Skip to navigation – Site map
Le cheval dans la péninsule Arabique

Insight on the Ancient Arabian Horse from North Arabian Petroglyphs

Sandra L. Olsen


The origin of the Arabian horse breed has been the topic of much speculation. The earliest clear depictions of horses with Arabian attributes are in Egyptian New Kingdom tomb art, although they were likely first obtained as war booty during Asian battles. Rock art from the 1st millennium BCE to early 1st millennium CE indicates a possible start date for their arrival in the Arabian Peninsula. This study focuses on petroglyphs near Taymāʾ, Ḥāʾil, and al-ʿUlā oases, important caravan way stations along the incense route. Regional artists developed a characteristic style for depicting horses and camels, distinguishing them temporally and spatially. From its initial appearance the breed was closely associated with chariots, but its eventual value to equestrians is equally evident in Arabian rock art. Petroglyphs in this region elucidate physical conformation, military roles, and high status. Practical aspects and cultural indicators such as training, tack, grooming, and henna body painting are revealed in the detailed illustrations.

Top of page

Author's notes

I especially wish to express my tremendous gratitude to HRH Princess Adela bint Abdullah bin Abdulaziz al-Saud and HH Prince Faisal bin Abdullah bin Mohammad al-Saud and their Layan Cultural Foundation for supporting our research. I also want to offer our appreciation to the Saudi Commission for Tourism and National Heritage, headed by HRH Prince Sultan bin Salman bin Abdulaziz al-Saud, which has been instrumental in providing permission, transportation, and contacts with the local SCTNH authorities, ensuring the achievements of our project. We wish to express our thanks, also, to Dr. Ali bin Ibrahim al-Ghabban, the SCTNH Vice-President for Antiquities and Museums, our gifted photographer, Richard T. Bryant, and my colleagues in the field, Dr. Majeed Khan and Chris Beard.
I wish to express my gratitude to the editors of this compellation, Jérémie Schiettecatte and Abbès Zouache, for all of their work and initiative in bringing the authors together for this fine volume. It is also my pleasure to acknowledge the tremendous benefit this article has received from the comments of the two anonymous reviewers. Without their input, I would not have been able to comment on the graffiti written in ancient Arabic scripts, which I feel has greatly strengthened my interpretations of the figures.

Full text


  • 1 Olsen & Culbertson, 2010.
  • 2 Ibid.

1The inception of the Arabian horse as a breed extends far into antiquity,1 and identification of its place of origin is clouded by the ambiguities of the archaeological record. Despite the puzzling nature of the Arabian horse’s ancient past, many tantalizing clues about its development are revealed in various Egyptian, Arabian and Near Eastern art forms, as well as some Egyptian skeletal remains. Evidence today indicates that this gracile, yet powerful equine manifestation had emerged at least by ca. 1450 BCE in Egypt and was closely associated with light chariots from the beginning of its appearance in the archaeological record. Further data elucidate its social status, military roles, use by equestrians and details regarding its grooming, conformation and tack.2

2In fieldwork conducted by our team in 2011 of petroglyphs across western Saudi Arabia from north to south, it was found that there were distinctive styles in the depiction of domestic horses and that these varied from one area to the next and chronologically.

  • 3 Macdonald, 2004.

3This study is restricted to some of the key examples of images in the northern part of Saudi Arabia that possess characteristics of the Arabian horse breed and date from circa 9th century BCE to the 4th century CE, based on the co-occurrence of them with Ancient North Arabian inscriptions.3 The region where these sites are distributed encompasses the vicinities around Taymāʾ Oasis, in Tabūk Province, al-ʿUlā Oasis (ancient Dedan), in al-Madīnah Province, and Ḥāʾil Oasis, in Ḥāʾil Province (Fig. 1). The sites are grouped together because of their shared details, which imply a cultural standard or artistic style for the depiction of horses, as well as the closely associated camels.

Fig. 1: Map of Arabia showing sites in this study.

4[Image non convertie]

5The North Arabian artistic style produced an idealized caricature of the Arabian horse, capturing the key morphological features and body conformation standards still valued in the modern-day breed. Every individual living horse of the time may not have actually manifested all the critical attributes, and ancient artists probably preferred to depict their horses in the most flattering way. However, the standardization of the artistic representations over this large area can be interpreted as reflecting a shared ideal that has continuity spanning perhaps as much as three millennia.

  • 4 Boessneck, 1970; Bokonyi, 1993; Clutton-Brock, 1974; Chaix, 2000; Ducos, 1971; von den Driesch & Pe (...)

6The preferred way to identify Arabian horses in the archaeological record would be through morphological characteristics in skeletal remains, including the overall size, a dished rostrum on the cranium, or a missing lumbar vertebra in complete burials. A thorough study comparing numerous living breeds and ancient remains needs to be done before it can be determined which morphometric characteristics would be diagnostic for Arabians and when they appear in the archaeological record. Unfortunately, many regions that probably had early Arabian horses lack the archaeological skeletal evidence because of poor bone preservation, because early archaeologists failed to save the horse remains they excavated, or, in some cases, because it is difficult to distinguish equine species based on bone fragments. Across the Near East, sample sizes are very small and the bones so fragmentary that large-scale metric analyses are difficult. The best material comes from ancient Egypt, where several skeletons are known and have been measured and studied in detail.4 These show different types, with varying degrees of primitive and more specialized characteristics, but many are relatively gracile, as one would expect for the Arabian breed.

Arabian Horse Color and Conformation

7In order to evaluate ancient depictions of horses and identify them specifically as the Arabian breed or type, it is useful to enumerate the identifiable conformational characteristics that are recognized today in living Arabian horses and that provide a standard for comparison. Many of the features of the Arabian horse are adaptive and reflect the development of the breed in arid regions.

  • 5 Blunt, 2002.
  • 6 Weise, “A Reference Guide: Arabian Coat Coloration”.
  • 7 Amir, 1997.

8Arabian horses today can be found in a wide range of colors, including gray, chestnut, roan, bay, and rarely black, but all have a black skin for protection from the harsh desert sun. Most individuals have some kind of marking, such as sabino (white facial markings and leg stockings, as well as other occasional small spots on the body), or rabicano (a roan-like combination of white and dark hairs), but breeders and owners in Arabia regard too much white as “bad medicine”. In her travel journal, Lady Anne Blunt reported seeing some “undesirable white” in horses in northern Arabia in 1878, but her Crabbet stud farm stallion Mesaoud had a sabino coat.5 Although too much white has been distained in recent years, Abbas Pasha had 11 mares in his Egyptian farm that were parti-color (piebald), including one of his favorites, Faras Saouda.6 In the Book of the Good and Evil Signs of Horses, and their Auguries for Good or Evil, by Khalil Rizkallah al-Khuri of Gaza, as dictated by Shaikh ʿIsa al-ʿUqaidi,7 al-Khuri details the positive or negative omens associated with very specific color markings on a horse.

  • 8 Sadek et al., 2006.

9The modern Arabian horse has a very distinctive conformation that gives the animal both grace and incredible endurance. It has a light build and ranges from around 142 to 157 cm at the withers.8 Features on the head of the Arabian breed include: a short, wedge-shaped skull with a concave or straight facial profile across the bridge of the nose, a broad forehead, small ears, large eyes that provide a wide visual range, and nostrils that can flare to the diameter of a teacup to take in plenty of air when running. A strongly desirable characteristic is the “jibbah”, a boss or bulge on the frontal bone between the orbits that expands the frontal region of the brain and contains sinuses to manage the hot, dry air of the desert. A large jibbah also accentuates the dishing of the nasal bone and contributes to the concave facial profile. The muzzle has a very narrow anterior end. For this reason, Arabians may be fitted with special bridles with smaller mouthpieces that are on average 1.3 cm narrower than those for other horses of similar body size.

10The throatlatch, the base of the jaw where the head and throat intersect, should have good volume for the windpipe, referred to as a “well-set windpipe deposit.” To achieve this, the jowls should be deep and spread wide between the mandibular branches, especially in comparison to the narrow muzzle. In a fine living specimen, a person can place their fist up inside the throatlatch externally, between the branches. The throatlatch should be lean, with scant fat deposits, exposing the windpipe region to the cooling effects of air when it runs in the hot desert.

11The ideal Arabian has a “mitbah,” or powerful arched neck, set on high, with the head turned downward, rather than a “ewe” neck that turns upward or one that is excessively thick, short or long. An arched neck improves collection in the horse, which shifts the weight more onto the hind limbs and prepares the back for springing. Collection has many advantages, including enhancing sudden bursts when escaping predators or jumping. The Arabian neck is set well back into the withers, which are moderately high.

  • 9 Stecher, 1962.

