Weaponry

Few reliefs, early Arabic and Graeco-Roman texts as well as rock art from the Iron Age to Late Antiquity illuminate weaponry in south-western Arabia. In addition, excavated finds reflect the contemporary weapons in south-eastern Arabia. Outside analogies are essential to reconstruct the historic situation.

Fighting is a basic aspect of desert life which conditions all manner of behaviour, even the fabric of gender identification itself. Weaponry spanning from the Iron Age to Late Antiquity is known to us in south-west Arabia, illuminated indirectly by early Arabic, few Graeco-Roman texts and recently from local rock art. While Ancient South Arabian texts are rich in conflict-related vocabulary and description of military infrastructure, rarely is a weapon ever mentioned. On the other hand, contemporary weaponry in south-eastern Arabia is known from excavated finds, particularly in recent years. In comparison, Mesopotamian cuneiform sources and representations in the visual arts contrast with how minimal our surviving South Arabian sources are. We can only understand the weaponry of Arabia by means of outside analogies, especially those of Romans, Parthians and their satellites. Even if not documented, pre-Islamic military equipment such as standards must have existed, as with contemporary lands and in early Islamic times. The Arabic ṣilāḥ adequately frames the topic, connoting both offensive weapons and defensive armour and equipment, but excluding the related topic of fortifications.

South-western Arabia

Rare indeed are Ancient South Arabian texts which mention arms (Sabaic: s¹lḥ; Robin 2019: 146). Exceptions include ks³w (military garment), s¹lṭ (divinatory arrow), ʿqbt, ʾṣnʿ (fortress) and rḥl (saddle). In contrast, cuneiform sources contain hundreds of such mentions (Seidl & Stol 2015). From their own vantage point, nobles in pre-Islamic Arabic poetry mention their own weapons (Beeston 1976: 13), but not those and soldiers of the lower classes (e.g. slinger: Ar. sing. rāfiḍ). This includes par excellence treatises on equitation (Late Islamic, Ar. sing. furūsiyya), with their pre-Islamic roots. Thus, important distinctions, such as the difference between light and heavy cavalry, find no expression.

The sword finds most frequent mention in early Arabic sources (Schwarzlose 1886: 124 n. 1; Bosworth 2004: 735). For the Graeco-Roman, Christian and Islamic world, like the horse, it is a rank-indicator. On the other hand, when Beeston positivistically wrote: “I do not believe that this [bow and arrow] was, at least in Arabia, a normal weapon of war” (1976: 13), he ignored the reality of contemporary Parthian and Roman warfare, which could hardly exist without archers and slingers, e.g. at the battle of Carrhae (today, SE Turkey). Sources such as Xenophon (Anabasis 3, 316–17; ca. 400 BCE) show the basic but unappreciated role of slingers in ancient warfare. The literary documentation of Roman and enemy slingers, battlefields and sites where sling balls have occurred is long, geographically disparate and analogous to our topic (Völling 1990: 55–7). The pre-Islamic origins for weapons described in Islamic texts are clear enough for most, despite missing musnad mentions. Hundreds of find-spots throughout the Near East for sling and archery equipment point to a temporal and spatial universality of these weapons from the Stone Age to recent times, even if only rarely in Ḥaḍramawt (ethnographic occurrence: Korfmann 1972: 6–8). If the bow is one of the “most important cultural goods of humanity” (ibid.: 45), its pre-Islamic existence cannot be dismissed as the result of a defective local heritage chain.

While a knight should be an excellent huntsman and archer (e.g. the reliefs of Ḥaẓiyān and Kāthibat, see below), professional archers belong to a lower social class than do knightly warriors, which explains why archery finds little mention in Arabic texts. Further down the social scale, essential are slingers for battle (Latin sing. fundator; Ar. sing. rāfiḍ) which in South Arabia are archaeologically even less visible. These need no elaborate equipment, as opposed to expensive bows, swords, armour etc.

Rare are Graeco-Roman textual descriptions of the local South Arabian weaponry, e.g.: “Here the barbarians joined battle with the Romans, and about ten thousand of them fell, but only two Romans; for they used their weapons in an inexperienced manner, being utterly unfit for war, using bows and spears and swords and slings, though most of them used a double-edged axe” (Strabo Geog. xvi,4,24).

