The Case of Guelta Oukas, Morocco

Print Friendly, PDF & Email


Very recently several petroglyphs at the rock art site of Guelta Oukas in the Anti Atlas Mountains of southern Morocco were severely damaged. However, the mutilation was limited to two panels with mainly depictions of cattle and – moreover – to specific body parts of those zoomorphic images. In this paper I argue that this is not just another case of unwanted vandalism. Instead, I propose that the mutilation at Guelta Oukas could represent an instance of ‘negative’ rock art, involving the desecration of the images.

by Maarten van Hoek

TRACCE index

TRACCE Online Rock Art Bulletin #37, —- 2015

The Case of Guelta Oukas

Desecrated Rock Art Panels in the Anti Atlas of Morocco

 By Maarten van



Any suitable rock panel can be chosen to manufacture rock art upon. The creation of rock art on a rock panel usually meant – especially in prehistoric times – that the panel, the site and the surrounding landscape became sacred. Very sporadically it happened that the rock art panel was desecrated by people of a subsequent culture. However, in my opinion the term ‘desecration’ is rather ambiguous and needs further refinement and explanation.

Let us assume that indeed – in the mind of the manufacturer – the creation of the first rock art image ritually charges the rock, the site and the immediate environment with sacred power. This way the rock and the site are consecrated. That was the goal – or one of the goals – of the creator of the markings and/or images. Every subsequent image will further enhance this sanctification and will ritually charge the rock and the site even more. Then there are two general possibilities. The subsequent image completely respects the earlier image and is placed distinctly separate from the earlier image, or the subsequent image is (partially) superimposed upon the earlier image for whatever reason. In both cases there is no question of desecration, even if the subsequent images are manufactured by a much later, different culture that is unaware of the symbolic meaning of the earlier images. Theoretically the goal of the next manufacturer is still the same: the ritual and symbolic sanctification of the rock panel and the site.

Modification of earlier images – implying the changing of the layout – may be regarded as borderline cases. However, in most instances it cannot be decided if the modification by – for instance – a subsequent culture concerns an instance of disrespectful desecration or just an indifferent or even a respectful adaptation of an existing image. Some fully pecked images may have been altered in such a way that a different image is created. A possible example is the image of a purported ‘hunter’ on one of the petroglyph panels at the Poison Spider site in Utah, USA (Figures 1 and 2). This petroglyph may originally have been the image of a quadruped that has been altered into an anthropomorph in a hunting scene. Earlier I suggested a similar adaptation for a petroglyph at Foum Chenna (southern Morocco) where the image of a horse was possibly changed into a camel. In many other cases it cannot be decided if a petroglyph was fully pecked out originally, or that the originally outlined image – possibly containing interior details – was carefully obliterated. However, an already fully pecked image may also just have been re-pecked.

Click on any illustration in this article to get a larger picture (click the < back arrow [top left] or – in some cases – click on the illustration to return to the article). Click the hyperlinks (in underlined red) to go to other web pages, which will open in a new window (click the X icon [top right] to return to the article).

Figure 01

Figure 1. Petroglyph on a rock panel at the Poison Spider site, Utah, USA (Site 3 at Potash Road according to Castleton 1987). Photograph by Maarten van Hoek.

Figure 02

Figure 2. Drawing of the same petroglyph (left) and my personal interpretation of the same petroglyph separating the purported earlier zoomorphic petroglyph from the alleged later additions. Drawings by Maarten van Hoek.

What I also do not regard to be a form of desecration is the violation of the site through vandalism. Although the shameful addition of modern graffiti and other damage may be regarded by indigenous people as a form of desecration (and I fully agree with them), the goal of the vandal is not to desecrate the sacred nature of the site or to nullify the power of an image. In most cases the vandal is not even aware of the sacred character of the site and the – often hidden – symbolism of the images. And even if vandals know the value of rock art, some of them just do not mind robbing, damaging or destroying rock art. Also for those reasons I do not consider modern graffiti to be a form of rock art. For the same reason also the obliteration of for instance petroglyphs by domestic or ritual grinding activities is not considered to represent desecration. At the most it is an indication of indifference.

Therefore, in this work desecration is defined as the deliberate attempt to annihilate or use the power of a very specific rock art image by erasing the image or parts thereof. This means that frequently other images at the panel or other panels at the site will be left untouched. Although it is often impossible to prove, in this paper the instances of desecration directly aim at the annihilation or use of the original power of one image or several images at a rock art site. Thus, desecration is power-related. Vandalism is not.

In this paper I focus on petroglyphs only and more specifically on petroglyphs from the Anti Atlas mountain range in southern Morocco. However, also certain sites with rock paintings may be the target of desecration. One example is the heavily mutilated and vandalised (unfortunately ‘legitimate’ desecration may well trigger vandalism) rock painting at Vilacaurani in the High Siera of Arica in the extreme north of Chile, South America, where a row of anthropomorphs seems to have been mutilated especially in the genital areas, while several zoomorphs (mainly camelids) feature mutilated heads, bellies and legs.