12Arabians are “short-coupled,” i.e. they have a relatively short back compared to other breeds. The ideal ratio between the topline (withers to croup) and underline (point of elbow to stifle) is 1:2 in an Arabian, meaning its back is short compared to its belly. Most horses have 18 thoracic, 6 lumbar, and 5 sacral vertebrae. Many Arabian horses are missing a lumbar or more rarely a sacral vertebra,9 which contributes to their characteristic topline brevity. The Arabian’s chest should be deep and the ribs well-sprung to give the largest lung capacity possible. Given that they were initially bred as chariot horses in hot, dry environments, this is not surprising.

13The Arabian’s croup is one of its distinguishing characteristics. Ideally it should be horizontal because a good hip angle improves performance and speed. The tail carriage in Arabians is usually higher than in other horses. This is partly due to the angle of the hip, but having a tail lifted, or “flagged” also keeps the animal cooler when in motion.

The Arrival of the Horse and Wheeled Vehicles in Arabia

  • 10 Goldhahn, 2009.
  • 11 Jacobson-Tepfer, 2012.
  • 12 Littauer, 1977.
  • 13 Macdonald, 2009.
  • 14 Cherlenok, 2006.

14In the Arabian Peninsula, the earliest signs of the introduction of domesticated equines (probably horses), as well as the chariot (or possibly cart), appear in northern Saudi Arabia as schematic images in rock art. Examples of two equines pulling a light two-wheeled vehicle are now numerous and resemble ones found across Eurasia from Scandinavia10 to Mongolia11, Central Asia12, and North Africa13. The Arabian images are stereotypic, with two recumbent equines, usually back-to-back, with a draft pole between them connected to the chariot box or cart. The two wheels have four spokes each and are splayed out flat to the sides of the chariot body or cart so their structure can be shown (Fig. 2). These geographically widespread symbols are usually dated to the Late Bronze Age/Early Iron Age, i.e. the last half of the 2nd to the early 1st millennium BCE, and may shed light on the timing for the arrival of domesticated equines in Arabia. It should be stated that this author strongly believes these images represent horses and chariots, rather than donkeys and carts, since this stylized form of depictions is so widespread across the whole Eurasian steppes, where donkeys were unknown and where horses were buried with chariots, rather than carts.14

Fig. 2: Early schematic image of horses and chariot at Jubbah, Ḥāʾil Province.
Photo by Richard T. Bryant.

15[Image non convertie]

North Arabian Horse Petroglyphs and their Artists

  • 15 Macdonald, 2010.
  • 16 Eichmann et al., 2006.

16The people of these northern Arabian oases began to prosper in the 1st millennium BCE from the frankincense and myrrh trade,15 but the settlement at Taymāʾ Oasis was already surrounded by a 15-km wall by the 19th-18th century BCE cal., based on charcoal samples.16 Significant settlements around vital wells, like Biʾr Haddaj, at Taymāʾ, reflect well-developed agricultural communities, as well as the wealth generated from camel caravans bringing the precious incense products up from the south to be distributed to even more distant lands and cultures.

  • 17 Macdonald 2010.

17The petroglyphs, with a few noted exceptions, seem to have been generated by the local inhabitants, mostly nomadic herders, judging from the subject matter. The inscriptions associated with petroglyphs in the environs around Taymāʾ most often consist of a person’s name (probably the author’s signature), although a few are prayers or label the types of animal pictured. Images of Arabian horses and camels are regularly accompanied by Ancient North Arabian inscriptions in the form of graffiti. Hundreds of examples of graffiti written on cliff faces in the region around Taymāʾ provide some, albeit limited, insight into the people who left their marks. The nomads had some amount of literacy, although there was little standardization in how the letters were arranged and the bulk of the inscriptions are a single or very few words.17 The associated images stress the high value placed on their domestic animals, particularly horses and camels.

  • 18 Ibid.
  • 19 Ibid.

18A reference to ancient Taymanitic writing by Yariris, a Hittite regent of the city of Carchemish, circa 800 BCE,18 provides an indication of the antiquity of that local script. The earliest determined date for writing in the Taymāʾ region itself is in the mid-sixth century BCE, based on the mention of the Neo-Babylonian King Nabonidus (reign 556–539 BCE, Dynasty XI, or Chaldean Dynasty), who lived at Taymāʾ for ten years.19 The use of this script seems to have dwindled a few decades after Nabonidus introduced Aramaic in the oasis settlements themselves.

  • 20 Nehmé, 2005, p. 171.

19Most of the graffiti associated with petroglyphs and most likely written by nomads living outside the settlement is Thamudic B, C, or D, rather than Taymanitic. Thamudic D survived at least until the 3rd century CE on a tomb at Madaʾin Ṣāliḥ (known in ancient times by the Nabataean name Hijra or in Greek and Latin as Hegra).20 The vast majority of inscriptions are done in Thamudic B.

The North Arabian Artistic Style

20During the 1st millennium BCE and the first few centuries after, nomadic herders and possibly the occasional caravan traveler left their marks on sandstone surfaces. Their choice of subject matter included a variety of animals and fewer humans, but they chiefly depicted horses and camels, according to a regional style. Some smaller figures of ibex, oryx, gazelle, lion, dog, and other animals may also have been produced around this time, but often without the same standardized stylistic characteristics. Typically, the horses and camels dominated over other figures with their size and presence in the center of the petroglyph panels.

21The animals were made by scratching shallow incisions through the rock’s patina to make a lighter colored outline via exposure of the underlying unweathered stone. In some cases, a zone inside the outline or even the entire animal was scraped to indicate a coat pattern or color, artificial pigmentation, tack or other characteristics.

22The North Arabian style has numerous unique features. The camels usually have a muzzle with drooping lips, a large eye, two short pointed and erect ears, a long outstretched neck that slopes downward, a prominent semi-circular hump, vertical forelimbs that are straight and nearly parallel, pinched loin, and tufted tail. Most diagnostic for this style were the muscular, slightly flexed hind limbs that are rotated so that the rear is shown in an otherwise side view of the animal (Fig. 3).

Fig. 3: Camel and horse lower right panel at al-Naṣlah 2.
Photo by Richard T. Bryant.

23[Image non convertie]

Fig. 4: Arabian horse figure at the site of al-Naṣlah 2, with a wasm on its hindquarter.
Photo by Richard T. Bryant.

24[Image non convertie]

25Horses are also done in a style characteristic of this time and place (Fig. 4). They show many of the Arabian characteristics, including a short muzzle, concave facial profile, large eye, small ears, strongly arched neck, short topline, powerful hindquarters, prominent croup, and high tail carriage. Like the camels, the horses have a rotated posterior that faces the viewer, despite the fact that the rest of the body is viewed from the side. This “biangular” style of presentation is a key feature of this North Arabian artistic style, but seems to be more typical of camels and horses than wild game, like ibexes, for example. One possible conjecture for this would be the artists’ wish to emphasize breeding and, thereby, fertility in their domestic stock.

26The geographic range of this style has not been completely determined, but it seems to be most concentrated in the environs of Taymāʾ, extending eastward 240 km to Jubbah and 315 km to Ḥāʾil, and southwest 120 km to al-ʿUlā (ancient Dedan). Camels and other animals at Jubbah reflect this style, but horses are less common there.

Treatment of Horses in the North Arabian Petroglyphs

27The petroglyphs investigated in this study yield a significant amount of information about how Arabian horses were kept, trained, groomed, ridden, used in pulling chariots, served military functions, and more. Thus, despite the paucity of archaeological equine skeletal material so far, our knowledge of the roles they played and how they were cared for can be elucidated in Arabian rock art. Through this evidence, it becomes clear that Arabian horses were not only important as a means of transportation, but were also highly regarded and maintained as status symbols.

Horse Training

28Arabian rock art is an important source for learning more about the grooming, training and handling of early Arabian horses. The best possible example of horse training can be seen at al-Naṣlah, a site 47 km south of the oasis of Taymāʾ (Fig. 5) (also see below under heading Horse Figures by Site). The man in this scene is keeping the horse’s head in the right position by holding a lead attached to its halter taut in his left hand. In his right hand he holds an arrow vertically with the tip up. The projectile is out in front of the horse in its line of sight, perhaps to act the way a goad or a modern halter whip gets the horse to focus during training.

Fig. 5: Man training Arabian horse, using an arrow as a goad, al-Naṣlah 1.
Photo by Richard T. Bryant.

29[Image non convertie]


  • 21 Linduff & Olsen, 2011.
  • 22 Olsen, 2006.

