beaters have been recorded at several Neolithic sites
in the Philippines and Indonesia (Sieveking 1956;
see also Tolstoy 1991 for Mesoamerican examples).
Incised stone tools have also been recovered from
the late Neolithic site of Skara Brae in the Orkneys,
with one notable example consisting of an elaborately
engraved mace head (Malone 2001, 246). At several
Neolithic sites in Scandinavia (such as Nämforsen), a
number of red slate daggers and other tools engraved
with ‘zig-zag’ and ‘diamond’-like pa�erns have been
recorded (Goldhahn 2002, 54-5; Fig. 16f). Ataman has
also described a small assemblage of pressure flaked
obsidian projectile points incised with figurative and
non-figurative imagery, found at the Neolithic site
of Can Hasan III, Turkey (Ataman 1986; 1988; Figs.
16a-d). The flat ventral portions of these points had
been incised with non-figurative geometric pa�erns,
probably with a sharp implement of chert or flint.
Interestingly, microscopic analysis and experimental
replication indicated that the incisions were made
before the points were finished (Ataman 1986, 340).
Engraved stone tools have occasionally been
recorded among hunter-forager societies of the
Holocene period. For example, in mid to late Holocene
contexts in eastern Australia, a stone axe with en-
graved depictions of a fish and a boomerang has
been reported (Bramell 1941, 18; McCarthy 1976, 72).
McCarthy also made note of specialised use-polished
tools made from slate, found throughout the eastern
part of South Australia (McCarthy 1976, 40). A number
of these implements bear engraved animal tracks, sets
of parallel lines, grid and linear pa�erns. The enig-
matic ‘morah’ grindstones from the North Queensland
rainforest consist of large flat slabs of hornfels slate
engraved across the upper surface with deeply incised
grooves, some arranged into pa�erns around the rim.
McCarthy refers to them as a specialized milling stone
restricted to northeastern Queensland (except for a
pecked example from the Simpson Desert) (McCarthy
1976, 56, 64). The slabs of slate are quarried and most
are smaller than about one foot (30 cm) long, with
a series of parallel transverse (sometimes longitu-
dinal) incisions across a grinding depression. It has
been suggested they were used to process poisonous
seeds with the grooves serving to drain away toxic
secretions (which seems impracticable to McCarthy
since grooves run into the middle). ‘Morah’ stones
appear to be distinct in design and function from a
range of other incised stones referred to by McCarthy
(1976) including engraved stones (pp. 62, 68) message
stones (p. 74), ‘cylcons’ (pp. 62, 75), phallic stones (p.
77), bone-shaping stones (p. 68), and tjurunga sacred
stones (pp. 65, 76).
In Britain, stone incisions relating to stone-tool
production include the markings made on the walls of
Neolithic flint mines in Sussex (Thomas 1999; Russell
2000). While not on loose stones, such finds deserve
mention here due to their Neolithic date, association
with stone tool production and, in particular, close
resemblance to the imagery on the dolerite artefacts
at Hiregudda. In addition, the marks on the walls of
mines at sites like Cissbury and Harrow Hill high-
light the possibility that the incisions found on some
engraved stones were made prior to their extraction
from quarries.
Archaeological interpretations of engraved stones
Some of the engraved stone examples mentioned
above carry imagery that is clearly representational,
and hence allows the artefacts upon which it is found
to fit fairly comfortably within the category of ‘port-
able’ or ‘mobiliary’ art (see Abadía & Morales 2004).
However, many are marked by incisions that are far
more enigmatic. These include a range of marks and
lines, and parallel or crosshatched pa�erns like those
found at Hiregudda are a not uncommon feature.

