This is a free oprint – as with all our publications
the entire book is freely accessible on our website,
and is available in print or as PDF e-book.
PETER M.M.G. AKKERMANS
THE ARCHAEOLOGY AND EPIGRAPHY OF JORDAN’S
NORTH-EASTERN DESERT AND BEYOND
© 2020 Individual authors
Published by Sidestone Press, Leiden
Lay-out & cover design: Sidestone Press
Photograph cover: Peter Akkermans – Jebel Qurma Archaeological Landscape Project
ISBN 978-90-8890-942-9 (softcover)
ISBN 978-90-8890-943-6 (hardcover)
ISBN 978-90-8890-944-3 (PDF e-book)
Introduction: landscapes of survival 9
Peter M.M.G. Akkermans
First inhabitants: the early prehistory of north-east Jordan 17
New techniques for tracing ephemeral occupation in arid, 37
dynamic environments: case studies from Wadi Faynan and
Wadi al-Jilat, Jordan
Populating the Black Desert: the Late Neolithic presence 59
Yorke M. Rowan, Gary O. Rollefson and Alexander Wasse
Flamingos in the desert: how a chance encounter shed light on 79
the ‘Burin Neolithic’ of eastern Jordan
Alexander Wasse, Gary Rollefson and Yorke Rowan
Pastoralists of the southern Nefud desert: inter-regional contact 103
and local identity
The works of the old men in Arabia: a comparative analysis 117
Defending the ‘land of the devil’: prehistoric hillforts in the 145
The Late Chalcolithic and Early Bronze Age of the badia and 165
beyond: implications of the results of the rst season of the
‘Western Harra Survey’
Stefan L. Smith
East of Azraq: settlement, burial and chronology from the 185
Chalcolithic to the Bronze Age and Iron Age in the Jebel Qurma
region, Black Desert, north-east Jordan
Peter M.M.G. Akkermans and Merel L. Brüning
Identifying nomadic camp sites from the Classical and Late 217
Antique periods in the Jebel Qurma region, north-eastern Jordan
Harmen O. Huigens
The Nabataeans as travellers between the desert and the sown 235
Will M. Kennedy
The desert and the sown: Safaitic outsiders in Palmyrene territory 255
Jørgen Christian Meyer
The north-eastern badia in Early Islamic times 265
Depicting the camel: representations of the dromedary in the 287
Black Desert rock art of Jordan
Nathalie Østerled Brusgaard
Bows on basalt boulders: weaponry in Safaitic rock art from 305
Jebel Qurma, Black Desert, Jordan
Keshia A.N. Akkermans
‘Your own mark for all time’: on wusūm marking practices in the 317
Near East (c. 1800-1960 AD)
Rock art in Saudi Arabia: a window into the past? First insights 333
of a comparative study of rock art sites in the Riyadh and Najrān
Grati and complexity: ways-of-life and languages in the 343
Hellenistic and Roman harrah
Michael C.A. Macdonald
Gaius the Roman and the Kawnites: inscriptional evidence for 355
Roman auxiliary units raised from the nomads of the harrah
Ahmad Al-Jallad, Zeyad Al-Salameen, Yunus Shdeifat and Rafe Harahsheh
Remarks on some recently published inscriptions from the 363
harrah referring to the Nabataeans and the ‘revolt of Damasī’
Two new Safaitic inscriptions and the Arabic and Semitic plural 391
Phillip W. Stokes
In: Peter M. M. G. Akkermans (ed.) 2020: Landscapes of Survival - The Archaeology and
Epigraphy of Jordan’s North-Eastern Desert and Beyond, Sidestone Press (Leiden), pp. 305-316.
Bows on basalt boulders: weaponry in
Safaitic rock art from Jebel Qurma, Black
Keshia A.N. Akkermans
The Safaitic rock art of Jordan’s Black Desert is a fascinating yet under-examined subject.
In this contribution, I discuss the representations of weapons in the rock art of the Jebel
Qurma region in north-east Jordan. Additionally, I will give an overview of the material
evidence of weaponry produced by recent excavations in the region’s burial cairns.
Detailed visual analysis distinguished four categories related to weaponry in the rock art:
bows, pole weapons, swords/daggers, and shields. Patterns in the use of these objects vary
for each category. Most notable are the rm association of lances with riders on animal-
back, and the archers that are predominantly depicted on foot.
Keywords: weaponry, rock art, Jordan, Jebel Qurma, desert communities, bow-and-
arrow, Near East
The north-east of Jordan is home to the ‘Black Desert’, which is an extensive, arid region
dominated by dark basalt uplands (harrah), alternating with vast gravel plains (hamad).
Despite the region’s forlorn ambience and harsh climate, it holds a wealth of archaeological
remnants of many dierent periods. Inscriptions in Safaitic and petroglyphs were
carved into many of the dark basalt boulders that litter the landscape, and these show
a variety of gurative and geometric motifs (Brusgaard 2015; 2019; Brusgaard and
Akkermans, in press). Whereas the Safaitic texts have received considerable attention
(see e.g. Al-Jallad 2015 and references therein; Littmann 1943; Macdonald 1993), research
on the contemporaneous petroglyphs has remained limited. A notable advancement is
the recent doctoral dissertation focussing on the rock art of the Jebel Qurma area by N.