Desecration of a (rock art) image may be triggered by political, economical, religious or even personal motives or a combination thereof. If the destruction of such religious icons is performed by a group of the same religion or culture, the violation may be regarded to be a form of iconoclasm, often also referring to aggressive statements or actions against a well-established status quo. However, in several cases the destruction of religious images is clearly performed by people with another religion or culture and then the term iconoclasm is not suitable. In most instances in rock art there is also no question of “damnatio memoriae”, as it is not the intention of the eraser to erase the memory of the depicted person or of the manufacturer of the rock art image. Moreover, although there are exceptions, names are generally lacking in rock art.

However, in ancient Egypt the greatest importance was attached to the preservation of a person’s name. Pharaohs who obliterated another Pharaoh’s name were thought to have destroyed the person, and it was believed that this effect extended beyond the grave. For that reason many cartouches carrying the name of a Pharaoh were obliterated in ancient Egypt (Figure 3). In fact, the names in the cartouches had been usurped – that is, they had been erased and re-carved with the names of a later Pharaoh. In addition, the sculptor often also reworked the facial features and bodily proportions of the Pharaoh. Many faces (Figure 4) and limbs, and sometimes complete bodies were thus erased and reworked. It is safe to claim – in ancient Egyptian iconography – that heads are more often the target of desecration than any other body part. Interestingly, in rock art desecration often aims at specific images and/or parts thereof and also in rock art the heads in anthropomorphic and zoomorphic imagery are more often found desecrated than any other part of the body. Yet, in some cases also the complete images were obliterated.

Figure 03

Figure 3. Desecrated cartouche on one of the walls of the temple at Karnak, Luxor, Egypt. Photograph by Maarten van Hoek.

Figure 04

Figure 4. Desecrated face of an image on one of the walls of the temple at Karnak, Luxor, Egypt. Photograph by Maarten van Hoek.


The Case of Guelta Oukas

The rock art site of Guelta Oukas, also referred to as Ukas, Oukas, Oukass, Jebel Ukas and Oukkas, was reported to the scientific world probably for the first time by Lieutenant Klug in 1939 and subsequently by George Simoneau (1977) who included the site into his national inventory (Blanc et al. 2003: 80). Also Susan Searight included a brief entry of the site (S-63 – Oukkas) in her extensive study about Moroccan rock art (2001: 305). However, it was not until 2003 that a comprehensive survey of the site was carried out by Claude Blanc, Werner Pichler and Alain Rodrigue and for that reason I will use the name Guelta Oukas to refer to the site. Their publication contained two sketches of the location of the rock art panels (though not the location of the site in the larger geographical context) together with numerous illustrations comprising a few photos and many accurate black and white drawings of the petroglyphs (though not all).

To visually support this paper about the Guelta Oukas rock art site, I have uploaded a video onto YouTube that offers more graphical information about the site than is offered in this publication. The video is in French (please excuse any error) in order to reach especially people in northern Africa. You can watch the video using the graphical link below, or, alternatively, you may prefer to click the following link (YouTube) so that the video appears next to the TRACCE publication.


Erratum: at 3.40 min. in the video the correct term for “camping” should read “campement” in French.

My maps in the current paper/video reveal the exact location of Guelta Oukas in the south of Morocco (Figure 5). Although I know that some people object to revealing locations of rock art sites, there should – in my opinion (see Issue 3) – not be any problem publishing the exact location of Guelta Oukas this time, as (prior to 2015) Jacques Gandini already revealed the location of Guelta Oukas most accurately in his web site using an interactive Google Earth map.  In his web page Gandini also offers detailed information about how to reach the site and he also offers the exact co-ordinates of the site (29° 24,74’ N and 08° 51,48’ W).  Strangely, in his ‘Itinerary’ Gandini remarks that “Les cartes anciennes signalent une guelta que nous n’avons pu retrouver”, while the ‘guelta’ is definitely there; out of sight from the rock art site however.

Figure 05

Figure 5. Map of the location of Guelta Oukas in southern Morocco. Map by Maarten van Hoek, based on Google Earth Relief Maps (a more detailed map is included in my video).

Also a few earlier YouTube videos and web pages feature photos of some rock art panels at Guelta Oukas (Adrar24tv-2012  – Casa Trotter-2013  – Casa Trotter-2014 and François Le Bourg-2015). In Google Earth a few visitors have marked the site with photographs they uploaded. Only the photograph by Josimon is incorrectly positioned 2.75 km to the NNE of the site. The two identical photos by Delfina and Mario have been positioned very near the site. They also offer an interactive Google Earth map as well as the co-ordinates of the site (29° 24′ 45.46″ N  and   8° 51′ 27.20″ W).

In January 2010 Tony Kerr uploaded photographs of some rock art panels at Guelta Oukas and he kindly granted me permission to use his photos. In April 2014 Heribert Bechen uploaded 51 photographs of the Guelta Oukas site, all of which he shared high-resolution copies with me, simultaneously generously granting me permission to use all his material in my paper and video. Some of his photos also appear in Google Earth and he provided the following (GPS) coordinates for one of the rock art panels (Panel 7; see video): 29° 24′ 44.82″ N and 8° 51′ 28.36″ W.