Page 13

13
Engraved Stone Artefacts from Neolithic South India
While a�empts to a�ribute such designs to various
pragmatic activities, such as tool sharpening and ar-
row/spear sha� straightening (e.g. Solecki & Solecki
1970), laundry scrubbing, vegetable grating, animal
branding or childrens’ play (e.g. Stewart & Rupp 2004)
are occasionally made, they are o�en unconvincing
(see Eirikh-Rose 2004). Controversy surrounds many
such interpretations. A common alternative has been
to suggest that the marks served as symbols that allow
abstract cultural ideas to be expressed and commu-
nicated. The interpretation of the engravings as nota-
tional systems provides a clear example of this kind
of reasoning. From such a perspective, the marks are
seen as possible coded entries that allowed the record-
ing or computation of abstract numerical entities. The
interpretation of the ‘chessboard’ or ‘ladder-pa�ern’
incisions at the Neolithic mines in Sussex, England as
indicative of a primitive type of tally-system (Russell
2000) provides a salient example of this kind of read-
ing, as does the suggestion that the Urkan pebble (see
Table 4) may have functioned as a device for recording
cyclic and seasonal notations (Hovers 1990, 321; see
also Belfer-Cohen 1991, 579; Marshack 1972). Consider
also Rosenberg’s thoughts on the ‘notched batons’
from Hallan Çemi Tepesi (see Table 4):
To the naked eye there is no evidence of wear within
the notches or elsewhere on these objects; the notches
were simply cut into the stone, as if to keep a formal
count of something. If so, whatever was being tallied,
it was arguably socially, economically, or politically
important enough to record permanently on a highly
uniform (i.e. formal) class of objects (Rosenberg
1999, 28).
Akin with such interpretations are those that link the
marks to early forms of writing or signification. With
respect to the incised tabular scrapers from the Levant,
for example, Rosen has argued that while their func-
tion and meaning remain enigmatic, it is nonetheless
‘…tempting to correlate these incised symbols with
early writing’ (Rosen 1997, 75). Others similarly argue
for a communicative function for the markings on
engraved stones. Garfinkel, for example, observed
that the Yarmukian incised pebbles bear marks that
are similar to the geometric pa�erns that are found
on contemporary stamp seals in the northern Levant
and Mesopotamia, as well as on pintaderas (small
stone and clay objects with carved relief designs) from
the Neolithic period in Byblos (Garfinkel 1993, 125),
thus implying that they may have served a similar
communicative function (see also Eirikh-Rose 2004
and Stewart & Rupp 2004). Cauvin argues that the
Neolithic marks on stone plaques in the Levant (see
Table 4) form part of a ‘universe of ‘signs’’ that appear
at the same time on various forms of material culture
(Cauvin 2000, 48).
Perhaps one of the most common arguments
made about stones engraved with non-figurative im-
agery is that they represent anthropomorphs, animals
or other beings or things, a somewhat paradoxical
situation given the apparently non-representational
nature of much of the imagery. For example, Kenoyer
suggested that the elongated oval pebble from Upper
Palaeolithic Gar-I-Asp was engraved with markings
representing eyes, a nose and mouth, forming an an-
thropomorphic figure (Kenoyer 1993, 239). Similarly,
Gopher and Orrelle have interpreted a number of
the Yarmukian incised pebbles to be figurines repre-
senting the female body at different stages of sexual
development (Gopher & Orrelle 1996). According to
this model, pairs of short slit marks at one end of elon-
gated pebbles are interpreted to represent ‘…young
girls, perhaps at menarcheal rites, whose body lines
are completely concealed, only the eyes being visible’
(Gopher & Orrelle 1996, 267). Single vertical slit-lines
on figurine pebbles, they further suggest, represent
the genitals of young girls, whereas parallel vertical or
horizontal lines converging to form grid-like pa�erns
are interpreted as the labia of older women, parturi-
tion scars, or scarring from genital mutilation (Gopher
& Orrelle 1996, 267). In addition, the incised pebbles
from Kamikuroiwa in Japan have been interpreted to
represent bare-breasted women with long hair and
skirts made of hanging cords (Aikens & Higuchi 1982,
107). Hermansen also interprets a piece of sandstone
with rough linear engravings at one end (from the
Neolithic site of Basta in Syria-Palestine) as an anthro-
pomorphic figurine (Hermansen 1997, 334). However,
given the non-figurative nature of the markings, he
concludes: ‘…this piece was apparently discarded in
unfinished condition’ (Hermansen 1997, 334).
Ethnographic perspectives on stone
Archaeology, therefore, has tended to interpret the
marks on engraved stones as a�empts to represent
ideas, numbers, words, things or people on the surfac-
es of the stones. Accordingly, li�le a�ention has been
paid to the stones themselves within such accounts.
It is as though people just needed a handy surface for
marking, and stone simply fulfilled that purpose. And
yet engraving stone surfaces is o�en no easy ma�er,
and, compared to marking so�er materials like wood,
bone, antler, shell and clay, could hardly have been
the most convenient way to temporarily store data,
record information or communicate messages. So
why stone? While archaeology is somewhat reticent
on this question, ethnography, in contrast, is full of

Page 14

14
Brumm et al.
insights on stones in general and why people mark
and modify them.