Brusgaard (2019), which comprises the rst systematic, contextual study of the contents
and motifs of the rock art of the Black Desert (see also Brusgaard, this volume). Other
important studies have focused mainly on specic motifs, such as women and chariots
(e.g. Macdonald 1993; 2009).
The majority of gurative depictions in the rock art of the Jebel Qurma region consists
of zoomorphic gures, while anthropomorphic representations make up a relatively
small share (Brusgaard 2015; 2019). The main category of material culture depicted in
the petroglyphs is that of weaponry. Studies on the motifs and objects in rock art can
provide valuable insights into the material culture used by its makers (May et al. 2017),
especially in regions where the archaeological record is complicated by numerous issues.
For example, artefacts found in the burial cairns in the Jebel Qurma region are often
306 LANDSCAPES OF SURVIVAL
poorly preserved, due to the porous construction of the
cairns, their continuous re-use, looting, and the region’s
extreme climatic conditions (Akkermans and Brüning
2017, 135). Such conditions also impacted the preservation
of weaponry in these tombs, and excavations therein
retrieved only a handful of examples. In this article, I will
discuss both the material and rock-art manifestations of
weaponry in the Jebel Qurma region in the Classical and
Late Antique periods.
Weaponry in the material record
In the Jebel Qurma area, burial tombs were the only places
that contained preserved weaponry. Two tombs yielded
some weapons made of iron, in the form of at least ve
arrowheads, a javelin, and a potential spearhead (Fig. 1)
(Akkermans et al. 2020). The javelin and four of the
arrowheads were found in the original (but collapsed)
grave chamber of a so-called ‘ring cairn’ at the site of
QUR-80. The cairn had undergone several phases of re-use
and alteration since its initial construction. Looting activity
in antiquity left the original grave heavily disturbed.
Moreover, a subsequent phase of re-use levelled the top
half of the ring cairn and used it as a foundation to build a
straight-walled ‘tower tomb’ on top of it. A fth arrowhead
and the possible spearhead were found in a burial in a
round tower tomb at the site of QUR-98. Iron weaponry
similar to the pieces presented here was found earlier at
sites like Lachish, Al Khadr, and Beer-Sheba, and suggest
the use of the two Jebel Qurma cairns between the late
ninth and seventh century BC (see Cross and Milik 1956;
Gottlieb 2004; 2016).
The arrowheads have simple, leaf-shaped blades
and lozenged tangs. Unfortunately, they are all
severely corroded, which makes precise typological
identication dicult. Most of them have a mid-rib that
is strongly pronounced. Nearly all of the arrowheads
are fragmented: only one has its tang still attached.
The arrowheads range between 4.5 cm and 6 cm in
length, 1.0 cm and 1.5 cm width, and, depending on
the presence or lack of a mid-rib, between 0.3 cm and
0.8 cm in thickness. The javelin (found in the tomb at
QUR-80 together with four arrowheads) has a short and
relatively thick leaf-shaped blade with a long, rounded
tang. It measures 10 cm by 1.5 cm and is 1.0 cm thick,
but the object’s corroded exterior belies its dimensions.
A central rib on the blade is either fully lacking or
masked by corrosion as well. Eight iron fragments were
also found together, all of which have a central rib on
one side of the blade. They could be re-joined to form
the long, narrow blade of a single spearhead about
17 cm in length, 2 cm in width, and 0.6 cm in thickness.
Alternatively, the pieces may come from two somewhat
shorter blades. The corrosion of the iron does not rule
out either of these options.
Figure 1. Weaponry found during the 2017‑2018
eldwork campaigns in cairn burials in the Jebel Qurma
region. First row: severely fragmented blade, possibly a
lance head. Second row: front and side view of a javelin
blade with a long tang. Third row: front and side views of
three arrowheads. The central rib is clearly visible on the
second point from the left. Two possible rounded tangs
are lacking adjoining blades (photograph: Jebel Qurma
Figure 2. Armour found in cairn burials of the Jebel
Qurma region during the 2017‑2018 eldwork
campaigns. First row: front and side views of four armour
scales of various sizes. Note the curvature at the bottom
of the plates. The two scales on the left are still joined
with their rivets. Second row: front and side view of
a simple rectangular wrist‑guard made of sandstone
(photograph: Jebel Qurma Project Archive).
BOwS ON BASALt BOULDERS
A more common nd in the burial cairns were the
bronze armour scales, which were about 2.5 cm long,
1.5 cm wide and 0.1 cm thick (Fig. 2). They came from
graves dated to the mid and late rst millennium BC
(Akkermans and Brüning 2017; Akkermans et al. 2020),
but never were present together with iron weapons
or with any other weaponry. Generally, their shape is
either polygonal or rounded, with a wider top that
tapers at the base. On multiple occasions the scales still
possessed their rivets, used for attaching the plate to
the fabric corselet underneath. Most of the scales are
curved at the wider end, presumably to further aid
adherence to the under-cloth. The scales were found in
several tombs but, curiously, their quantity per grave is
very low. Most scales have been found as single pieces
or, occasionally, in small groups of only two or three
per cairn. Perhaps they were segments that fell o
during looting activities in antiquity which removed
the armour corselets from the tombs. Alternately (and
probably more likely), the armour scales may have
served a ritual purpose in the mortuary practices of the
region. Maran (2004, 23) ascribes an apotropaic element
to individual appearances of armour scales; a single
scale would have contained protective qualities equal to
a complete, plated armour coat.