Selected examples

Archery

The double convex bow is well documented from contexts. Reliefs, Sasanian (Fig. 1) at Ḥuṣn al-ʿUrr (cf. Yule 2019b: 69, fig. 104c; 131, fig. 104c); Reliefs, Sabaean: Ḥaẓiyān, the archer and Kāthibat (Yule & Robin 2007: 362, figs. 165-6); Bayt al-Ashwal (Yule 2019b: 67, 137, fig. 96). Rock art: Ḥimà, Saudi Arabia (Robin 2019: 149, fig. 5; 151, fig. 8).

Axe

Reliefs: Fenestrated axe: Depicted in Yule and Robin in the so-called ‘knight relief’ (Fig. 2) and ‘page relief’ (2007: 360, figs. 160-2). Iron halberd (pres. h. 30.5 cm) from Wādī Ḍuraʾ, Umm Hunayka, gr. 3, 2nd–4th century (Breton & Bāfaqīh 1993: 56, pl. 17, fig. 48; pl. 30, fig. 94).

Chariot

Not documented.

Elephant

According to the Islamic-Arabic tradition, Abraha (r. ca. 533–after 558) used elephants in his campaign against Mecca. The Maʿadd are targeted by an expedition of Abraha mentioned in the South Arabian inscription Murayghān 3 but there is no indication that it incorporated elephants. Rock art: al-ʿUlā in the Ḥijaz: Fig. 4; in Ḥimà, there are three such images (Robin 2018). The use of battle elephants lost its meaning with the introduction of firearms.

Equitation & camel riding

Horseback riding in rock art: Ḥimà and Murayghān (Robin et al. 2017, Robin 2019: figs. 18-25). Horseback riding engraving (Breton & Bāfaqīh 1993: pl. 24, fig. 69). Horseback riding relief from Ẓafār (Fig. 2). Metal statue from Ghaymān (Yule 2019b: fig. 80).

Camel riding in rock art: Ḥimà (Robin 2019: figs. 9-10). Metallic vessel engraving which shows a camel saddle: Wādī Ḍūraʾ, Umm Hunayka, stray-find (Fig. 5; Breton & Bāfaqīh 1993: 60, pl. 11, fig. 28, pl. 25, fig. 70.4).

Given analogies, as in the Roman period, such as the famous Numidian cavalry with minimal equipment, the use of the camel comes as no surprise.

Helmet, armour, military clothing

Reliefs: scale armour & helmet on the ‘knight relief’ and ‘page relief’ (Fig. 2, both Yule & Robin 2007: 358–60, figs. 157–62). The ‘page’ wears armour padding. The Bashlik (hood) is worn on the relief of Ḥuṣn al-ʿUrr (Fig. 1).

Horse tack

Noseband from al-ʿIrāfa tomb ar1, 5th century CE (Ẓafār Museum, zm2565; Yule* 2019b: 39, 186, figs. 45, 83, 111, 120).

Lance/spear

Main cavalry weapon. Relief: Sasanian relief from Ḥuṣn al-ʿUrr (Yule 2019b: fig 104a) ‘knight relief’ from Ẓafār (Fig. 2). Rock art: Ḥimà, portrait of Tamīm with an over-dimensioned lance (Robin 2019: 153–4, figs. 12-14).

Spears have an iron butt at the lower end which can be stuck in the ground when the weapon is not being carried (Ar. zuǧǧ; Bosworth 2004: 735).

Shield/buckler

The ‘knight relief’ from Ẓafār (Fig. 2) shows a large multi-purpose shield and a buckler, the latter used basically for close fighting, poor protection against missiles. Another Ẓafār relief (zm4) shows a shield in a (idealised) Romanising heroic depiction (Yule 2019b, fig. 88). A buckler may be fashioned of thick leather or wicker, weighs ±1 kg, is also useful as a fist.

Sling

David and the Philistine Goliath are the classical Iron Age example (1. Sam. XVII 40–48). Ethnographers and ancient historians take it more seriously as a strategic weapon than do prehistorians. Slinging is every bit as deadly as archery (Korfmann 1972: 3) and can kill a horse (ibid. 1972: 15). In pre-Islamic Arabic texts (Schwarzlose 1886: 319), the flexible sling (Latin: Funda, Ar. miqlāʿ) differs from the staff sling and slingshot. Evidence comes in the form of accumulations of stones of homogeneous size in certain contexts. Sling bullets have different ballistic properties, lead being superior (Fig. 7). The range is most effective within 100 m, up to 300 m, and as effective as archery (Baatz 1990: 61). The accuracy depends on distance. The bullet weight is 50–500 g.