This remote site is located on the left (east) bank of the Oued Tamanart; an intermittent river that flows SSW from the Anti Atlas mountain chain in southern Morocco to join the river Draa further south. As the area borders the Sahara (Figure 6), the river is mainly dry and for most of the time comprises a flat, dry river bed sparsely grown with low grasses and herbs, while at several places some low trees and bushes border the dry valley floor. At the spot of the rock art site the river is also called Assif N’Int (Figure 7). The valley floor is about 950 m O.D., while the tops of the hills that border the valley are just above 1000 to 1100 m O.D., causing the rather steep hill slopes to rise more than 100 m above the valley floor. At this point the valley floor has a width of 475 m from east to west. The hill slopes bordering the valley consist of beautifully folded sedimentary rocks of red and brown coloured strata (Figure 8).

Figure 06

Figure 6. Map of the location of Guelta Oukas in the Anti Atlas of southern Morocco. Map by Maarten van Hoek, based on Google Earth Relief Maps.

Figure 07

Figure 7. Map of the location of Guelta Oukas in the Oued Tamanart. 1: The rock art site; 2: the ‘guelta’; 3: the conspicuous isolated hillock in the Oued Tamanart. Map by Maarten van Hoek, based on Google Earth.

Figure 08

Figure 8. View from Guelta Oukas, looking west across Oued Tamanart (blue arrow) and across the isolated hillock towards the mountains of folded, colourful strata. Photograph by Tony Kerr (2010).

The rock art is found on a low ridge projecting SW from Jebel Oukas, which in turn is part of a long, NNE-SSW running, higher mountain ridge, called Tazougart. The site is located on the important north-south route through the Oued Tamanart. When coming from the north it is almost impossible to distinguish the low ridge against the background of dark sediments of Tazougart (Figure 9). However, travellers from the north knew that the spot could easily be identified by a small and isolated, cone-shaped hillock some 535 m NW of the rock art site. However, the main reason the rock art has been created at this spot is the fact that behind the ridge and some 275 m east of the southernmost point of the ridge, is a permanent water basin; the ‘guelta’ (Figure 10), which is still used nowadays by people travelling up and down the valley (Figure 11). From the south the site is more easily recognised, not only by the cone-shaped hillock, but also because of the more verdant vegetation that fringes the path to the ‘guelta’.

Figure 09

Figure 9. The rock art site of Guelta Oukas (green arrow) looking SE across the dry valley floor of the Oued Tamanart (blue arrow). The site is indistinguishable against the dark sediments of Tazougart. Photograph by Heribert Bechen (2014). The video offers further details regarding this picture.

Figure 10

Figure 10. The ‘guelta’ east of Guelta Oukas rock art site (2 in Figures 7 and 14). Photograph by Heribert Bechen (2014).

Figure 11

Figure 11. A Berber man returning from a visit to the ‘guelta’, looking NE. The rock art site is just out of sight. Photograph by Heribert Bechen (2014).

The importance of the site is further emphasised by several ancient structures. Claude Blanc, Werner Pichler and Alain Rodrigue (2003: 80) mention three ancient tombs just west of the ridge (each marked with + in Figure 7). Remarkably, they do not mention the large boat-shaped burial site in the secondary valley leading to the ‘guelta’, photographed by Heribert Bechen in 2014 (Figure 12). Moreover, Heribert Bechen also made photographs of a paved and walled structure near the southern tip of the ridge that could have been an ancient camping place (Figure 13). It was constructed a little higher up the west facing slope of the ridge and thus out of reach of possible torrents. But most significant is the rock art at this site where at least 178 petroglyphs have been recorded. Jacques Gandini mentions the presence of rock paintings ‘in a rock shelter’, but so far no rock paintings were recorded at this site or its immediate surroundings.

Figure 12

Figure 12. Boat-shaped burial site (feature 4 in Figure 14) in the secondary valley leading to the ‘guelta’. Unfortunately the tombs have been looted. Photograph by Heribert Bechen (2014).

Figure 13

Figure 13. Paved and walled structure (feature 5 in Figure 14) near the southern tip of the ridge at Guelta Oukas. Photograph by Heribert Bechen (2014).

Although there are numerous rock surfaces suitable for rock art manufacture on both sides of the path to the ‘guelta’ and around the basin, until now rock art was only reported on the ridge (Figures 14 and 15). Moreover most of the petroglyphs are concentrated on the higher part at the foot of two rock stacks (Figure 16). The smaller but more prominent of the rock stacks, called the “Grande frise du site principal” and the “point d’orgue” by Claude Blanc, Werner Pichler and Alain Rodrigue (2003: 81, 87) has the majority of the petroglyphs. Although several petroglyph panels have a view over the Oued Tamanart, the majority overlook the path to the ‘guelta’ (which itself is invisible from the rock art site), emphasising its importance.

Figure 14

Figure 14. Map of the rock art panels at Guelta Oukas. 1: The rock art site; 2: the ‘guelta’; 4: the ancient burial-site; 5: the purported ancient camping place. The approximated locations of the rock art panels (yellow squares) are based on Blanc et al. (2003). The two desecrated panels have been indicated with red squares. Map by Maarten van Hoek, based on Google Earth.