Excavations in another burial chamber identied
a remarkable example of non-metallic weaponry, a
sandstone archery wrist-guard. The object measures
about 7.3 cm long, 2.5 cm wide, and 1.0 cm thick, and has
a straight-sided outline with rounded corners. A single
perforation is present on either end of it. The object
appears to be made out of a layered type of sandstone.
The layers have disintegrated in places, with the surface
of one side of the object starting to crumble and ake o.
From what is left of the brace, its cross-section appears to
have been straight on both sides.
The wrist-guard stems from a much earlier context
than the other objects discussed so far. It was found
in the burial chamber of a ring cairn at the site of
QUR-207, which radiocarbon evidence has dated to
the late Early Bronze Age IV period, c. 2010-1890cal BC
(cf. Akkermans and Brüning, this volume, Table 1).
Remarkably, the grave has yielded no other objects
related to archery, such as metal or stone arrowheads.
In fact, this context ts well into the wider discussion
regarding the functionality of these stone wrist-guards;
some scholars ascribe a more decorative or symbolic
signicance to them rather than a practical use, based
on their decoration and their intended position on the
arms (Woodward et al. 2006; Fokkens et al. 2008). The
burial at QUR-207 oers little help in interpreting the
object: the relatively poor preservation of the skeletal
remains did not allow for a reconstruction of how and
where the object was worn on the body.
Weaponry in rock art
Extensive eldwork in the Jebel Qurma area since 2012
has yielded many thousands of Safaitic inscriptions
and petroglyphs (Brusgaard 2019; see also Brusgaard,
this volume). The present analysis is based on rock art
documented in the region during surveys between 2012
and 2016. In this time span, more than 300 sites were
meticulously searched for petroglyphs. The survey
documented all encountered rock art, forming an extensive
corpus of both ancient and modern petroglyphs that
portrayed a large variety of subjects. Most of Jebel Qurma’s
Figure 3. Stacked bar
chart showing the types
of weapons and weapon
combinations, as well as the
number of anthropomorphic
gures holding them. The
unidentiable objects (‘uncl.
object’) are also included.
Figures mounted on an
animal are displayed in light
ochre and gures on foot
are indicated by dark ochre.
308 LANDSCAPES OF SURVIVAL
rock art can be found on large, isolated basalt boulders in
the landscape. The petroglyphs tend to cluster in specic
locations high up on the hills, which oer a good overview
of the landscape (Brusgaard 2019). Carving, pounding, or
pecking removed parts of the dark upper surface of the
rock, revealing a lighter stone base underneath.
As with most rock art, dating the petroglyphs in the Jebel
Qurma region has proven to be dicult. Conventionally,
they are dated between the rst century BC and the fourth
century AD, although this dating is based on meagre
evidence and informed speculation; the chronological
boundaries are, in fact, unknown (Al-Jallad 2015, 17). The
surveys and excavations in the Jebel Qurma area have
shown that rocks with petroglyphs were regularly re-used
as building material for tombs that probably date as early as
the third and second centuries BC (Akkermans and Brüning
2017; Akkermans et al. 2020). This observation strengthens
the presumption that the Safaitic rock art of Jebel Qurma is
several centuries older than previously believed.
The 2012-2016 eldwork yielded a total of 4637 rock-art
motifs. Anthropomorphic gures take up 562 (about 12%)
of this amount, including those mounted on animals.
Of these, 268 gures are shown holding objects, most of
which are weapons. A number of these gures hold ‘lead
ropes’ and ‘whips’; since these are not classed under
weaponry, they are excluded from this study. 56 objects
could not be identied. In general, the petroglyphs of the
Jebel Qurma area display little detail, making it dicult to
identify weapons. Altogether there were 198 gures with
a clearly depicted weapon or a combination of weapons,
divided over 172 individual rock-art panels (Fig. 3).
The depicted weaponry can be divided into four
categories: (1) bow-and-arrow; (2) pole weapons; (3) swords
and/or daggers; and (4) shields. These types of weaponry
could be used individually and in any combination. The
shields, although technically not a weapon, are considered
part of the weaponry equipment as defensive gear.
The largest share of the petroglyph weapon assemblage
from Jebel Qurma is made up by the bow-and-arrow,
with 99 anthropomorphic gures holding this weapon. In
rare occasions the quiver is indicated as well, being worn
on the back with the etchings of the arrows pointing
upwards (Fig. 4). The arrows are generally depicted as
simple lines that are occasionally crowned by undened
arrowheads. The shape of the bows, on the other hand,
is clear and remarkably uniform throughout all its
depictions. Its appearance corresponds with that of the
composite bow, which is recognisable by its relatively
small size and its distinct arch, reminding one of a
cupid’s lip (Bowden 2012, 48). Several bows show a slight
curvature at their outer ends, classifying these as double-
convex composite bows.
The shape of the convex bow is the result of its structural
composition: the body is built upon a thin wooden core that
functions as its ‘skeleton’, onto which other components
are xed (Zutterman 2003, 126; Miller et al. 1986, 183). The
materials for these additional components are selected for
their capability to withstand and adapt to the pressure and
tension on the bow’s body when it is drawn (Zutterman
2003, 121). In antiquity, materials for bows generally
encompassed wood, bone, horn, sinew, and glue made
from the swimming bladders of sh (Paterson 1966, 70-77;
Bowden 2012, 44). Sinew was applied to the outer sides of
the bow and would be stretched when the bow was spun
(i.e. exed and held tight by a bowstring). The ‘belly’ of the
bow, which faced the archer, was made of horn and would
contract when the bow was used.