30Tack that is shown in this sample of petroglyphs includes: bridles and halters, riding crops, saddle blankets, harnesses on chariot horses, and possible armor or caparison worn on the neck and breast of one chariot horse. Saddles and stirrups are lacking, but this is not surprising, given that they are not essential or were perhaps even unknown in the region at the time. The earliest known saddles date to the 8th century BCE in Xinjiang Uyghur Autonomous Region, in Northwest China,21 but they did not become common elsewhere until 400–200 BCE or even later. Many cultures preferred not to use the saddle, unless engaging in hand-to-hand combat on horseback or employing thrusting weapons while mounted. Stirrups are later still, appearing only around the 1st century BCE–1st century CE in India.22

Wusūm: Tribal Markings in Brands

  • 23 Khan, 2000; 2013.

31A wasm is a symbol placed by members of Bedouin on livestock as brands and incised into cliff faces and on boulders to indicate territorial boundaries and tribal affiliation. In addition, they served as a signature on documents, as grave markers, and to indicate a member of the tribe had been in that place.23 These are fairly simple geometric designs that have been used for millennia and in some cases have been reused later by modern Bedouins. When they appear in the rock art on a domestic animal, such as a horse or camel, they represent livestock ownership and depict brands. Examples have been found on a few of the horses, the best one being at al-Naṣlah 2 (Fig. 4).

Shaded Areas on Horse Figures

32Many of the Arabian horses depicted in the rock art show zones within the outlines of their bodies that have been scraped, exposing the light-colored rock surface underlying the natural patina. These can occur on the neck, both fore limbs, and/or both hind limbs, and the tip of the tail. There are several possible interpretations as to what these represent.

33It is typical for pelage color patterns to be depicted in various domestic and wild animals in Arabian petroglyphs, so this must be considered as one possible explanation for the shaded zones on the horse figures. As noted above, most Arabian horses have some small patches of white or other color, such as a blaze on the face or socks on one or more of the feet, but a piebald coat with large areas of white on the body is rare in the Arabian breed and has generally been discouraged in breeding.

34In the petroglyph horse figures, the leg zones terminate at precisely the same level on the left and right sides of the animal’s limbs and are solid, rather than “spotty” or irregular in outline. While some Arabian horses do exhibit fairly uniform white socks on more than one leg, repeated occurrences of identical zones on many horse figures on the same restricted anatomical parts argue for an interpretation other than coat patterns. Possible alternatives for these zones are presented here.

35For the feet, leg wraps could be one explanation for the shaded zones. Leg wraps help support the horse and prevent the hoof on the opposite foot or thorny vegetation from scraping the thin skin of the cannons. Modern leg wraps normally are confined to covering just the cannon bones (metacarpals and metatarsals), and this could account for multiple examples in the rock art (see below). Horses in 16th century Persian illuminations often show leg wraps, although this is very late compared to the Arabian rock art.

36An alternative explanation for these markings is that they represent clipping patterns. During the winter, a horse that works hard and sweats can take two or more hours to dry its heavier coat, with the animal getting chilled before the sweat evaporates. Even in the desert this can be a problem, since the temperature can drop suddenly once the sun sets. Grooms who clip horses use different patterns depending on what form of exercise the horse will be performing. Many of the modern clip patterns leave the coat on the legs all the way up to the elbow and stifle, while removing the hair on the head, neck and body. Some of the examples seen in our sample do run too high up on the legs to represent leg wraps, so could be explained by leaving the hair all the way up on the legs when clipping the coat on the body. This type of light-colored zone sometimes occurs on the neck, in addition to the legs, in which case it could represent a clipping pattern that left the coat on the neck and legs unclipped, while removing it from the trunk. Alternatively, the horse may have been fitted with a “neck sweat”, which would serve to keep the neck thin and conditioned.

  • 24 Personal communication, June 2016.
  • 25 Miczak, 2001.

37While all of these are possible explanations for zones on the horse figures, my colleague, and expert on Arabian horses, Cynthia Culbertson,24 suggests that the markings could reflect the application of henna to various parts of the horse. In fact, the use of henna on horses, donkeys, and other animals is a tradition that has been widespread, in Egypt, the Levant, Greece, Iran, India, Pakistan and elsewhere. Henna dye has been used by numerous cultures and in a variety of ceremonies, most commonly on brides for their wedding day, but also as a way of celebrating holidays, battle victories, births, circumcisions, birthdays, and so forth. Cows were painted with henna in India for festivals, sheep were decorated prior to slaughter in the Near East, and brides were sometimes given a painted cow as a wedding gift.25

  • 26 Harvard Museum of Art, object number 1958.76.

38Numerous depictions in medieval art from various cultures depict the faces, bodies, hooves or paws, and tails of horses, donkeys, mules, sheep, cattle and salukis decorated with henna. Salukis sometimes had it applied on holidays or to their paws for “protection”, because of its proclaimed medicinal qualities. This tradition continues today, particularly for hunts, in order to toughen and cool the paws. An image of a saluki with henna on its ears, legs, and tail can be found in a watercolor folio in an illustrated manuscript entitled Nighttime in a City, attributed to Mir Sayyid ʿAli, now in the Harvard Art Museum/Arthur M. Sackler Museum. 26It dates to the Safavid period of the Persian Empire, circa 1540, and comes from Tabriz, Iran.

  • 27 Miczak, 2001.

39Henna is derived from a plant known by a single species, Lawsonia inermis. The plant is a tall shrub or small tree with white flowers that is believed to have originated in Egypt or Ethiopia.27 To be used as a dye, dried leaves are crushed into a powder and mixed with water, tea, or lemon juice, and other ingredients may be added. According to traditional medicinal beliefs, it has numerous benefits, including advancing wound healing, an anti-inflammatory, analgesic, anti-bacterial, and anti-fungal that would keep nails and hooves healthy, and divert the “evil eye”. Henna flowers were also the source for a highly valued perfume, known in Latin as cyprinum.

  • 28 Ibid.
  • 29 De Moor, 1971.
  • 30 Miczak, 2001.

40The earliest evidence for the use of henna as a dye is during the Egyptian Neolithic Nagada II period (3650–3300 BCE), at Hierakonpolis, where mummies displayed traces of it in their hair and wigs.28 The mummy of Ramesses II (1320 BCE) exhibited henna stains on his hair, fingertips, palms, feet, and wrappings. Henna was used at least as early as the Bronze Age of Syria, where it is referenced in the Ugaritic legend of Baal and Anath.29 Ugarit thrived from 1450 to 1200 BCE. Henna reached Arabian trade routes by at least 900 BCE,30 so it would have been available when these horse petroglyphs were made.

41For horses, the legs, mane, tail and more rarely the head could be colored with henna. Sometimes even the body was covered with the dye. Sir John Chardin, in the third volume of his original Voyages de Monsieur le chevalier Chardin en Perse et autres lieux de l’Orient, published in 1686 (pp. 371–372), reported the use of henna on horses in Persia in the 17th century:

  • 31 Chardin, 1988, p. 169.

“They have a way also in Winter of Painting the Horses with Henna, that yellow Paint above-mentioned, used likewise by Men and Women; they anoint their Legs and their whole Bodies, up to their Breasts with it, and sometimes their Heads; they say that it keeps them from the Cold, tho’ it is rather used for Ornament; for in several Parts it is done at all Seasons: They paint on the King’s Horses for distinction sake, a broad Tagged Lace, with Flourishes like those of Coronets”.31

  • 32
  • 33 BM object number 1974,0617,0.19.4.

42Persian illuminations sometimes depict henna on horses, including a fine polo pony from the 10th century reproduced in Lady Wentworth’s 1938 book Thoroughbred Racing Stock. Another example of a Persian groom and horse, dating to the 17th century, shows all four legs tinted with henna. This folio is in the famous Moraqqaʿ-e Golšan, or Golšan Album, which is a compilation from 11th/17th-century paintings, drawings, calligraphy, and engravings by Mughal, Persian, Deccani, Turkish, and European artists, kept in the Golestān Palace Library, in Tehran.32 A print in the British Museum from an album folio probably from Deccan, India,33 shows 18th century women playing polo on ponies with legs and the tips of their tails painted with henna.

  • 34 Fuller, 1961, p. 84-86.

43Until very recently, in Shiraz, Iran, the practice was conducted on horses, mules, and donkeys to celebrate Nowruz, the Persian New Year, which begins on the vernal equinox. Still today, the Muslim villagers in Bawārij, Lebanon, celebrate the Springtime Festival by painting henna on their horses’ heads and tails.34 According to Fuller:

  • 35 Ibid.: 84.

“First comes the Thursday-of-the-Animals. On this day it is said all animals meet and mate, reproducing their kind. Household working animals are given a rest on this day, while henna, as a sign of blood and life, is dabbed upon their foreheads.” and, on the final Thursday of the Spring Festival, “The animals are decked in blue beads and colored garlands for the occasion and the tails of the horses are dyed in henna”.35

  • 36 BM object number 1920,0917,0.3.