Stone is in many cultural contexts considered to
be not a passive entity or blank canvas, but instead a
very meaningful, indeed o�en animated, substance
(Boivin 2004a; Brumm 2004; Taçon 1991; see also Boast
1997, Kopytoff 1986 and Knappe� 2002 for discussion
of the agency of material objects in some cultural
contexts). This is especially true for many small-scale
foraging and farming societies in India, Southeast
Asia and the Western Pacific and Australasian re-
gions, most notably central and northern Australia,
Indonesia, New Guinea and Melanesia. For these
indigenous people — many of whom had recent, li�le
or no knowledge of metals at the time of ethnographic
documentation — stones are o�en considered to be
sentient, volitional beings who interact freely with
the world of humans.
Table 4. Examples of engraved stones from a range of archaeological sites and time periods in Southwest Asia.
Site/LocationAge Description
Qafzeh Cave (Israel) 100,000 bpBroken Levallois core incised with parallel and overlapping non-
figurative lines on the outer cortical surface. Incisions were made
a�er the core was flaked.
Quneitra (Israel)50,000 bpSupposed flint plaque engraved with non-figurative imagery.
Hayonim Cave (Israel)‘Aurignacian’Two engraved limestone slabs, one incised with lines interpreted to
resemble an ungulate.
Piece of basalt with a deep central groove and several parallel
incisions.
Urkan e-Rub II (Lower Jordan Valley) 19,000–14,400 bpLimestone pebble incised on both sides with complex repeated
geometric pa�erns.
Karain and Öküzini caves (southern
Anatolia)
17,000–12,000 bpSeveral incised stone pebbles and small tablets at both sites.
From Karain, a sub-spherical pebble featuring natural markings
that had been ‘completed by thinly carved, man made incisions as
well as dots which seem to have an intentional order’ (Anati 1968,
25).
Central Béqaa Valley (Lebanon)
12,500–10,200 bpSeveral limestone pebbles engraved with sets of parallel lines.
Wadi Hammeh 27 (Jordan)
12,500–10,200 bpSeveral limestone pebbles and fragments incised with a range of
non-figurative pa�erns.
Limestone pebbles with central grooves and adjacent geometric
incisions.
Nahal Oren (Israel)12,500–10,200 bpBasalt ‘sha�-straightener’ with a complex grid-like design incised
on one side.
Salibiya I (Lower Jordan Valley)
12,500–10,200 bp Fragment of an incised limestone plaque�e.
Hayonim Cave (Israel)
12,500–10,200 bpPebble criss-crossed with net-like incisions.
Ochre-stained limestone fragment with net-like incisions.
Flat limestone fragment with an incised pa�ern of horizontal lines.
Hallan Çemi Tepesi (eastern Anatolia) eleventh millennium ��
Several pendulous objects made from so� metamorphic rock and
featuring arrangements of variably spaced notch marks (and in
one example, an incised ‘hourglass design’); known as ‘notched
batons’.
Cafer Höyük (eastern Anatolia)8500 bp Piece of green stone with a central encircling groove and incised
geometric pa�ern.
Jerf el Ahmar, Çayönü, Mureybet,
Sheikh Hassan, Cafer (Levant)
‘Neolithic’Grooved stones with engraved geometric and linear motifs; ‘sha�-
straighteners’.
Aşıklı Höyük mound (Anatolia)Eighth millennium �� Small polished stone plaque incised with ‘V’ and ‘O’ pa�erns.
Jerf el Ahmar (Syria)‘Neolithic’ Double-sided stone plaques featuring incised figurative images
interpreted by Cauvin to represent a range of animals, including
a small owl, large insect, serpent-shape, straight lines or snake-
shapes, and a grid motif with a snake.

Page 15

15
Engraved Stone Artefacts from Neolithic South India
A review of the ethnographic literature suggests
that many indigenous societies hold the belief that de-
ceased relatives, spirit entities and other supernatural
potencies dwell within particular stones, or indeed,
are embodied as the stones themselves. Belief in the
‘life-force’ (a�er Taçon 2004) of stones commonly in-
volves large immobile rocks, such as standing stones
and other types of megaliths. For example, among
Naga Hills villagers in recent northern India, it was
believed that the souls of the dead became infused
into stone menhirs, which in turn made the land fertile
(Hu�on 1927). Likewise, in the jungles of Pengkalan
Kempas in eastern Malaysia, stone megaliths were
associated with certain noted ancestors, and referred
to by local people as batu hidup or literally, ‘living
stones’ (Chandran 1973, 97). And as King noted of the
‘Maloh’ of Indonesian Borneo (Kalimantan), stones are
perceived to be both the potential receptacles of spirits
and material manifestations of the human soul (King
1975, 108; see also Linehan 1940 and Sukendar 1985 for
further ethnographic insights from Indonesia).