The materials used and the bow’s design made the
composite bow highly ecient. Performance equivalent to
that of the ‘self-bow’ or ‘laminated bow’ could be achieved
with less eort. A self-bow required much more strength
to launch an arrow of similar dimensions and weight at
the same speed as if one used a composite bow (Loades
2016, 5). In practical terms, this means that the composite
bow can be kept spun for a longer time, culminating in
a more precise aim (Miller et al. 1986, 185). The upward
curling notches of recurve composite bows would help
to reduce the recoil shock generated upon release of the
arrow. As a result, it causes the arrow to deviate less from
its intended ight path (Bowden 2012, 44).
The technology of the composite bow was by no
means restricted to the Black Desert, nor was it a ‘new’
technology. The earliest appearances of composite bows in
Mesopotamia date to the third millennium BC. Later, the
Hittites, Egyptians and Assyrians (among others) all used
composite bows of a rather triangular shape (Miller et al.
1986, 180; Zutterman 2003, 120-123). With the foundation of
the Persian empire by Cyrus II around 550 BC, the triangular
composite bow slowly faded out of usage. The recurve
composite bow replaced the triangular composite bow
across a wide geographic area, which can also be seen in
the rock art found in the Jebel Qurma area, and continued
to be used for another 1800 years (Rimer 2004, 85).
It is important to realise that the manufacture of a good
composite bow would have taken considerable time and
craftsmanship, and it could take a couple of months up to
two years before one was completely nished (Paterson
1966; Loades 2016, 5). The addition of an element to
the bow’s wooden base required a thorough period of
drying before adding the next element. Due to its lengthy
manufacturing time, previous research has proposed that
professional bowyers in sedentary communities produced
composite bows in large batches of a couple of hundred
at a time (McEwen 1978; Miller et al. 1986, 184, who point
out: ”A completed composite bow was a tour de force of
precision engineering and bonding …”).
BOwS ON BASALt BOULDERS
The current evidence suggests that the desert
populations of the Jebel Qurma area were nomadic
or semi-nomadic hunters and pastoralists (see e.g.
Akkermans and Huigens 2018; Akkermans 2019; Huigens
2019). Therefore, it is unlikely that they made the bows
themselves. If the rock art scenes are accurate depictions
of nomadic life in the desert and of its associated material
culture, we may wonder where the desert groups obtained
these bows from and whether they indicate specic types
of interaction between the desert and the settled regions
(which undoubtedly existed; cf. Macdonald 2014).
While the nature of such trade requires further
denition, a more direct interpretation is that the remaining
weapon categories depicted in the corpus of rock art from
the Jebel Qurma area were locally fabricated; its required
craftsmanship notwithstanding, the production process of
weapons such as lances, arrows, and spears is less lengthy
and requires a lesser variety of raw materials. A small,
temporary metal workshop identied at the site QUR-595
demonstrates that at least some part of the pastoralist
population in the region could work iron in a rudimentary
way (Huigens 2019; Akkermans and Brüning 2020). However,
radiocarbon samples of charcoal inside the installation at
QUR-595 produced a date between 100 to 385 AD, which is
considerably later than the suggested date of the physical
weaponry in the area. Additionally, there is a clear disparity
in the level of specialism needed to manufacture or repair
simple tools for everyday use and to forge high-quality
blades. With only one metal-working installation identied at
present, it would be rather premature to make assumptions
on the local people’s prociency in metal working.
At Jebel Qurma, 81 out of the 256 (c. 32%) weapon depictions
represent pole weapons. Pole weapons are part of the class
of ‘melee weapons’ or close-combat weapons, and t within
the sub-category of the pointed arms (Woosnam-Savage 2004,
417). Divergences in the length and mode of use in the rock art
suggest that they depict more than one kind of pole weapon.
Presumably spears, lances, and javelins were all part of the
repertoire of pole weapons in the harrah. Dierentiating
between these three weapons often results in a semantic
discussion, because it is unclear what the conventional terms
‘lance’ and ‘spear’ actually dene (Potts 1998, 183). The modern
Arabic word rumḥ’ can refer both to the lance and the spear.
The Safaitic word rmy, in turn, is connected to the Arabic verb
ramā, meaning ‘to throw’. Gordon (1953, 68) describes a spear
as “just a dagger at the end of a shaft”, which functions as “a
thrusting weapon with a longer reach”. Potts (1998, 183) uses
the term ‘spear’ to refer to a “light projectile which could be
thrown over a considerable distance at an enemy and for
which the term ‘javelin’ is sometimes employed”. The term
‘lance’ refers to “a much heavier and longer weapon which,
although it could be thrown a short distance, was more
commonly hand-held and used for thrusting in close combat.”
(ibid.). Most of the pole weapons in the rock art of Jebel Qurma
range from fairly long to very long in relation to the gure that
holds them, and are therefore most likely to be lances (Fig. 5).
In addition, they have a strong association with depictions
of anthropomorphic gures on equids and camel-back (see
below). The shorter pole weapons are often held by gures
on foot, and are regularly accompanied by a shield. These are
interpreted as spears.
Figure 4. Depiction of an archer with a composite bow. Note the quiver carried on the gure’s back, with the arrows
sticking out from the top (photograph: Jebel Qurma Project Archive; tracing by the author).