44The zones on the tips of horse tails in the Arabian petroglyphs provide the strongest evidence for the use of henna, since no tack or gear would cover the very end of the skirt of a full, unplaited tail (Fig. 6). Comparable examples of henna dipped horses’ tails can be found in illuminated manuscripts, like one at the British Museum.36 This particular example was made in India during the Mughal Dynasty, circa 1650.

45Based on the evidence in this study of petroglyphs, it is likely that henna was being used to decorate the bodies of horses in northern Saudi Arabia during the 1st millennium BCE to early 1st millennium CE. It cannot be ruled out that some of the figures depict clipping or leg wraps, as well.

Fig. 6: The horse on the right has a full tail skirt with the tip shaded, al-Daksh, near Taymāʾ. Photo by Richard T. Bryant.

46[Image non convertie]


47A number of grooming practices are recorded in the images of horses illustrated in Arabian rock art. The mane could be handled in varying ways. In some cases, the mane appears to be trimmed all the same length, with it either lying down or clipped short enough to stand erect. In the latter case, it is critical to examine the overall morphology of the equine to ensure that it is not an onager (Equus hemionus) or African wild ass (E. africanus), both of which naturally have an erect mane. In many other examples, the mane seems to be roached, or hogged, i.e. trimmed very close to the neck so that it does not show at all. The advantage of clipping the mane for an equestrian archer is obvious, since a long, flowing mane could easily get entangled in the bow and arrow when the cavalryman is poised to shoot while in motion. Likewise, for a chariot horse, a long mane might become entwined with the reins. Examples of modifying manes can be seen in the art of many earlier or contemporaneous cultures, including Egyptian, Assyrian, Persian, Chinese, and Etruscan. Modern polo ponies have their manes roached to avoid tangling with the mallet.

  • 37 Olsen & Bryant, 2013.

48Tails were also managed in different ways. Banging, or cutting the tip straight across the bottom, is shown on figures of horses with particularly full tails (see examples below). Several tails are shown as slender lines that then flare out to an array of multiple short lines (see examples below). Some scholars have mistaken this tasseled tail as an indication that the equine illustrated was an ass or onager, but the overall conformation of the animals in the Arabian rock art discussed here normally makes it possible to distinguish between horses and asses or onagers. Zebras, asses, and onagers all have short hair on the sides of the dock, with long hairs emanating just from the lower part of the dock, creating a tassel that begins about halfway down the tail. Mules and hinnies have a tail like a horse, that is, nearly all the hairs are long throughout, instead of having a tassel at the end. Images of onagers or African wild asses do occur in Arabian petroglyphs, but these are interpreted as earlier, probably Neolithic, and their body conformations make it clear that they are not horses.37

49There are a few viable explanations of the “tasseled” tail on true horses seen in several Arabian petroglyphs: the tail may be wrapped in a tail bandage, the tail hairs may be plaited from the root of the dock down to about half the length of the tail, the dock may be clipped on the sides, or hairs may be pulled out along the sides of the dock, all of which can leave the skirt full. Plaiting of the dock is used today in field hunters, show hunters, and in equitation, but not dressage because performance can be restricted if the braid is too tight. Tails are plaited in polo ponies to avoid tangling with a mallet, and it is easy to see how this would also be advantageous in battle when long lances or other weapons are used by cavalrymen. Plaiting would also be beneficial for chariot horses to prevent entanglement of the tail with the harness. One example of a petroglyph, at ad-Ḍabiyah 1, near al-Shamālī, Ḥāʾil Province (Fig. 7) shows a full tail, but three concentric curved stripes near the dock may represent a tail bandage or wrap. The appearance of the tail in many petroglyphs compares most closely with modern horses’ tails that have had hairs pulled or clipped along the dock, especially on either side. This gives the illusion of a more rounded croup, which is definitely accentuated in the horse figures.

Fig. 7: Horse with a possible tail bandage, ad-Dabiyah 1, near Ḥāʾil.
Photo by Richard T. Bryant.

50[Image non convertie]

Horse Figures by Site

51Outside Egyptian New Dynasty tomb art, the most realistic depictions of ancient Arabian horses can be found in northern Saudi Arabia, in the modern provinces of Tabūk, Ḥāʾil, and northern al-Madīnah. The horse images from the sites in this study are presented below with descriptions elucidating features of each.

Taymāʾ Region

52Al-Naṣlah 1, located 2.3 km northeast of Taymāʾ oasis, is a beautiful natural geological feature, or mushroom rock, that has a depiction of a man standing beside an Arabian horse (Fig. 5). The formation consists of two boulders on short, slender pedestals separated only by a straight, narrow crevice. The two rocks have a large flat surface on one side, but are rounded and irregular on the other surfaces. The panel on the right boulder has a central figure of an Arabian horse, done in the traditional North Arabian style, with a man holding its lead in one hand and an arrow in the other. The horse has a halter depicted on its face, a prominent forelock, probably a clipped mane, and a full tail. The entire body, except for the face, is shaded, which may mean the horse was a solid color naturally. The man seems to be wearing a futah (men’s cloth wrap covering the lower part of the body). The oldest inscriptions, probably made at the same time as the horse and man, occur below the horse. Another group of ancient characters appears encircled above and to the right and claims the drawing. Newer individual characters, wusūm, are found between the horse’s legs and behind it. As mentioned above, this finely executed scene probably illustrates the act of training a horse, using the arrow to gain its attention.

53Al-Naṣlah 2 is a broad dome located a short distance from al-Naṣlah 1. When approaching this massive outcrop, one is struck by the beauty of the sandstone and the surrounding desert. The sand seems to sweep up around it like waves along the seashore. A delicate, but large Arabian horse is carved in the center of the vertical stone face (Fig. 4), and smaller camels and horses are portrayed in a separate grouping on the far right (Fig. 3). The main horse figure faces left and is drawn in the typical style for Northern Arabian horses dating to the 1st millennium BCE to early 1st millennium CE, based on repeated examples of associated graffiti. Notable characteristics include the fact that the end of its rostrum has been shaded to depict a color pattern, that it is wearing a grazing muzzle, or that it has been painted with henna. It has a distinct forelock, but the mane appears to have been clipped quite short. The tail is thick and long, but has been “banged” to make the skirt level all the way across the bottom. There is a distinctive wasm, or brand, on its croup and an inscription in front of the horse at chest level.

54The top of this horse’s tail consists of a very faint preliminary line, and the outlines of three of the feet are open at the bottom. This figure appears to be unfinished, since it is often found that the artist will sketch out a light drawing of a figure to make sure the proportions are right before making deeper and broader lines for the final picture. With nomads, it is possible the artist found it necessary to move on to fresher pastures for his herd before finishing his drawing, particularly if the artwork was started near the end of his seasonal stay in the area. One query that emerges is whether artists often took multiple seasons or years to complete a scene, particularly the more complex ones, since these were not sedentary people and could not always afford to linger long enough to completely execute their works of art. Some images clearly take very little time to produce, while deeply incised, very large, or intricate scenes with numerous figures would require a greater investment of time and effort. When unfinished examples are recorded, it could mean that the work was interrupted and the artist upon returning to the area lost interest or perhaps never did return to the exact location again. Sometimes much later individuals came along and altered an illustration, renewed it by scraping inside the figure’s body or going over the outlines, drew a circle around it, added an inscription, supplemented it with additional figures, obliterated it, or overwrote with a new illustration. It is very common to find a palimpsest of figures and inscriptions from a wide range of time periods.

  • 38 Ibid., p. 72.

55On the far right side of the al-Naṣlah 2 outcrop is a complex scene with figures and inscriptions created at different times by different artists (Fig. 3). The largest of the camels and a small horse in this grouping are done in a manner consistent with the North Arabian style, including the biangular turning of the hindquarters. These may be contemporaneous with the large central figure and may even have been done by the same artist. The large camel, unusually, dons a horse’s tail that has shading on the lower half, as if dipped in henna. Whether the artist was confused or expressing a sense of humor by putting one species’ tail on another’s body is unknown. The small Arabian horse is completely filled in with shading. The man riding it is a stick figure consisting of a vertical line for his body topped by a round head and sprouting two horizontal lines in front for his arms. A long line stretches out from this horse’s head to the waist of a man sitting on the hump of another camel. Significantly, this creature is drawn in a different style: that which is seen at al-Shuwaymis, where a wādī in southwestern Ḥāʾil Province has produced some of the oldest, Neolithic images, as well as numerous later camels.38 The Shuwaymis style camels may reflect the visitation of people from that region. This makes the “interaction” between the Arabian horse and the Shuwaymis-style camel perplexing. The patinas are basically indistinguishable, so it is difficult to determine which came first or if they were contemporaneous. It is also not clear whether the long line represents a rope lasso or the horseman’s lance, especially since it emanates from his horse’s head, not the man’s arm.