Figure 15

Figure 15. The rock art site at Guelta Oukas, looking NW towards the ‘Site Principal’ on top of the ridge. There are more rock art panels further south on the ridge (not shown in this photo). Photograph by Heribert Bechen (2014).

Figure 16

Figure 16. Map of the rock art panels at the ‘Site Principal’ at Guelta Oukas. The band of harder limestone is coloured brown. The approximated locations of the following rock art panels (yellow squares) are based on Blanc et al. (2003). 1: Panel 22-II; 2: Panels 29 – 30 – 31; 3: Panels 37 – 38; 4: Panel 7; 5: Panel 12; 6: Panel 3; 7: Panel 41. The two desecrated panels (1 and 6 in this map) have been indicated with green squares. Map by Maarten van Hoek, based on Google Earth.

Figure 17

Figure 17. Photo showing the location of several rock art panels at the ‘Site Principal’ at Guelta Oukas. 1: Panel 22-II; 2: Panels 29 – 30 – 31; 3: Panels 37 – 38; 4: Panel 7; 5: Panel 12 (out of sight); 6: Panel 3 (out of sight); 7: Panel 41. Photos with petroglyphs on several of these panels appear in the video. Photograph by Heribert Bechen (2014).

Most images at Guelta Oukas belong to the so called ‘Pecked Cattle Style’ and indeed, almost 50 % of the petroglyphs depict bovines. According to Susan Searight this style developed around 2500 B.C. and slightly later than the most distinctive Tazina Style that is characterised mainly by images of wild animals. She moreover suggests that the Tazina Style and the Pecked Cattle groups continued to be manufactured during the 2nd millennium B.C. (2001: 188). The Oued Tamanart is noted for its Pecked Cattle sites, but Susan Searight remarks that the southern part of this valley is one of the few areas where Tazina Style and Pecked Cattle petroglyphs occur together. That area seems to have been a contact point between two very different styles (2001: 160; Fig. 43).

Importantly, Claude Blanc, Werner Pichler and Alain Rodrigue (2003: 81) notice that there are many cases of superimposition (75 %), some of which are rather remarkable (Figure 18). This superimposition is also amazing as there are numerous untouched boulders with surfaces that are very much suitable for rock art production. This selectiveness points to a long use of the specifically selected petroglyph panels at this sacred place.

Figure 18

Figure 18. Superimposition of a large petroglyph of a goat onto bovine images on Panel 12 at Guelta Oukas. Photograph by Heribert Bechen (2014) with the superimposed goat added in yellow by Maarten van Hoek (2015).

The petroglyphs of quadrupeds are either unsexed, show male sex (for instance on Panel 4) or are of ambiguous sex. Also Blanc et al. recognise the problem of distinguishing the anatomical detail below the belly of bovines. It may either depict male sex or – especially when very large – an udder (2003: 83). Similar ambiguous petroglyphs of bovines occur in the High Atlas for which Alain Rodrigue remarks (1999: 60): “l’étroitesse de la poche sous-ventrale ressemblant dans ces cas précis à un scrotum beaucoup plus qu’à des pis”.

Unambiguous representations of female animals do not seem to occur at Guelta Oukas. Yet, according to Blanc et al., one pregnant (?) bovine on Panel 12 seems to have a smaller bovine petroglyph – a ‘foetus’ – at the belly area. However, a detail photo made by Heribert Bechen (Figure 19) does not show the alleged ‘foetus’ as clearly as the sketch by Blanc et al. (2003: 93). In my opinion it is just one of the several unpecked areas that – especially in bovine imagery – often have been left untouched (on purpose?) by the manufacturer of the rock art. Moreover, as the sexed body part of the bovine in Panel 12 (Figure 19) is not very large, this bovine may – in my opinion – equally be male.

Figure 19

Figure 19. Detail of the belly area of a bovine petroglyph on Panel 12 at Guelta Oukas, showing the purported ‘foetus’ (green arrow). Photograph by Heribert Bechen (2014).

The Desecrated Panels

In this paper the focus is on the recent damage of the bovine petroglyphs on Panel 3 and on Panel 22-II, but it is unclear whether the person, who exclusively damaged those two panels, aimed at mutilating images that he possibly interpreted as only female bovines. Most likely this did not matter.

Panel 3: Panel 3 (indicated as feature 6 in Figures 16 and 17) is found on the NW facing wall of the major rock stack, which is in fact the western end of the thick, inclined layer of dolomitic limestone (see Figures 15, 16 and 17). Blanc et al. recorded three petroglyphs of bovines on this panel (2003: 91 – Fig. 3); in my opinion all three possibly male (Figure 20). At the time of publication of their survey all three petroglyphs were undamaged. Also Tony Kerr photographed the left (north) facing bovine in a still undamaged condition in 2010 (Figure 21). Importantly, Claude Athané photographed the whole panel (Figure 22) in an undamaged condition in December 2012 (Bernard Fardet and Claude Athané of Casa Trotter 2015: pers. comm.). However, in April 2014 Heribert Bechen made a photographic record of Panel 3 that clearly showed that all three bovines of Panel 3 had been mutilated (Figures 23 and 24). Based on all the information available, it proves that the mutilation took place between December 2012 and April 2014.