310 LANDSCAPES OF SURVIVAL
Swords and/or daggers
Depictions of swords and/or daggers are quite rare in the
Jebel Qurma region. Altogether only 15 swords/daggers
have been identied, although the actual number may
be higher. Apart from cross-guards and pommels, the
swords/daggers oer little visual clues for recognition
in the rock art. For example, the distinct lunate pommel
daggers, so characteristic for rock carvings in the Arabian
Peninsula (Newton and Zarins 2003; Aksoy 2017, 6), do
not occur in the Jebel Qurma region and, perhaps, in the
Harrat al-Sham at large.
The basis for identifying a sword or dagger in the
Jebel Qurma petroglyphs is if the object is held at an
outer end and seems relatively short. The most clear-cut
example in the corpus is that of a horseman carrying a
stick-shaped object at the waist (Fig. 6). A shorter bar that
diagonally crosses the front side of the rider’s body could
portray the cross-guard of the sword. While the carving
does not clearly depict the belt itself, the position of the
objects suggests that it is hanging from the hips, just as
a sword would when carried in a sheath. This coincides
with Michael Macdonald’s observation that Safaitic rock
art very rarely depicts horseman gures wielding a sword,
but they are shown occasionally with a sword at their belts
(Macdonald 2007, 282).
A second notable depiction of a sword comes from a
carving of a gure holding a long object in front of a camel
(Fig. 7). This object slightly widens before turning back into
a point at the outer end. What is most remarkable about
this depiction, however, is the clear curvature in the blade.
Its sickle-shaped blade is reminiscent to that of the Assyrian
sappara and the Egyptian khopesh. As early as 1917, Sir
William Flinders-Petrie noted with regard to the khopesh:
“the peculiarity of the type is the deep hollowing of the
back, and the projecting of the edge far in advance of the
handle. By its great curvature it was intended for a wiping
cut.” (Flinders-Petrie 1917, 27). The khopesh is not identical
to a sickle sword, although both types of weapons have
considerable similarities in shape. The sickle sword was
sharpened on the concave inner side of the curving blade,
like that of an agricultural sickle. The khopesh, on the other
hand, was sharpened on the convex outer side or, on rare
occasions, sharpened on both sides. There is evidence for
the use of khopesh-like arms or sickle swords in the third
to rst millennium BC in Assyria, Babylonia, Phoenicia,
Anatolia, Palestine, and Egypt (Gordon 1958, 23-24).
If the gure at Jebel Qurma is indeed holding a
khopesh or a sickle sword, it would be the only curved
sword represented in the local repertoire. The image is not
accompanied by a Safaitic inscription and therefore cannot
Figure 5. Scene with swordsmen/spearmen, archers, and a rider with a lance, all attacking a carnivorous mammal. At
least four gures carry a sword at the waist. The panel shows a clear divide in weapon choice between those on foot
and those on animal‑back. The Safaitic inscription and later additions are not traced (photograph: Jebel Qurma Project
Archive; tracing by the author).
BOwS ON BASALt BOULDERS
be unambiguously dated to the Safaitic period. The style of
the image also deviates slightly from the general Safaitic
rock-art corpus at Jebel Qurma, strengthening the suspicion
that the image belongs to a dierent time period. It is also
worth mentioning that the curvature of the sword perfectly
follows the curvature of the rock onto which the image was
made. Since it follows its curvature so well, it is possible that
the depiction did not intend to represent a curved sword at
all but formed into a curve because of the uneven surface
of the rock face. However, the neck of the camel depicted
next to the gure holding the discussed object also stretches
over the curvature of the rock but remains unaected by it.
At least 28 gures are depicted holding shields. One panel
shows three gures holding shields while drawing bows,
but most shields are accompanied by spears and swords
(see Fig. 8). All the shields are relatively small and round, but
Figure 6. Figure holding a lance and carrying a sword at the waist, while seated on an equid. Note the diagonal bar
crossing the shaft of the sword and the way in which it is carried at the waist (photograph: Jebel Qurma Project Archive;
tracing by the author).
Figure 7. A gure standing next to a camel carries a shield and curved sword. The sword curves alongside the edge
of the boulder face (indicated in grey), while the head of the camel is not aected by this curvature (photograph: Jebel
Qurma Project Archive; tracing by the author).
312 LANDSCAPES OF SURVIVAL
their patterning varies: nine have a circle in the middle, four
bear a central cross, two show feather-like decorations, one
has cross-hatching, and one has several lines radiating from
the centre towards the shield’s outer edges. The remainder
of shields either have decoration that is unclear or no
ornamentation at all. It is unclear what the shield adornment
represented. Circles on shields could be depictions of the
shield boss (umbo), which attaches the grip of the shield to
the shield itself using a convex, round piece of material in the
centre of the shield. Linear decoration such as cross-hatching
might imply leather strips on the front.
Lines and cross-hatching are not exclusive to shields.
Lines cover the bodies of a (very) small number of
anthropomorphic gures (Fig. 9). Among the nine known
patterned gures, the number of gures also handling
weapons is relatively high: three carry a sword and a shield,
and two hold a spear and a shield. Three scenes depict
gures both with and without patterned torsos. The patina
of the patterns in these scenes does not stand out from the
rest of the image, ruling out that they were later additions.