56Al-Daksh is a broad dome-shaped outcrop with a smaller one in front. Both domes have a smooth, flat surface, with the most anterior one serving as a canvas for ancient artists. Located 42 km south of Taymāʾ Oasis, this panel has three elegant Arabian horses facing left and drawn with dramatic outlines (Fig. 6). The one on the left is more shallowly incised and appears unfinished. The central one is the largest, and the Thamudic B text in front of it claims the drawing. This encircled text and that on its body seem to have been incised with the same tool as this horse. The circle around the text in front and the first character inside the circle cut through the neck of the first -and earlier- horse on the left. The small horse on the right appears to be the newest, based on its patina. It was incised with a narrower tool than the one used to create the center horse. The similarities in terms of the style and strokes used to create all three horses are more than between these horses and others in the region, which would suggest that the three al-Daksh figures were made by the same artist. However, the text above the right hand horse claims that figure and is a different person’s name than the one associated with the central horse. The two-line Thamudic B text above the left hand horse does not refer to the horses. One can never be certain that the names of the people written adjacent to or on animal figures actually represent signatures of the true artists, and in many cases it is clear that they were added later. However, the differences in patina and tool width, plus the overwriting of the encircled text on the small horse on the left indicate that the creations of the three horses were not contemporaneous. Their strong similarities perhaps indicate a compulsion to conform to certain standards for drawing horses, as even the junctures of lines imply that a set order was followed in the process of outlining the animals.

57These three horses have the typical Arabian breed features of a concave facial profile, large eye, small ears, strongly arched neck, short topline, powerful hindquarters, and high tail carriage. The central figure and the one on the right clearly have a prominent forelock or perhaps a plume headdress, and the manes on all three horses are lacking, suggesting that they were roached. The tail of the small horse on the right and possibly that of the central horse have a shaded zone at the terminus most likely to indicate dipping their tails in henna.

58Both the central figure and the one on the right have an incised line running from just behind the ears down to the throatlatch. This may be part of a halter looped over the neck, since the larger figure also has a line that comes down from the neck to behind the horse’s elbow that is probably the lead. The six characters on its body were added later. Six small ostriches in a line are located between the first and middle horses.

59Jabal Ḥabīb (Fig. 8), also near Taymāʾ, has the outline of a single horse that faces left. The figure is so incredibly similar to those at al-Daksh that it could have been made by one of the artists from the other site. At the very least it conforms strongly to the same style. The facial features, as well as the lines of the body, feet, and tail are nearly identical to those shown at al-Daksh. The horse has a forelock or plume and a roached mane, just like those at al-Daksh. The artist cleverly positioned the head so that the pupil of the horse’s eye is indicated by a white pebble inclusion in the sandstone (Fig. 9). The tip of the tail has a line across it similar to those at al-Daksh, but it is not shaded. The demarcation still may indicate it was meant to represent a henna dipped tail. There is no evidence of a halter or bridle.

Fig. 8: Horse at Jabal Ḥabīb, near Taymāʾ, that closely resembles those at al-Daksh.
Photo by Richard T. Bryant.

60[Image non convertie]

Fig. 9: Close up of Jabal Ḥabīb horse’s head, showing white pebble eye.
Photo by Richard T. Bryant.

61[Image non convertie]

Fig. 10: Horse wearing a halter and lead at Jabalah, near Taymāʾ.
Photo by Richard T. Bryant.

62[Image non convertie]

Fig. 11: Horse with a “renewed” rider hunting an oryx at Jabalah, near Taymāʾ.
Photo by Richard T. Bryant.

63[Image non convertie]

64Jabalah is a tall, broad sandstone outcrop 53 km south of Taymāʾ with a huge panel consisting of four large camels and numerous smaller ones, three Arabian horses, a cavalryman on horseback, an ostrich, and an ibex. One horse faces left (Fig. 10), has a very concave face and a prominent forelock or plume on its forehead. No mane is depicted, so it may have been roached. The tail is full from the dock to the tip and may have been banged at the end. It is wearing a halter with a long lead. There is a Thamudic B inscription in front of the horse that claims it. The second Arabian horse also faces left, has a short clipped mane, a full tail, and a line from behind the ears to the throatlatch perhaps representing a halter. A third horse faces right, has a very concave facial profile, a thick mane, and a tail that appears to be braided, pulled or clipped from the root to about a third of the length of the tail before it spreads out (Fig. 11). This may be a juvenile based on its small face and brushy mane. A tiny rider, added later, was drawn just as a vertical line with a horizontal line probably representing one of the arms behind. Another long line that could represent a lance runs from the rider to a stick-figure oryx, also younger than the original horse image.

65Jabalah 2 consists of a later image of a cavalryman on a horse, facing right (Fig. 12). The figure is quite distinct from the earlier North Arabian style horses in its straight facial profile, long ears, straight neck, bent forelimbs, medium high tail carriage and the way the hind limbs emerge from the body. Based on these features, this particular equine may be a type other than an Arabian horse and may have been introduced from outside this region. The man has a very long lance in one hand and the bridle’s reins and a curved sword in his other. His legs dangle straight down, with no indication of stirrups.

Fig. 12: Later lance-yielding cavalryman at Jabalah 2, near Taymāʾ.
Photo by Sandra L. Olsen.

66[Image non convertie]

67Uqula is a massive castle-shaped jabal 15 km west of Taymāʾ with spires on top. On the southeast side, it has a long, smooth, vertical face with several groups of figures scattered along it. The central panel (Fig. 13) has a large camel on either end, with a group of smaller figures in the middle, including a little ibex that is being attacked by dogs, a pair of possibly young horses, a couple of smaller camels, an ostrich, another tiny horse and a large adult horse. All of the animals face right, except for one dog. The graffiti below the pair of equines say “By Bġr are the two horses” (frs¹n). The two horses have long manes, but their tails appear to have been plaited, pulled or clipped for most of their length so the tail skirts flair only near the tip. Both horses have shading on the forelimbs all the way up to the elbows, probably representing henna. The larger Arabian horse has no visible mane, so it may have been roached, and the tail has a tassel at the end as though it was plaited, pulled or clipped down most of the length of it, as well. It has no shading on the legs. The powerful arched neck on this figure suggests it was an adult.

Fig. 13: Complex scene with numerous animals, including two possible juvenile horses and an adult horse at Uqula, near Taymāʾ. Photo by Richard T. Bryant.

68[Image non convertie]

69Jabū is a jabal located 47 km southwest of Taymāʾ with a broad vertical surface decorated with finely executed animal scenes. Miscellaneous smaller figures are sprinkled on cliffs and boulders around the corner from the main panel. Two Arabian horses face left, one behind the other, separated by just a couple meters (Fig. 14). Likely done by the same artist at the same time, the horses have a concave, unshaded face and distinct forelock. The manes, necks and all four legs are shaded, probably representing henna coloring. The forelimbs are shaded all the way up to the elbow, while the hind shading only goes up to the hock. The hooves are not shaded. The tail on the first horse is full and shaded, but the one on the other figure is a single line, as if it had been wrapped, plaited, pulled or clipped. A small circular object suspended midair between a hind leg and the tail of the first horse may indicate defecation. The second horse has a diminutive rider who is holding the reins of its bridle and has a long lance in the other hand. He is depicted with both legs on the side visible to the viewer, as if riding sidesaddle, but it may just be the artist’s desire to show both appendages. Behind the second horse is a bow case with the end of a bow protruding out. Normally, bow cases are attached to the rider, but this one appears to have just fallen off. A few meters farther to the right is an older outline of a lion, also facing left.

Fig. 14: A pair of horses probably decorated with henna at Jabū, near Taymāʾ. The first horse may be defecating and the second has a small rider who has apparently just dropped his bow case.
Photo by Richard T. Bryant.

70[Image non convertie]

71Ḥafīrat Bard, 55 km south of Taymāʾ, is an enormous cliff face with pronounced vertical streaks of alternating red and white caused by the flow of rain over its edge through the millennia. To gain access to the rock face, the artists had to ascend 2-3 m to a ledge on which they stood while creating their images. A series of figures from left to right includes: a group of six long-necked camels and a fat-tailed sheep, followed by an Arabian horse with a rider, another fat-tailed sheep, small cavalrymen, more camels, inscriptions, and a horse and chariot.

  • 39 Drews, 2004, p. 34.
  • 40 Littauer & Crouwel, 1979, fig. 38.