In addition to large immobile stones rooted to
permanent locations in the landscape, smaller, more
portable stones are or were o�en considered by many
indigenous peoples to be living entities. Many such
cultures describe portable stones moving around of
their own accord. For example, in the Tangma area of
Irian Jaya (West Papua, Indonesia), green schist axes
are believed to fly through the air at night, and also
travel underground through subterranean passage-
ways (Pétrequin & Pétrequin 1993, 375). Dani people
wishing to obtain these stones had to first catch them
in a special ceremony requiring the use of pork fat as
bait. In other Melanesian societies, ethnographers have
recorded examples of stones believed to walk around,
dance, light fires, transmit and cure disease, speak,
procreate and kill (Kahn 1990; Roe & Taki 1999).
Nagas from northern India kept small black
oval stones in their households that were believed to
be possessed by spirits (Hu�on 1926, 79). The stones
magically protected the rice crops and were even
reputed to do ba�le with the mice and rats that came
to eat the rice, bearing ‘scars’ from these fights in the
form of small incisions resembling rodent teeth-marks.
Indeed, Hu�on wrote that, to the Nagas ‘…[a]ny
stone…that is at all out of the ordinary is liable to be
regarded as the abode of a deity…it seems possible to
treat any smooth stone complete in itself as the abode
of a spirit’ (Hu�on 1926, 79).
Some small stones may require food and sus-
tenance like any living being. Religious specialists
or balien of the Taman of Indonesian Borneo, for
instance, ‘feed’ their spirit-stones rice, glutinous rice,
palm wine, and rice wine, with the stones in return
ensuring bountiful rice harvests and protecting their
owners from danger (Bernstein 1997; see also King
1975). Similarly, the Nagas of northern India are said
to have kept their rodent-fighting stones carefully
hidden in specially woven ra�an baskets, occasionally
taking them out to be rubbed with pig fat (Hu�on
1926). Hampton noted similar practices among the
Langda of Irian Jaya, who rubbed cores selected for
quarrying with sacralized pig fat, in order to make the
stones beautiful and please the spirits inside (Hamp-
ton 1999, 257).
On this note, it is important to point out that it
is not only megaliths and small natural pebbles that
become imbued with such deeper symbolic meanings.
Many indigenous people also regard the stones used
to make tools as alive. As Pétrequin and Pétrequin
recalled of the green schist axes used in the above-
mentioned Tangma area in Irian Jaya:
…the green schists themselves are treated with re-
spect. The slabs are gathered in the stream-beds and
are ground to give an even shape. If a rough-hewn
blade breaks, it is placed vertically along the walls
of the rocky overhang of Biganme. Everything takes
place as if a certain force pre-existed in the rock be-
fore the consecration of the finished object (Pétrequin
& Pétrequin 1993, 375).
A similar perspective is found amongst the Langda
stoneworkers of Irian Jaya, where Stout recorded
that:
…the adze makers view the stones they work with
as living, intentional subjects. Thus, knappers will
speak of stones being ‘angry’ if they fail to fracture
as desired and will call out to them using their ‘secret
names’ as they search for them at the quarry sites
along the river. The boulders themselves are believed
to grow with age as people do … with ‘old stone’
(wisy-ya) being darker and stronger than ‘young
stone’ (ya-babau). Social relations with stone are an
important part of production, and care must be taken
to avoid angering pieces through improper practices
such as placing finished pieces on the ground in an
improper orientation (Stout 2002, 704).
When knapping, bifacial axe blanks that broke un-
expectedly were said by stoneworkers to have fallen
ill and died (Hampton 1999). Similarly, it is reported
that Langda people bring stone axes that break in
the fields back to their villages for respectful discard,
claiming that they ‘feel sorry’ for the tools (Toth et al.
1992, 92).
For many indigenous people, stone used for
making tools is believed to grow in the ground like a
plant, and even give birth to young. For example, at
the Ngiliptiji blade quarry in eastern Arnhem Land,