As with the case of the shields, it is unclear whether the
patterns are of functional or decorative nature. Stripes and
other lines also adorn some carved dromedaries, equids,
bovines and other animal species. Such patterns do not t
the animal’s coat colours or armour, suggesting they serve
as decoration (Brusgaard 2019). As for lines added to the
human gures, they could represent (plate) armour, straps
carrying swords, bows or quivers, or even clothing details
(although Safaitic engravings rarely indicate clothing;
Macdonald 2007, 274).
Modes of weapon use
Analysis of the employment of weaponry in the rock art
of Jebel Qurma shows distinct relational patterns between
weapon use and gures mounted on animals. The majority
of the 79 riders that carry weaponry ride equids, probably
either horses or mules/hinnies. Some ride dromedary
camels. The category of pole weapons has strong aliations
with mounted gures, as 62% of weapon-wielding riders
carry a spear or lance. On the other hand, the share of pole-
weapon use for gures on foot is about 5%.
Bow use shows quite the opposite trend: although bows
make up almost 66% of all weapons handled on foot, they
comprise less than 1% of the weapons used on animal-back.
The latter observation is surprising, as the composite bow
is highly suited for use on horse-back for several reasons.
Firstly, the composite bow has an elongated drawing time,
resulting in an improved aim. This is especially benecial
when the bow is operated while riding a moving animal.
Secondly, the size of the bow can be kept to a small size. A
small bow allows a mounted archer to move uidly and turn
his upper body all the way to the rear of the horse if needed,
known as the ‘Parthian shot’ manoeuvre (Overtoom 2017,
103; Zutterman 2003, 134). It is hard to explain the lack of
Figure 8. Two gures on foot holding short spears and small round shields. A stripe pattern decorates the torso of each
gure (photograph: Jebel Qurma Project Archive; tracing by the author).
BOwS ON BASALt BOULDERS
riders carrying bows from a functional perspective. Instead,
the recurring depiction of archers on foot, apparently
involved in short-range combat, might point to specic
cultural conventions (K. Akkermans 2017). The preference
for pole weapons seems to conrm a statement by Potts
(1998, 185) on the lance being “the principal weapon of
most ancient and indeed much modern cavalry.”
Several conclusions can be drawn from the study of weapons
in the material record and the rock art of the Jebel Qurma
region. Recent excavations of burial cairns have yielded
a few iron arrowheads, an iron javelin, a possible iron
spear, a stone wrist guard, and some bronze armour scales
(cf. Akkermans et al. 2020). The distribution of the bronze
armour scales is rather curious, with only one to three
scales per grave. This pattern can be explained as residual
evidence of looting activities or, more likely, as a testament
to the apotropaic use of armour scales in funerary rituals.
The majority of these objects are poorly preserved, which
makes more specic typological identication dicult. In
terms of chronology, the majority of the iron weapons come
from burial contexts that are likely to pre-date the Safaitic-
period petroglyphs. However, the temporal relationship
between the material and iconographic evidence of
weapons requires further research.
It is possible to discern two main categories of weapons
in the rock art: the bow-and-arrow, of which the composite
bow is especially prevalent, and pole weaponry. Of the latter,
the majority probably represent lances, which were used
primarily by riders. Spears only make up a small share,
and are used by pedestrian gures. Apart from the unclear
objects, 198 anthropomorphic gures can be recognised as
holding weaponry. Of these, 99 hold a bow (or a combination
of a bow and another weapon) and 82 hold a lance/spear (or
a combination of a lance/spear and another weapon).
There is a clear distinction between the use of the bow-
and-arrow and the use of pole weapons. The bow is almost
exclusively depicted with gures on foot. Pole weapons
are shown primarily with anthropomorphic gures riding
animals, in which case these weapons likely portray
lances. Despite being dicult to identify, there are at least
15 recognised instances of swords.
Shields are commonly depicted together with
other weapons. All shields are small and round or
ovoid in shape. The various patterns engraved on the
shields are particularly intriguing, yet their meaning
remains uncertain at present. However, there are rare
instances of similar patterns depicted on the torsos of
anthropomorphic gures as well, perhaps signifying
clothing, armour, or decoration.
It is still dicult to interpret these ndings on the theme
of weapons in the rock art of Jebel Qurma. The discovery
of new weapons by future eldwork campaigns will aid
in comparing between the weapons shown in petroglyphs
and those in the material record. As of now, chronological
Figure 9. A possible raiding scene revolving around two camels. Two gures on foot use composite bows, while two
riders seated on a single horse wield a long lance. Both riders have horizontally striped bodies. The Safaitic inscription is
not traced (photograph: Jebel Qurma Project Archive; tracing by the author).
314 LANDSCAPES OF SURVIVAL
discrepancies between the rock-art weaponry and its
physical counterparts impede any direct conclusions.