72The first Arabian horse with a rider faces left (Fig. 15). It has a full mane and a pulled, clipped or plaited tail, and all four legs are shaded. The rider, who was modified or renewed later based on his lighter patina, has two swords at his waist and his feet are dangling straight down below the horse’s belly. Curiously, both legs appear to be on the opposite side of the horse, as though the rider was mounted in a “sidesaddle” position, albeit without a saddle, similar to the one at Jabū. A few Syrian terracotta figurines from circa 2000 BCE also show their riders in this position.39 A seal impression from the Bronze Age Hittite city of Kültepe, Anatolia, from the same period clearly has a rider with both legs on the same side. The person is seated on a cushion of sorts on the equine’s croup, in other words, in the “donkey seat”.40 The horse figure illustrated at Ḥafīrat Bard is wearing a bridle and the rider is holding the reins. The fact that the reins and part of the person’s body have a patina similar to that of the outline of the horse suggests that the original image did include a rider. Whether the original rider carried swords or other weapons cannot be discerned because of the renovations of the rider’s image.

Fig. 15: Horse and cavalryman at Ḥafīrat Bard, near Taymāʾ.
Photo by Richard T. Bryant.

73[Image non convertie]

Fig. 16: Chariot with three men at Ḥafīrat Bard.
Photo by Richard T. Bryant.

74[Image non convertie]

  • 41 Khan, 2012, p. 52.
  • 42 Sacco, 2013, p. 210.
  • 43 Olsen & Culbertson, 2010, p. 112.
  • 44 Curtis & Tallis, 2012, p. 41.
  • 45 Israel Museum, Jerusalem, Accession number IAA 1978-1.
  • 46 Kawami & Olbrantz, 2013, p. 76.
  • 47 Calvert, 2013.
  • 48 Hawass, 2005, p. 196.
  • 49 Sacco, 2013, p. 208.
  • 50 Manassa, 2013, p. 146.
  • 51 Chard, 1937, p. 317.

75The Ḥafīrat Bard chariot (Fig. 16) is pulled by a horse that has shading on its face, neck, all four legs and tail. The tail was pulled, clipped or plaited from the root of the dock to about 1/3 of the way down. The horse has a net-like pattern on its neck, breast and trunk that probably indicates it is wearing some form of armor or caparison. Another example can be found at ad-Dabiyah,41 although that one is restricted to the horse’s neck. Some Assyrian bas-reliefs of horses, like ones pulling the chariot of Ashurbanipal from the Southwest Palace at Nineveh, show armors over the horse’s neck and chest that have a similar appearance to that in the Ḥafīrat Bard image.42 Another example from the British Museum is an openwork ivory panel carved as a horse drawn chariot with a driver and a fragment of a passenger (BM 132939) from Fort Shalmaneser, at Nimrud, Iraq, dating to the 9th-8th century BCE. The horse on this Syrian or Phoenician style piece has features reminiscent of the Arabian breed, with a delicate muzzle, erect neck and high tail carriage.43 The horse has a patterned trapper (i.e. caparison),44 encircling its girth from shoulder to loin. Similar, even earlier trappings are indicated on a gold plaque found in a Late Bronze Age Canaanite temple at Lachish, Israel, dating to the 13th century BCE.45 The plaque has a nude female, possibly the Semitic goddess Astarte, standing on the horse’s back.46 In that case, the armor covers the neck and whole trunk, including the croup. Finally, Egyptian depictions of pharaohs Thutmose IV (1401–1391 or 1397–1388 BCE),47 Tutankhamun (1332–1323 BCE),48 Ramesses II (1303–1213 BCE)49 and Ramesses III (1218–1155  BCE)50 also show caparisons on their chariot horses. Without actual examples, it is not always clear whether the horse trappings are purely ornamental or were sufficiently padded to provide real protection for the horse. Lightweight ones may have just served as sweat cloths to prevent the horse’s perspiration from reaching the chariot driver and passengers. The horse of Senenmut, an architect and government official under Egyptian 18th Dynasty Pharaoh Hatshepsut, was buried with a thin linen cloth with leather trim and straps to attach it to the horse’s back51 that was probably just a sweat cloth.

76The Ḥafīrat Bard chariot box holds three men, probably a driver and two warriors. The reins run from the horse’s head to the hand of the man on the left, who also holds a sword upright. A second line emanating from the horse’s head is problematic. It may also be a rein, but it turns downward at a right angle. Its point of origin seems too high to be one of the yoke traces. The man in the middle does not seem to be armed and may actually be the charioteer, even though the reins do not clearly extend over to his hand. This could simply be a problem with the small size of the image and lack of detail. The passenger on the right is wielding a bow, perhaps to defend against an infantryman carrying a sword and shield standing behind the chariot. The box of the chariot is made up of rows of smaller squares that represent its construction. The draft pole is indicated below the chariot box with the eight-spoked wheels coming off of it in a strange perspective that, like earlier schematic chariot images, shows both wheels facing toward the viewer. The wheel rims have segments within them that represent the felloes and possibly wheel clamps or wraps that held the spokes in place.

Al-‘Ulā Region

77Al-ʿUlā, in al-Madīnah Province, in the northwest, is a large fertile valley with a long history. Important as a way station on the incense trade route, during the 1st millennium BCE, it was the capital of the kingdom of Dedan, and then the home of the Liḥyān Empire. There are a few carvings in the sandstone cliffs lining the valley of small horse done in the same style as those around Taymāʾ. In one particularly clear example located near the town’s police station (Fig. 17), the horse has an entirely shaded body, a mane, a pulled or clipped tail, a bridle and a rider. Both of the rider’s legs are shown on the side facing the viewer. The drawing is claimed as a horse in the Thamudic B inscription to the right.

Fig. 17: Horse and rider image at al-ʿUlā.
Photo by Richard T. Bryant.

78[Image non convertie]

79Umm al-Rākib (Fig. 18) is an alcove with steep, parallel sides located near al-ʿUlā, that has a small petroglyph of two horses and riders, one following closely behind the other. Roughly scratched into the hard sandstone, their outlines are somewhat jagged. The men hold the reins in one hand and the first rider seems to have a riding crop in the other hand. The riders are seated with their leg extended fully and toe pointing forward. The horses appear to have some type of saddle blanket. The manes and forelocks are lightly indicated and both have full tail skirts. The first horse probably has a pulled, clipped or plaited dock.

Fig. 18: Horses and riders at Umm al-Rākib, near al-ʿUlā.
Photo by Richard T. Bryant.

80[Image non convertie]

81Qīʿān al-Ṣanīʿ (Fig. 19) located near al-ʿUlā, is significant in having a small panel with a chariot. The panel also includes two camels, a large lion, and inscriptions. The chariot is pulled by one visible equine, but the four reins suggest another. The drawing is less realistic than most and the horse is not drawn in a way that indicates any Arabian features. It is all shaded in and the back is elongated and flat. The neck is not arched and the facial profile is convex. The tail has been braided or wrapped to leave just a tassel at the end. The chariot box appears to have just the driver, and the wheel has around ten spokes, an exceptionally high number. The main camel is behind the chariot and is done in the Shuwaymis style. Its tail is elevated, an indicator that it is a female, since female camels raise their tail when in estrus. The lion is directly below the horse and chariot and is depicted as equal in size to them. It is clear that the creator of this image, which is lightly incised and probably done quickly, did not follow the North Arabian, or Taymāʾ, artistic style. If the camel was done at the same time by the same artist who drew the chariot, then the person may have been from around Shuwaymis and perhaps from a later time.

Fig. 19: Simple line drawing of a chariot at Qīʿān al-Ṣanīʿ, near al-ʿUlā.
Photo by Richard T. Bryant.

82[Image non convertie]

Ḥāʾil Region

83Ad-Ḍabiyah 1 (Fig. 7) is a remarkable locality near the modern town of al-Shamālī, 170 km southwest of the modern city of Ḥāʾil, in Ḥāʾil Province. Although this and another horse figure nearby at ad-Ḍabiyah 2 (Fig. 20) are drawn according to the North Arabian conventions for 1st millennium BCE-early 1st millennium CE Arabian horses, they are distinct in their detailed presentation of their riders. The horse at ad-Ḍabiyah 1 faces left, has a concave facial profile, and its forelock is shown. The mane is lying down, with individual strands that are the same length, perhaps braided and trimmed. The tail has a high carriage and is lush, but there are 3 curved, parallel lines near the dock that may represent taping of the tail with cloth bands. The two hind feet are shaded, perhaps indicating henna or leg wraps. The horse is wearing a simple bridle illustrated by a line from behind the poll down to its throatlatch and reins going from the muzzle to the rider’s hand. In his other hand, the man is holding a riding crop. The rider is seated on a saddle blanket in the middle of the horse’s back with his leg fully extended forward, without stirrups. The man has a distinctive hairstyle or helmet, and pecked dots on his body suggest that he is wearing scalar armor. A bow case projects behind him from his waistband or belt. From his appearance, the rider must have been a cavalryman. A brief inscription was placed directly in front of the horse, but it does not mention the drawing. Several camels of a different style and an ostrich are on the same panel. Another boulder just in front of the panel is covered with camels and the image of a strange sprawled body of a man.