Figure 20

Figure 20. Drawing of the petroglyphs on Panel 3. Based on Fig. 3 in Blanc et al. (2003).

Figure 21

Figure 21. Photo of a still undamaged bovine petroglyph on Panel 3 at Guelta Oukas. Photograph by Tony Kerr (2010).

Figure 22

Figure 22. Photo of the complete and still undamaged Panel 3 at Guelta Oukas. Photograph by Claude Athané (2012).

Figure 23

Figure 23. The desecrated bovine petroglyph on Panel 3 at Guelta Oukas. Photograph by Heribert Bechen (2014).

Figure 24

Figure 24. Two desecrated bovine petroglyphs on Panel 3 at Guelta Oukas. Photograph by Heribert Bechen (2014).

Panel 22-II: The same is true for the petroglyphs on Panel 22-II (indicated as feature 1 in Figures 16 and 17), which features a large collection of bovine petroglyphs (Figures 25 and 26); possibly all involving male bovines (Blanc et al. 2003: Photo 3, 96 – Fig. 22-II). Significantly, this panel is located high up the SE facing part of the smaller rock stack and out of sight from the valley of the Oued Tamanart. In fact the panel is even out of reach when standing in front of the rock stack. Indeed, Blanc et al. confirm that the petroglyphs “ont été réalisés en hauteur et sont relativement inaccessibles”. Only when dangerously standing on a small ledge of rock it is possible to touch the petroglyphs. Yet, most of the petroglyphs on this important panel were mutilated between December 2012 and April 2014 (Figures 27 and 28), most likely by the same person who damaged Panel 3 judging by the analogous brightness of the patination and the similar style and technique of the mutilation.

Figure 25

Figure 25. Drawing of the petroglyphs on Panel 22-II. Based on Fig. 22-II in Blanc et al. (2003).

Figure 26

Figure 26. Photo of the complete and still undamaged Panel 22-II at Guelta Oukas. Photograph by Tony Kerr (2010).

Figure 27

Figure 27. The desecrated bovine petroglyphs on Panel 22-II at Guelta Oukas. Photograph by Heribert Bechen (2014).

Figure 28

Figure 28. Detail of a desecrated bovine petroglyph on Panel 22-II at Guelta Oukas. Photograph by Heribert Bechen (2014).

Indeed, both Panel 3 and Panel 22-II show the same kind of damage and thus it is likely that one person decided to climb to the highest point of the rock art site and decided to mutilate only those two panels. He (or – less likely – she) most likely will have passed several other petroglyph panels that were left untouched. Moreover, the panels that were chosen are hard to reach; especially Panel 22-II. This means that the person, who mutilated the panels, deliberately chose those two panels, ignoring many other panels with petroglyphs of bovines that are easier of access.

It moreover seems that it was a man with a specific goal. Although at first sight the damage seems to be randomly executed, it is interesting to notice that the specific body parts are preferred over random mutilation. Especially the heads have been the target, as well as the genital areas and the tail-areas of the animals. As most petroglyphs are damaged in this way, instead of randomly obliterating one or more animal, it seems that the damage represents some kind of desecration. This desecration most likely was triggered by personal circumstances, probably of economic or religious nature, although – without informed knowledge – one can only guess.

But it is also very likely that the desecration had something to do with cattle. Perhaps a bovine-related disease or some other personal disaster triggered this specific person to mutilate those petroglyph panels. Therefore it is possible that the mutilation of Panels 3 and 22-II at Guelta Oukas is not an instance of pure vandalism. It may be a form of ‘negative’ rock art, especially if indeed religious or cultural motives have triggered the mutilation. It is therefore interesting to compare the desecration at Guelta Oukas with some other instances of mutilation, both in Moroccan and in global context.

The Case of Tanoumrit: The Tazina Style site of Tanoumrit is found 350 km to the ENE of Guelta Oukas (location shown on a map in the video). Susan Searight includes an entry of this site (SE-23; Tanoumrit S, Tazzarine) in her study of Moroccan rock art (2001: 297) in which she mentions a ‘frieze of equid-ostrich-rhinoceros (light patina)’. This frieze concerns three vertically arranged petroglyphs (all depicted in outline) that are described by André Simoneau as an ‘association d’un rhinocéros, d’une antilope et d’une autruche’ (1976: Fig. 20). Although the photograph by André Simoneau (Figure 29) is not that clear, it seems that the panel was undamaged before 1976.

Figure 27

Figure 29. Frieze of three zoomorphic petroglyphs at Tanoumrit. Photograph (may have been flipped horizontally) by André Simoneau (1976: Fig. 20).

However, a photograph of the same panel uploaded onto the internet showed considerable damage to all three petroglyphs (Figure 30). Because of the depth of the original Tazina Style petroglyphs the images still survived the mutilation. This photo (Maroc gravures 3) was taken by Antoine Sigogneau in 2011. He kindly sent me a high-resolution copy of the photo and simultaneously granted me permission to use this photo. It proves that only the frieze of animals was the target of hacking. The damage is rather patinated, which may point to a rather early mutilation (shortly after 1976?). At first sight the mutilation is randomly executed, but especially the head and the neck of the antelope (equid?) seem to have been the target (Figure 31). Again this may have been done by a local for personal reasons, possibly of religious or cultural nature.