Historic and recent bullet finds in Yemen and Ḥaḍramawt: Korfmann 1972: 8 (no. 26).

Rock art: Dhī al-Ṣawlaʿ near Ẓafār (Fig. 6 upper left)?

Standard

Standards are for military rallying points and symbols of power. Relief: Ẓafār Museum zm0450 (Yule 2019b: fig. 50). Rock art: possibly Fig. 3, if not an over-dimensioned sword.

Sword/scabbard

Stabbing and thrusting weapon worn from a baldric or at the waist in a colourful scabbard (Schwarzlose 1886: 208–9). Straight swords are for hand-to-hand combat rather than as a cavalry weapon. However, cutlasses are a sub-recent.

In pre-Islamic and early Islamic Arabic poetry, Swords frequently were named after victorious places and places of origin, real and imagined (Schwarzlose 1886: 130–5). Damascene and Indian blades were renowned (Schwarzlose 1886: 127).

Relief: Ẓafār z607 (Yule 2019b: fig. 100). Rock art: Ḥimà, possibly on Fig. 3 and the portrait of Tamīm (Robin 2019: 156, fig. 14). Scabbarded, single-edge short sword (class S16) from Wādī Ḍuraʿ, Umm Hunayka gr. 3, 3rd–4th century CE (Breton & Bāfāqih 1993: no. 45, pl. 16: figs. 42–3; pl. 29: fig 90).

War machine

Pre-Islamic desert Arabs had no war machines (Schwarzlose 1886: 319).

South-eastern Arabia

The vast majority of known SE Arabian metallic artefacts are weapons made of copper alloy from Early Iron Age (EIA) contexts. However, only 8 % of the 215 find-classes of metallic artefacts from this region are from single-phase contexts (al-Jahwari et al. 2021: 90, 187) which conditions dating reliability (chronological overview: Fig. 9). The vast majority of the EIA weapons are made of copper-alloy, although composite iron and copper alloy weapons appear. The paucity of weapons in the 3 rd millennium BCE is a result of survival chance (Fig. 8) since they are abundantly known in surrounding regions. The weaponry of LIA burials shows clear social class distinctions (Yule 2019a: 155, fig. 6) not recognisable in EIA burials, which are indeed rarely preserved. The two periods show radical differences in the weaponry.

Archery

Since the Upper Palaeolithic, archery was an essential occupation, despite a local lack of evidence until the Wadi Suq period, when the first evidence for metallic points comes to bear (Figs. 8-9; Yule 2019a: 163). The vast majority of iron arrowheads appear in the Samad Late Iron Age (LIA) and période préislamique récente (PIR). In the Samad LIA the finds consist of arrowheads as well as metal cramps from the quivers. Several miniature copper-alloy convex bows, quivers and arrows came to light in Mudhmar East in Adam in eastern Oman in EIA II context (Fig. 10). The only bow fragment to survive came to light in a grave at ed-Dūr in the U.A.E. dated to the PIR (Yule 2019a: 161). Bows were possible too valuable to sacrifice in a burial.

Axe

Axes in copper-alloy are numerous in the EIA. Both large, elaborate, non-functional ones with a small shaft-hole and small functional ones occur (Fig. 9).

Chariot

A representation of a chariot is engraved in a bronze phiale, a surface find which came to light in ed-Dūr (Fig. 11). It is drawn by two overlapping horses, has a driver, an archer, a drawbar, but no convincing wheels.

Elephant

The Sasanians used war elephants. Oral traditions implausibly mention the use of combat elephants in Oman (Yule 2019a: 151).

Helmet, armour, military clothing

An engraved phiale from ed-Dūr depicts two riders mounted on horses and camels who wear headgear and possibly wearing patterned shorts (Fig. 12).

Horse tack

Fig. 12 shows bareback riders, assigned to the PIR. Samad LIA horse bits have survived (Yule 2019a: 156, fig. 7).

Lance/spear

Lance and spear heads of copper-alloy are numerous in the Wadi Suq period (Fig. 9). Few in iron have survived. Depictions of speer-throwing riders have been attributed to the PIR (Fig. 12; Yule 2019a: 156, fig. 6). Ar. Qanāt, originally “bamboo, reed shaft” is used by synecdoche for the whole weapon. Organic artefacts have poor survival chances.