Fig. 20: Horse and rider at ad-Dabiyah 2, near al-Shamālī, Ḥāʾil Province.
Photo by Richard T. Bryant.

84[Image non convertie]

85Ad-Ḍabiyah 2, roughly 100 m from ad-Ḍabiyah 1 is a similar scene of a man riding a horse. The horse and rider face left, as is typical, and the horse is drawn in the standard manner for an Arabian, with a concave face, large eye, small ears, short topline, level croup, and high tail carriage. As with the other image, there is a line running from behind the ears to the throatlatch, but this bridle has a noseband and chin strap on the muzzle with a rein coming off the bridle where the cheekpieces would be. The horse’s mane may be clipped to stand erect or lie down evenly and the tail has a full skirt. The mane, tail, and both hind legs up to the stifle are shaded. This shading of the legs, and perhaps even the mane and tail, may indicate the use of henna.

  • 52 Khan, 2013, p. 52.

86The stoutly built man has the reins in one hand, a riding crop in the other, and a bow case at his hip. Either his hairstyle or a helmet is shown, and there is a hint of armor on his chest and shoulders made by heavy pecking with the stone that made the image. He is positioned fairly far back on the horse, almost in a donkey seat, with bent knee on a saddle blanket with no indication of either a saddle or stirrups. There is an inscription in front of the horse and another above the horse’s croup, but neither mentions the horse or man. Nearby is another panel with two Arabian horses facing one another and wearing clothes on their necks.52

87Ḥiān ʿArnān (Fig. 21), also in Ḥāʾil Province, is located in a region sprinkled with delicate spires that give it an enchanted appearance. One face of this large sandstone block is densely packed with figures and inscriptions, many overwriting ones from earlier periods. The fact that it is one of the more complex petroglyph palimpsests reflects that this was clearly a very popular place along an ancient route. Animals on this panel include numerous ibexes, camels, lions, and stick figure warriors and cavalrymen. Beside a large outline of a camel is a well-executed Arabian horse, facing left. It has a notably large eye with the pupil and brow indicated. The full mane, all four limbs and tail are shaded. An X on the muzzle represents a bridle.

Fig. 21: Close-up of an Arabian horse in a large panel at Ḥiān ʿArnān, Ḥāʾil Province.
Photo by Sandra Olsen.

88[Image non convertie]

89A small panel on Jabal ʿArnān (Fig. 22), near the larger one at Ḥiān ʿArnān, has six horses, a camel, and some later stick figure quadrupeds. There are multiple inscriptions, apparently done by different individuals at different times. Three of the horses face left and three face right. Five of the equines are clearly Arabian horses, drawn more or less in the traditional North Arabian style. These have slight indications of manes in at least four cases, and tails that were pulled, clipped or plaited to a varying extent. One tail appears to have a full skirt but with segments that may indicate wrappings, while the other tails have been pulled, clipped or plaited. Four of the horses have riders illustrated by simple vertical lines rising off the horses’ backs. Four of the horses are identified as mares in the inscriptions. Only the horse at the bottom lacks an inscription.

Fig. 22: Small panel of seven horses at Jabal ʿArnān.
Photo by Richard T. Bryant.

90[Image non convertie]

91Al-Mismā (Fig. 23) is a small individual boulder that has one of the best depictions of a horse and rider. The entire body of the horse is shaded so that it appears to be all one color overall. It has a prominent forelock and a full, lush tail. The artist has cleverly left the bridle and reins unshaded across the face and neck of the horse, as well as the saddle blanket and the rider’s leg and hip. The effect is similar to a photographic negative in comparison to other horse petroglyphs. The rider has a beard and appears to be wearing a helmet and a cuirass, with perhaps a bow case and a sword projecting behind him. There is an ancient inscription in front of the horse, probably written contemporaneously with the figures. The first line gives a name, but the second line is illegible. Below and to the left are four much younger mounted stick figure cavalrymen wielding swords or short lances over their heads.

Fig. 23: Horse and rider on boulder at al-Mismā.
Photo by Richard T. Bryant.

92[Image non convertie]

93Jabal al-Yāṭib (Fig. 24), located 35 km east of the modern city of Ḥāʾil, is an impressive rocky hill covered with hundreds of images on its multifaceted surfaces. Perched near the top in a very precarious position is a small image of an Arabian horse with a rider pecked into the red sandstone with a small inscription below it. The equestrian is a simple figure with one hand on the reins and both legs bent and on the side facing the viewer. The tail is plaited so that it has a tassel at the end. Around the corner, a larger panel of figures scratched through a heavy layer of desert varnish shows camels done in the North Arabian style, alongside a lion, palm trees, and some stick figure horses and camels that are probably much later.

Fig. 24: Jabal al-Yāṭib horse and rider.
Photo by Sandra Olsen.

94[Image non convertie]


95In conclusion, there appears to be a North Arabian, 1st millennium BCE –early 1st millennium CE, artistic style for creating petroglyphs of Arabian horses and camels centered in Taymāʾ Oasis that spread outward as far as al-ʿUlā and Ḥāʾil. Several other conventions for depicting horses existed in different regions or time periods in the Arabian Peninsula, some even showing breeds other than the Arabian. Although the North Arabian style has numerous shared characteristics, it is possible to detect the work of different artists and, at the same time, to conclude that some images in a shared vicinity could have been made by one artist. During this peak period for the incense trade, it is likely that horses and camels, both critical means of transportation and valued property, would be heavily emphasized on caravan routes and in the environs of important oases. The images of horses elucidate the arrival of the Arabian breed into North Arabia and the high status it rapidly gained in society. The depictions of finely groomed animals, the probable application of henna, examples of training, and the fact that horses were the centerpieces of so many petroglyph panels all provide documentation for their significant social status. The military roles of 1st millennium BCE to early 1st millennium CE horses in chariot warfare and the cavalry are illustrated by the weaponry and armor of the men who drove the vehicles and rode the horses. Associated North Arabian inscriptions provide a basis for dating the images to this period. In closing, it is important to recognize that even seemingly simple outlines of animals in rock art have the potential to reveal an enormous amount of information about a culture, its traditions and the behavior of its members. Often the revelations are things that would otherwise not be preserved in the archaeological record.

Top of page


Amir P., Oriental Horses of Pakistan, Islamabad, Asianics Agro-Dev., 1997.

Blunt, Lady Anne, A Pilgrimage to Nejd, the Cradle of the Arab Race, Piscataway, NJ, Gorgias Press, 2002.

Boessneck J., “Ein altagyptisches Pferdeskelett”. Mitteilungen der Deutschen Archaologischen Institus Abteilung Kairo 26, p. 43-47, 1970.

Bökönyi S., “Two horse skeletons from the cemetery of Kurru, Northern Sudan”, Acta Archaeologica Academiae Scientiarum Hungaricae 45, p. 301-316, 1993.

Calvert A. M., “Vehicle of the Sun: The Royal Chariot in the New Kingdom”, in A. J. Veldmeijer & S. Ikram (eds.), Chasing Chariots: Proceedings of the First International Chariot Conference, Leiden, Sidestone Press, p. 45-71, 2013.

Chaix L., “An Hyksos horse from Tell Heboua (Sinai, Egypt)”, in M. Mashkour, A.M. Choyke, H. Buitenhuis, & F. Poplin (eds.), Archaeozoology of the Near East IVB: Proceedings of the Fourth International Symposium on the Archaeozoology of the Southwestern Asia and Adjacent Areas, Groningen, ARC Publicatie 32, p. 177-184, 2000.

Chard T., “An early horse skeleton”, Journal of Heredity 28, p. 317-319, 1937.

Chardin, Sir J., Sir John Chardin, Travels in Persia: 1673–1677. An abridged version of the author’s Voyages du chevalier Chardin en Perse, et autres lieux de l’Orient, edited by L.-M. Langlès, 1711. Original English version published by Argonaut press, 1927. Mineola, New York, Dover Press, 1988.

Cherlenok E. A., “The chariot in Bronze Age funerary rites of the Eurasian steppes”, in S. L. Olsen, S. Grant, A. M. Choyke & L. Bartosiewicz (eds.), Horses and Humans: The Evolution of Human-Equine Relationships, Oxford, BAR International Series 1560, p. 173-179, 2006.

Clutton-Brock J., “The Buhen Horse”, Journal of Archaeological Science 1(1), p. 89-100, 1974.

Curtis J. & Tallis N., The Horse: From Arabia to Royal Ascot, London, British Museum Press, 2012.