Figure 28

Figure 30. The same frieze of the three zoomorphic petroglyphs at Tanoumrit; now harshly damaged. Photograph by Antoine Sigogneau (2011).

Figure 29

Figure 31. Detail of the head of the uppermost zoomorphic petroglyph at Tanoumrit. Photograph by Antoine Sigogneau (2011).

At nearby Aït Ouazik South a rock panel shows the possible obliteration of three zoomorphic (?) images that are hardly recognisable because of the many deeply patinated scratches (Figure 32). Obliteration also seems to have occurred further west at Foum Chenna where several petroglyphs of zoomorphs seem to have been pecked all over (Van Hoek 2014: Fig. 21).

Figure 30

Figure 32. Deeply patinated petroglyphs of possibly three zoomorphs that have been mutilated in ancient times. Aït Ouazik South. Photograph by Terry Little of TARA.

The Case of Tibaksoutine: Ahmed Taoufik Zainabi (2004: 38) has published a photograph of a petroglyph that – in his opinion – depicts a horse galloping (cheval au gallop). In the caption he labels the site as Tibaksoutine, a rock art site SE of Zagora (location shown on a map in the video), which is probably the same site as referred to by Susan Searight as SE-32; Jbel Tibasksoutine, Zagora (2001: 298). This outlined petroglyph may belong to the Tazina Style, but this is not certain, because Tazina Style petroglyphs usually concern wild animals. Moreover the style of this ‘horse’ petroglyph is slightly different. Interestingly, Tazina Style petroglyphs of animals are often associated with (mainly randomly executed) straight (polished) grooves. At first sight this also seems to be the case with the horse petroglyph. However, at Tazina Style zoomorphs those grooves hardly ever cross the animal completely. In this case however, the horse petroglyph is crossed by no less than nine vertical grooves (Figure 33). They systematically seem to cut the whole animal in pieces and especially the two crossing the neck seem to have been intended to (ritually?) kill the animal. In my opinion this instance may represent an example of desecration or even a symbolic, ritual sacrifice.

Figure 31

Figure 33. Petroglyph of a possibly intentionally sectioned zoomorph at Tibaksoutine, southern Morocco. Drawing by Maarten van Hoek, based on a photograph by Ahmed Taoufik Zainabi (2004: 38).

The Case of Toro Muerto, Peru: The horse at Tibaksoutine seems to have symbolically been cut in pieces. Grooves similarly ‘cutting’ the necks of zoomorphs have been reported from Toro Muerto; an enormous petroglyph complex of about 5000 m in length, located in the desert of southern Peru. A three meter long bicephalic snake on Panel Bc-016C has a short groove that seems to be cutting through the neck of the animal (Figure 34), while a crude groove seems to be decapitating the head of a feline on Panel Bc-003 (Figure 35), which – interestingly – is located only 80 metres to the SE of Panel Bc-016C. These ‘cutting’ grooves seem to have been executed at a (much?) later stage. This is probably also the case with the petroglyph of a llama on Boulder CAL-033 at Calaunsa, northern Chile, which has a scratched groove horizontally across the neck, possibly referring to the Andean practice of offering llamas at certain rituals.

Figure 32

Figure 34. Detail of the possibly decapitated head (green arrow) of the enormous bicephalic snake on Panel Bc-016C at Toro Muerto, southern Peru. Photograph by Maarten van Hoek.

Figure 33

Figure 35. Petroglyph of a possibly decapitated head (between the green arrows) of a feline on Panel Bc-003 at Toro Muerto, southern Peru. Photograph by Maarten van Hoek.

Interestingly, Toro Muerto has several instances of desecration that mainly concern the obliteration of faces of anthropomorphs and zoomorphs (Figure 36), although also a few instances occur where roughly the genital areas of zoomorphs have been mutilated (Figure 37). At Toro Muerto the obliteration of especially facial features of selected biomorphs was primarily focussed on the images of “dancers” and fantastic “snakes”; imagery that is suggested to convey fertility and water concerns (Van Hoek 2005, 2006), although other zoomorphs, such as felines and quadrupeds (mainly camelids) also are found with body parts erased.

Figure 34

Figure 36. Petroglyph of a desecrated zoomorph (green arrow) on Panel Ab-015 at Toro Muerto, southern Peru. Photograph by Maarten van Hoek.

Figure 35

Figure 37. Petroglyph of a zoomorph with mutilated rear end (green arrow) on Panel Da-061B at Toro Muerto, southern Peru. Photograph by Maarten van Hoek.

As there often is no question of random destruction of the entire image, this “positive” attitude may represent an indication of continuous respect for the symbolic value of the original image. Importantly, in many cases the obliteration appears to have been very carefully executed, and in most cases only the head area within the outlines of the image was the target of the obliterator. This may point to a rather high degree of intent. Thus, the careful removal of especially the facial details must have been most meaningful and might have conveyed a very specific emotion. This sentiment might be an articulation of discontent about the desired effects of the imagery. The instances of obliteration at Toro Muerto therefore might represent a second layer of symbolism idiosyncratically expressing the disappointment about the results of the imagery. To put it differently, to certain persons the once powerful symbols had “lost their face”.