Moreover, it is important to establish the degree of realism
in the images. For instance, the lack of archers riding animals
may be the result of societal customs of weapon use, if these
depictions can be considered representative or expressive of
the reality of these societies. Like all symbolic systems, rock
art has a social function (Layton 2001). The rock-art tradition
of the Jebel Qurma region is highly repetitive. It eectively
forms a visual language based on the standardised shapes
of animals and anthropomorphic gures. The panels display
high levels of naturalism: scenes related to weapons show no
apparent signs of otherworldly activity, mythical narration,
or ritual or religious practices. Instead, the rock-art scenes
seem to remain fairly mundane, depicting events that were
likely more exciting than the general activities of everyday
life, such as a raid or a successful hunting party. At the same
time, the scenes do not portray domestic activities, apart
from two highly unclear depictions of what may be human
intercourse. There is a complete lack of carvings of domestic
structures, vegetation, trading caravans, etc.. The rock art
of the Jebel Qurma area is an expression of local cultural
and social norms, rather than an assemblage of random
depictions of interest to individual carvers (cf. Brusgaard
2019). These observations about petroglyphs compliment
previous statements by Al-Jallad (2015, 3) regarding the
Safaitic inscriptions, in which he notes that the texts are
decidedly formulaic and uniform. The range of subjects
presented in these carvings is limited and highly selective,
and thus by no means an emanation of ‘unstructured
Indeed, it would be incorrect to simply “read [rock
art] as a mirror of society” (Walderhaug 1998, 298).
Nevertheless, the insights that rock-art analysis can
provide about past worldviews and interactions should
not be understated. For example, the persistent depiction
of double-convex composite bows suggests that these were
physically present in the Jebel Qurma region. However, as
mentioned earlier, producing composite bows is a delicate,
time-consuming task that is best executed in a sedentary
context, which suggests that the mobile or semi-mobile
pastoralist societies of the harrah maintained intimate
relationships with settled communities elsewhere.
A nal word of caution is warranted: the outcomes
of this analysis should not necessarily be taken as
representative of the Black Desert as a whole (cf.
Brusgaard, this volume). Weapon assemblages and
patterns of weapon use may vary considerably throughout
the wider harrah. The populace that used the Safaitic script
and produced the petroglyphs likely comprised several
individual cultural or ethnic communities dispersed over
a broad geographic area, each with their own practices,
preferences, traditions, and material culture (cf. Al-Jallad
2015; Macdonald 2009).
This research was carried out as part of the Jebel Qurma
Archaeological Landscape Project, directed by Prof. Peter
Akkermans (Leiden University). I would like to thank
Peter Akkermans for his kind invitation to contribute to
this book, and for all the motivating, understanding, and
insightful conversations we shared about the material.
I also express my gratitude to Timothy Stikkelorum for
providing me with many fascinating notes on archery.
My special thanks go to Nathalie Brusgaard, who played a
quintessential role in the execution of this study with her
kind and useful feedback.
Akkermans, K.A.N. 2017. Battles in basalt: a study of
the weaponry depicted in the Safaitic rock art of
Jebel Qurma, north-eastern Jordan. Leiden: Leiden
Akkermans, P.M.M.G. 2019. Living on the edge of forced
into the margins? Hunter-herders in Jordan’s
north-eastern badlands in the Hellenistic and Roman
periods.Journal of Eastern Mediterranean Archaeology
and Heritage Studies7, 412-431.
Akkermans, P.M.M.G. and Brüning, M.L. 2017. Nothing but
cold ashes? The cairn burials of Jebel Qurma, north-
eastern Jordan. Near Eastern Archaeology 80, 132-139.
Akkermans, P.M.M.G. and Brüning, M.L. 2020. A Late
Roman pastoralist ironworking site in the north-east-
ern Black Desert, Jordan. In: Ahrens, A., Rokitta-Krum-
now, D., Bloch, F. and C. Bührig, C. (eds.). Drawing the
threads together: studies on archaeology in honour of
Karin Bartl. Münster: Zaphon, 103-119.
Akkermans, P.M.M.G. and Huigens, H.O. 2018. Long-term
settlement trends in Jordan’s northeastern badia: the
Jabal Qurma Archaeological Landscape Project. Annual
of the Department of Antiquities of Jordan59, 503-515.
Akkermans, P.M.M.G., Brüning, M.L., Arntz, M., Inskip,
S.A. and Akkermans, K.A.N. 2020. Desert tombs: recent
research into the Bronze Age and Iron Age cairn
burials of Jebel Qurma, north-east Jordan. Proceedings
of the Seminar for Arabian Studies 50, 1-17.
Aksoy, Ö.C. 2017. A combat archaeology viewpoint on weapon
representations in northwestern Arabian rock art. Medi-
terranean Archaeology and Archaeometry 17, 1-17.
Al-Jallad, A. 2015. An outline of the grammar of the Safaitic
inscriptions. Leiden: Brill.
Bowden, J. 2012. The origin and the role of the composite
bow in the ancient Near East. Ancient Warfare
Magazine 5, 42-54.
Brusgaard, N.Ø. 2015. Pastoralist rock art in the Black Desert
of Jordan. In: Giraldo, H. and García Arranz, J.J. (eds.).
Symbols in the landscape: rock art and its context. Pro-
ceedings of the XIX International Rock Art Conference
IFRAO 2015. Tomar: Instituto Terra e Memória, 761-767.
BOwS ON BASALt BOULDERS
Brusgaard, N.Ø. 2019. Carving interactions: rock art in the
nomadic landscape of the Black Desert, north-eastern
Jordan. Oxford: Archaeopress.
Brusgaard, N.Ø. and Akkermans, K.A.N. In press. Hunting
and havoc: narrative scenes in the Black Desert
rock art of Jebel Qurma, Jordan. In: Davidson, I. and
Nowell, A. (eds.). Making scenes: global perspectives on
scenes in rock art. New York: Berghahn Books.
Cross, F.M. and Milik, J.T. 1956. A typological study of
the El Khadr javelin and arrow-heads. Annual of the
Department of Antiquities of Jordan 3, 15-23.