De Moor J. C., The Seasonal Pattern in the Ugaritic Myth of Balu, according to the Version of Ilimilku, (Alter Orient und Altes Testament), Kevelaer, Butzon and Bercke, 1971.

Drews R., Early Riders: The Beginnings of Mounted Warfare in Asia and Europe, New York, Routledge, 2004.

Driesch A. von den & Peters J., “Fruhe Pferde- und Maultierskelette aus Auaris (Tell el-Daba) ostliches Nildelta”, Agypten und Levant 11, p. 301-311, 2001.

Ducos P., “Le cheval de Soleb”, in M. S. Giorgini (ed.), Soleb II:Les Nécropoles, Firenze, Sansone, p. 260-265, 1971.

Eichmann R., Schaudig H.-P. & Hausleiter A., “Archaeology and Epigraphy at Tayma, Saudi Arabia”, Arabian Archaeology and Epigraphy 17, p. 163-176, 2006.

Fuller A., Buarij: Portrait of a Lebanese Muslim Village, Harvard Middle Eastern Monographs VI, Cambridge MA, Center for Middle Eastern Studies of Harvard University, Harvard University Press, p. 84-86, 1961.

Goldhahn J., “Bredaror on Kivik: A Monumental Cairn and the History of its Interpretation”, Antiquity 83(320), p. 359-371, 2009.

Hawass Z., Tutankhamun and the Golden Age of the Pharaohs, Washington DC, National Geographic Society, 2005.

Jacobson-Tepfer E., “The image of the wheeled vehicle in the Mongolian Altai: Instability and Ambiguity”, The Silk Road 10, p. 1-28, 2012.

Kawami T. S. & Olbrantz J., Breath of Heaven, Breath of Earth: Ancient Near Eastern Art from American Collections, Seattle WA, University of Washington Press, 2013.

Khan M., Wusum: The Tribal Symbols of Saudi Arabia, Riyadh, King Fahd National Library, 2000.

Khan M., The Arabian Horse: Origin, Development and History, Riyadh, King Fahd National Library, 2012.

Khan M., “Rock art of Saudi Arabia”, Arts 2, p. 447-475, 2013.

Linduff K. M. & Olsen S. L., “Saddles Discovered In Western China Zaghunluq (轧滚鲁克)(761 +61 BCE) and Subeshi (苏贝希)(480 +85 –230 +85 BCE) Xinjiang”, Chinese Journal of Eurasian Studies 1, p. 59-75, 2011.

Littauer M. A., “Rock Carvings of Chariots in Transcaucasia, Central Asia and Outer Mongolia”, Proceedings of the Prehistoric Society 43, p. 243-262, 1977.

Littauer M. A. & Crouwel J., Wheeled Vehicles and Ridden Animals in the Ancient Near East, Leiden, Brill, 1979.

Macdonald M. C. A., “Ancient North Arabian, I”, in R. D. Woodard (ed.), The Cambridge Encyclopedia of the World’s Ancient Languages, Cambridge, Cambridge University Press, 2004.

Macdonald M. C. A., “Wheels in a Land of Camels: Another Look at the Chariot in Arabia”, Arabian Archaeology and Epigraphy 20, p. 156-184, 2009.

Macdonald M. C. A., “Ancient Arabia and the written word”, in M. C. A. Macdonald (ed.), The Development of Arabic as a Written Language, Supplement to the Proceedings of the Seminar for Arabian Studies 40, Oxford, Archaeopress, p. 5-28, 2010.

Manassa C., “The Chariot that plunders foreign lands: “The Hymn to the King in His Chariot”, in A. J. Veldmeijer & S. Ikram (eds.), Chasing Chariots: Proceedings of the First International Chariot Conference, Leiden, Sidestone Press, p. 143-156, 2013.

Miczak M. A., Henna’s Secret History: The History, Mystery and Folklore of Henna, Lincoln NE, Writers Club Press, 2001.

Nehmé L., “Towards an Understanding of the Urban Space of Mada’in Saleh: Ancient Hegra through Epigraphic Evidence”, Proceedings of the Seminar for Arabian Studies 35, p. 155-75, 2005.

Olsen S. L., “Introduction”, in S. L. Olsen, S. Grant, A. M. Choyke & L. Bartosiewicz (eds.), Horses and Humans: The Evolution of Human-Equine Relationships, Oxford, BAR, International Series 1560, p. 81-113, 2006.

Olsen S. L. & Culbertson C., A Gift from the Desert: The Art, History, and Culture of the Arabian Horse, Lexington KY, International Museum of the Horse, Kentucky Horse Park, 2010.

Olsen S. L. & Bryant R. T., Stories in the Rocks: Exploring Saudi Arabian Rock Art, Text by S. L. Olsen, Photography by R. T. Bryant, Pittsburgh PA, Carnegie Museum of Natural History, 2013.

Sacco A., “Art and Imperial Ideology: Remarks on the Depiction of Royal Chariots on Wall Reliefs in New-Kingdom Egypt and the Neo-Assyrian Empire”, in A. J. Veldmeijer & S. Ikram (eds.), Chasing Chariots: Proceedings of the First International Chariot Conference, Leiden, Sidestone Press, p. 203-211, 2013.

Sadek M. H., Al-Aboud A. Z. & Ashmawy A. A., “Factor Analysis of Body Measurements in Arabian Horses”, Journal of Animal Breeding and Genetics 123, p. 369-377, 2006.

Stecher R. M., “Anatomical Variations of the Spine in the Horse”, Journal of Mammalogy 43(2), p. 205-219, 1962.

Weise M., A Reference Guide: Arabian Coat Coloration,, accessed 7/22/2016.

Top of page


1 Olsen & Culbertson, 2010.

2 Ibid.

3 Macdonald, 2004.

4 Boessneck, 1970; Bokonyi, 1993; Clutton-Brock, 1974; Chaix, 2000; Ducos, 1971; von den Driesch & Peters, 2001.

5 Blunt, 2002.

6 Weise, “A Reference Guide: Arabian Coat Coloration”.

7 Amir, 1997.

8 Sadek et al., 2006.

9 Stecher, 1962.

10 Goldhahn, 2009.

11 Jacobson-Tepfer, 2012.

12 Littauer, 1977.

13 Macdonald, 2009.

14 Cherlenok, 2006.

15 Macdonald, 2010.

16 Eichmann et al., 2006.

17 Macdonald 2010.

18 Ibid.

19 Ibid.

20 Nehmé, 2005, p. 171.

21 Linduff & Olsen, 2011.

22 Olsen, 2006.

23 Khan, 2000; 2013.

24 Personal communication, June 2016.

25 Miczak, 2001.

26 Harvard Museum of Art, object number 1958.76.

27 Miczak, 2001.

28 Ibid.

29 De Moor, 1971.

30 Miczak, 2001.

31 Chardin, 1988, p. 169.


33 BM object number 1974,0617,0.19.4.

34 Fuller, 1961, p. 84-86.

35 Ibid.: 84.

36 BM object number 1920,0917,0.3.

37 Olsen & Bryant, 2013.

38 Ibid., p. 72.

39 Drews, 2004, p. 34.

40 Littauer & Crouwel, 1979, fig. 38.

41 Khan, 2012, p. 52.

42 Sacco, 2013, p. 210.

43 Olsen & Culbertson, 2010, p. 112.

44 Curtis & Tallis, 2012, p. 41.

45 Israel Museum, Jerusalem, Accession number IAA 1978-1.

46 Kawami & Olbrantz, 2013, p. 76.

47 Calvert, 2013.

48 Hawass, 2005, p. 196.

49 Sacco, 2013, p. 208.

50 Manassa, 2013, p. 146.

51 Chard, 1937, p. 317.

52 Khan, 2013, p. 52.

Top of page


Electronic reference

Sandra L. Olsen, « Insight on the Ancient Arabian Horse from North Arabian Petroglyphs », Arabian Humanities [Online], 8 | 2017, Online since 05 November 2019, connection on 21 January 2020. URL : ; DOI : 10.4000/cy.3282

Top of page

About the author

Sandra L. Olsen

Curator in Charge, Division of Archaeology, Biodiversity Institute, University of Kansas

Top of page


Licence Creative Commons
Ce document est mis à disposition selon les termes de la Licence Creative Commons Attribution - Partage dans les Mêmes Conditions 4.0 International.

Top of page
  • Logo Centre français d’Archéologie et de Sciences Sociales
  • Logo Institut de recherches et d'études sur le monde arabe et musulman
  • Logo Ministère des Affaires étrangères et européennes
  • Logo CNRS - Institut des sciences humaines et sociales
  • Logo DOAJ - Directory of Open Access Journals
  • OpenEdition Journals