The Case of Talabre, Chile: A case similar to the desecrated images at Toro Muerto is found at Talabre, a rock art site on the eastern edge of the Atacama Desert in northern Chile. In this case the desecration only concerns images of camelids. On large, vertical cliff faces in the gorge of the Quebrada de Quesala (also known as Kezala) are several panels with a large collection of rather small and very large outlined camelid petroglyphs. Several of the camelids show smaller and larger areas where very crude hacking deeply obliterated body parts (Figures 38 and 39). Especially the heads, the tails and the genital areas have been mutilated, although in some cases also the bellies show smaller and larger areas of hacking, as if images of foetuses have been erased. Importantly the hacked areas are almost as deeply patinated as the original petroglyphs which may point to a great age of the desecration.

Figure 36

Figure 38. Petroglyph of a large camelid desecrated at several spots (green arrows) at Talabre, Quebrada de Quesala, northern Chile. Drawing by Maarten van Hoek, based on a photograph by Lucas Burchard Señoret (2014).

Figure 37

Figure 39. Petroglyph of two camelids desecrated at several spots (green arrows) at Talabre, Quebrada de Quesala, northern Chile. The smaller camelid at the belly area just possibly might represent a ‘foetus’. Drawing by Maarten van Hoek, based on a photograph by Lucas Burchard Señoret (2014).

Although there are many sites with numerous petroglyphs of camelids in the Atacama, I am not aware of comparable instances east of the Andes. Only in the area around Antofagasta de la Sierra, on the Altiplano of west Argentina, are some petroglyphs camelids that may have been mutilated in the same way and possibly for the same reason (Aschero 1999: 110, 111, 119 and 120).

The Case of the Procession Panel, Utah, USA: Much further north, in North America, Ekkehart Malotki and Donald Weaver specifically describe a number of zoomorphic petroglyphs on the so called Procession Panel in southern Utah, USA (2002: 72, Plate 73). They write: An intriguing aspect of the animals is the obvious punctuation of their bodies in very specific areas (head, heart area, extremities). This intentional act, obviously done at a much later time as is evident from the relatively light varnish on the damaged spots, may have been motivated by the idea of utilizing the power inherent in the animal fragments to obtain favourable outcomes for hunting- or ritual-related activities similar to those depicted on the panel. In fact, rock flour from the engravings may actually have been ingested by the ritual participants, an act referred to as geophagy, that is, the “consumption of earth (2002: 72). Several more zoomorphic and even some anthropomorphic petroglyphs on this extensive panel have similarly been desecrated (Figure 40). Those instances of partial desecration, apparently unique in this area, have many parallels with the Andean examples, but can also be compared with the Guelta Oukas case, especially as several zoomorphic petroglyphs of the Procession Panel also have been desecrated at their tail-ends.

Figure 38

Figure 40. Part of the Procession Panel in southern Utah, USA, showing the specifically selected areas of desecration (green arrows) of zoomorphic images. Drawing by Maarten van Hoek, based on a photograph by Matt McGrath (2009).

The Case of Butler Wash, Utah, USA: Another well known instance of desecration is the aptly called Desecration Panel on the San Juan River near Bluff, southern Utah, USA (Figure 41). Several ancient Basketmaker Anasazi petroglyphs of the Butler Wash to the Narrows – Site 2 were obliterated almost completely by a Navajo family for religious reasons (Castleton 1987: 228; Figs 7.58 and 7.59). Dennis Slifer (2000: 123 – 124; Fig. 113) writes the following about this practice: “This is an interesting example of rock art influenced by shamanic concepts related to the role of power in healing rituals. Power can have either positive or negative applications, such as witchcraft or sorcery. During the 1950s, this panel of predominantly Basketmaker petroglyphs was ritually desecrated by Navajos living nearby, after a Navajo witch was implicated in causing illness to others by utilizing ‘power’ taken from the rock art. This is believed to have been accomplished by scraping away some of the sandstone from the images, grinding it into a powder, and blowing into the flesh of victims with a small tube. The curing rituals which followed determined that harmony and healing could be achieved by destroying the ‘power’ of the images in the rock art. Certain figures in this panel, primarily the anthropomorphic images suggestive of Basketmaker shamans, were therefore ritually obliterated by chopping at specific parts of the petroglyphs with metal tools such as chisels or axes”.

Figure 39Figure 41. Two of the desecrated anthropomorphic petroglyphs at the Desecration Panel on the San Juan River near Bluff, southern Utah, USA. Drawing by Maarten van Hoek, based on a photograph by Kenneth Castleton (1987: Fig. 7.58).



Although there are only very few ethnographic accounts explaining desecration, the example from Butler Wash proves that often desecration is frequently executed by one or just a few individuals. It seems likely that in most cases religious and/or cultural motives explain instances desecration, which aims at (ritually?) destroying or utilising the power of the rock art images. This practice is what I have labelled ‘negative’ rock art in this paper.