Flinders-Petrie, W. 1917. Tools and weapons, illustrated by
the Egyptian collection in University College, London,
and 2,000 outlines from other sources. London: British
School of Archaeology in Egypt.
Fokkens, H., Achterkamp, Y. and Kuijpers, M. 2008.
Bracelets or bracers? About the functionality and
meaning of Bell Beaker wrist-guards. Proceedings of
the Prehistoric Society 74, 109-40.
Gordon, D.H. 1953. Swords, rapiers and horse-riders.
Antiquity 27, 67-78.
Gordon, D.H. 1958. Scimitars, sabres and falchions. Man
Gottlieb, Y. 2004. The weaponry of the Assyrian attack.
Section A: the arrowheads and some aspects of the
course of the Lachish siege battle. In: Ussishkin, D.
(ed.). The renewed excavations at Lachish, 1974-1994.
Tel Aviv: Institute of Archaeology, Tel Aviv University,
Gottlieb, Y. 2016. Beer-Sheba under attack: a study of
arrowheads and the story of the destruction of the
Iron Age settlement. In: Herzog, Z. and Singer-Avitz,
L. (eds.). Beer-Sheba III: the early Iron Age IIA enclosed
settlement and the late Iron IIA-Iron IIB cities. Winona
Lake, IN: Eisenbrauns, 1192-1228.
Huigens, H. 2019. Mobile peoples– permanent places:
nomadic landscapes and stone architecture from the
Hellenistic to Early Islamic periods in north-eastern
Jordan. Oxford: Archaeopress.
Layton, R. 2001. Ethnographic study and symbolic
analysis. In: Whitley, D.S. (ed.). Handbook of rock art
research. Walnut Creek, CA: AltaMira Press, 311-332.
Littmann, E. 1943. Safaitic inscriptions. Leiden: Brill.
Loades, M. 2016. The composite bow. Oxford: Osprey
Macdonald, M.C.A. 1993. Nomads and the Ḥawrān in the
late Hellenistic and Roman periods: a reassessment of
the epigraphic evidence. Syria 70, 303-403.
Macdonald, M.C.A. 2007. Goddesses, dancing girls or
cheerleaders? Perceptions of the divine and the
female form in the rock art of pre-Islamic North
Arabia. In: Sachet, I. and Robin, C.J. (eds.). Dieux et
déesses d’Arabie. Images et representations. Paris: De
Macdonald, M.C.A. 2009. Wheels in a land of camels:
another look at the chariot in Arabia. Arabian Archae-
ology and Epigraphy 20, 156-184.
Macdonald, M.C.A. 2014. Romans go home? Rome and
other ‘outsiders’ as viewed from the Syro-Arabian
desert. In: Dijkstra, J.H.F. and Fisher, G. (eds.). Inside
and out. Interactions between Rome and the peoples on
the Arabian and Egyptian frontiers in Late Antiquity.
Leuven: Peeters, 145-16.
Maran, J. 2004. The spreading of objects and ideas in
the Late Bronze Age eastern Mediterranean: two
case examples from the Argolid of the 13th and 12th
centuries B.C. Bulletin of the American Schools of
Oriental Research 336, 11-30.
May, S. K., Wesley, D., Goldhahn, J., Litster, M. and Manera,
B. 2017. Symbols of power: the rearm paintings of
Madjedbebe (Malakunanja II). International Journal of
Historical Archaeology 21, 690-707.
McEwen, E. 1978. Nomadic archery: some observations
on composite bow design and construction. In:
Denwoord, P. (ed.). Arts of the Eurasian steppelands.
London: SOAS, 188-202.
Miller, R., McEwen, E. and Bergman, C. 1986. Experimen-
tal approaches to ancient Near Eastern archery. World
Archaeology 18, 178-195.
Newton, L. and Zarins, J. 2003. Aspects of Bronze Age
art of southern Arabia: the pictorial landscape and
its relation to economic and socio‐political status.
Arabian Archaeology and Epigraphy 11, 154-179.
Overtoom, N.L. 2017. The Parthians’ unique mode of
warfare: a tradition of Parthian militarism and the
battle of Carrhae. Anabasis 8, 95-122.
Paterson, W.F. 1966. The archers of Islam. Journal of the
Economic and Social History of the Orient 9, 69-87.
Potts, D.T. 1998. Some issues in the study of the pre-Islam-
ic weaponry of southeastern Arabia. Arabian Archae-
ology and Epigraphy 9, 182-208.
Rimer, G. 2004. Archers and archery. In: Bradford, J.C.
(ed.). International encyclopedia of military history.
Walderhaug, E.M. 1998. Changing art in a changing
society: the hunters’ rock-art of western Norway. In:
Chippindale, C. and Taçon, P.S.C. (eds.). The archae-
ology of rock-art. Cambridge: Cambridge University
Woodward, A., Hunter, J., Ixer, R., Roe, F., Potts, P., Webb,
P.C., Watson, J.S. and Jones, M.C. 2006. Beaker Age
bracers in England. Antiquity 80, 530-543.
Woosnam-Savage, R.C. 2004. Edged weapons. In:
Bradford, J.C. (ed.). International encyclopedia of
military history. Oxford: Routledge.
Zutterman, C. 2003. The bow in the ancient Near East.
Iranica Antiqua 38, 119-165.