Buckler

The buckler (Ar. ters) is in evidence in the Hatraean and Palmyrene art, but only in evidence in the sub-recent period in SE Arabia and eastern Africa.

Sling

A clay ball 3.6 x 3.8 x 3.8 cm and 120 g, excavated from al-Maysar fortlet M34, Samad LIA (DA 4960) is a candidate for a sling bullet. Others came to light during the excavation of al-Rākī 2 which weigh 99, 107 and 139 g (unpublished DA 13032 & 13033).

Sword

Blades increase in length in SE Arabia during the EIA. They range in shades of grey from daggers to swords (Fig. 9). Rim-flanged daggers and swords are cast in both copper-alloy and iron (al-Jahwari et al. 2021: 50–7, 82–7) with several bimetallic composite (D22 class) daggers with good shape parallels especially in NW Iran. With the onset of the LIA heterogeneous forms appear.

Paul A. Yule

References and suggested reading

Sources

• Strabo Geog. XVI: The Geography of Strabo VII. Books XV-XVI. Translated by H.L. Jones (Loeb Classical Library, 241). 1930. Cambridge (MA), London: Harvard University Press, W. Heinemann.
• Xenophon, Anabasis: Xenophon. Anabasis, with an English Translation by Carleton L. Brownson (Loeb Classical Library), 1922, rev. 1980. Cambridge (MA), London: Harvard University Press, W. Heinemann.

Studies

• Al-Jahwari, N., P. Yule, Kh. Douglas, B. Pracejus, M. al-Belushi, A.T. ElMahi 2021. The Early Iron Age metal hoard from the Al Khawd area (Sultan Qaboos University) Sultanate of Oman (Archaeological Heritage of Oman, 6). Oxford: Archaeopress.
• Baatz, D. 1990. Schleudergeschoße aus Blei – eine waffentechnische Untersuchung. Saalburg-Jahrbuch 45: 59–67.
• Beeston, A. 1976. Warfare in ancient South Arabia (2nd–3rd centuries A.D.) (Qahtan fascicle 3). London: Luzac & Co.
• Bosworth, C. 2004. Ṣilah, Encyclopedia of Islam, 2^nd ed., vol. 12: 734–6. Leiden: Brill.
• Breton, J.-F. & M. Bāfāqih 1993. Trésors du Wādī Ḍuraʾ (République du Yémen): fouille franco-yéménite de la nécropole de Hajar am-Dhaybiyya (Bibliothèque archéologique et historique, 141). Paris: Geuthner.
• Korfmann, M. 1972. Schleuder und Bogen in Südwestasien von den frühesten Belegen bis zum Beginn der historischen Stadtstaaten. Bonn: Habelt.
• Robin, C.J. 2018, Les expéditions militaires du roi Abraha en Arabie centrale (vers 548-565 de l’ère chrétienne). CRAI 2018(3): 1313–1376. DOI: 10.3406/crai.2018.96589.
• Robin, C.J., 2019. À propos de l’armement des troupes ḥimyarites au VI^e siècle de l’ère chrétienne, in J.-F. Breton & F. Villeneuve (eds) La guerre en Arabie antique: 145–173. Paris: Geuthner.
• Robin, C.J. & S. Antonini de Maigret. 2017. Le cheval dans l’Arabie méridionale antique. Arabian Humanities 8. http://journals.openedition.org/cy/3284. DOI: 10.4000/cy.3284.
• Schwarzlose, F. 1886. Die Waffen der alten Araber aus ihren Dichtern dargestellt, ein Beitrag zur arabischen Altertumskunde, Synonymik und Lexikographie. Leipzig: Hinrichs.
• Seidl, U. & M. Stol 2015. Waffen im alten Orient. Bibliotheca orientalis LXXII(5-6): 613–626.
• Völling, T. 1990. Funditores im römischen Heer. Saalburg-Jahrbuch 45: 24–58.
• Yule, P. 2019a. Defense during the Samad period – a first attempt at an archaeology of conflict in south-eastern Arabia. JOS 20: 143–176.
• Yule, P. 2019b. Himyar. Late Antique Yemen / Ḥimyar. Tārīḫ al-Yaman al-qadīm. Aichwald: Lindensoft.
• Yule, P. & C.J. Robin 2007. Ḥimyarite knights and infantrymen, two reliefs. Arabia 3 (2005-2006): 261–271, figs. 157–69 on pp. 358–363.

_{Alternate spellings: Armament, Weapon, Weapons, Arms}