There may be many reasons to desecrate a rock art image. I suggest that the examples from the Andes indicate that desecration may have been performed out of discontent or disappointment about the desired effects of the original imagery. Although there is no ethnographic evidence to prove those Andean cases, this theory may explain – in general – why instances of desecration are often limited to only one site or even one or two rock art panels, like at Guelta Oukas. In some cases only specific images on one panel have been obliterated. An instance, in my opinion incorrectly labelled as vandalism by the photographer, is found at Tuba City, Utah, USA, where apparently specific Hopi Clan symbols have been erased, leaving the others untouched. Only earlier photographs can reveal which symbols were deleted, which again emphasises the enormous importance of making photographic records of rock art by both professionals and non-professionals.

The fact that in many cases specific body parts are the target of the desecrator proves that he or she had unambiguous motives. In very many cases the head, the male genital area and/or the tail area (the genital area of female quadrupeds) were the target, which may point to fertility related motives. The mutilation is often crudely executed (as if in resentment), like at Guelta Oukas, but in some cases it has been executed very carefully, only carefully erasing the face, Fine examples are the obliterated heads of snakes at Toro Muerto (Figure 42). This practice should not be confused with the technique of careful polishing of bodies or body parts, like the petroglyphs of ostriches and quadrupeds that are found on a rock wall of about 15 m long at Oued Mader (see photos by Antoine Sigogneau), north of Figuig in southern Morocco. In many cases their heads, tails and legs have been very carefully polished leaving the remainder untouched. This practice almost certainly does not involve desecration. It more likely is part of a specific style of petroglyph production.

Figure 40

Figure 42. Petroglyph of a carefully obliterated snake head on Panel Bc-006 at Toro Muerto, southern Peru. Photograph by Maarten van Hoek.

 Even though the mutilation of the two petroglyph panels at Guelta Oukas is of a very recent date, the fact that the mutilation mainly is focussed on the heads and genital areas probably points to intentional desecration. The fact that at Guelta Oukas only two panels and apparently only bovine petroglyphs were selected and consequently desecrated in a rather crude way, may indicate that (personal) discontent or disappointment, possibly about a cattle-related situation, formed the reason to desecrate those images.

Globally, vandalism at and destruction of rock art sites is increasing. Unfortunately the desecration at Guelta Oukas may also trigger vandalism at the site (and at other Saharan rock art sites), which should be prevented at all cost. Personally I very much regret any damage to rock art and rock art sites and even though the mutilation at Guelta Oukas may be regarded as a form of ‘negative’ rock art, I strongly advocate not touching ancient rock art in any way, not even for ‘legitimate’ personal, religious or cultural reasons.



This paper could never have been written without the help of several people, especially as I have not visited Guelta Oukas myself. In particular very helpful has been Heribert Bechen who provided me with many photographs of Guelta Oukas and the environment and kindly granted permission to re-produce them in this paper and in the accompanying YouTube video. In fact, the photos by Heribert Bechen triggered me to write this paper. Also Tony Kerr, Antoine Sigogneau, Terry Little of TARA – the Trust for African Rock Art – and Claude Athané kindly allowed me to use their photographs in this paper and/or in the video. All those photographs (most of which can be viewed by using the links in this paper) are their copyright.

Moreover, Heribert Bechen, Tony Kerr, and Claude Athané and Bernard Fardet (of Casa Trotter) provided useful additional information that has been used in this paper. However, I would like to emphasise again that only I am responsible for all the (information on) the illustrations, hypotheses, observations and conclusions presented in this paper.

Please be aware that any URL included in this paper (functional as per July 2015) may be broken or even hacked. Please notify me if any broken link occurs.


Aschero, C. A. 1999. Rock Art of the Puna Desert and Northwestern Argentina, In: Rock Art in the Andes of Capricorn. Museo Chileno de Arte Precolumbino. Santiago de Chile.

Blanc, C., W. Pichler and A. Rodrigue. 2003. Le site rupestre de Guelta Oukas (Morocco). Almogaren XXXIV; pp. 79 – 111. (Institutum Canarium), Wien.

Castleton, K. B. 1987. Petroglyphs and Pictographs of Utah. Volume Two: The South, Central, West and Northwest. Utah Museum of Natural History. Salt Lake City.

Gandini, J. n.d. Le site du Jebel Oukas ou de l’assif Youmkat. In: Le Maroc avant l’histoire.

Malotki, E. and D. E. Weaver jr. 2002. Stone Chisel and Yucca Brush. Colorado Plateau Rock Art. Kiva Publishing. Walnut, California.

Rodrigue, A. 1999. L’art rupestre du Haut Atlas marocain. L’Harmattan. Paris.

Searight, S. 2001. The Prehistoric Rock Art of Morocco. A Study of its extension, environment and meaning. PhD. Thesis. Bournemouth University.

Simoneau, A. 1976. Les rhinocéros dans les gravures rupestres du Dra-Bani. In: Antiquités africaines. Vol. 10; pp. 7 – 31.

Simoneau, A. 1977. Catalogue des sites rupestres du Sud-Marocain. Ministere d’Etat charge des Affaires Culturelles, Rabat.

Slifer, D. 2000. Guide to Rock Art of the Utah Region. Sites with Public Access. Ancient City Press, Santa Fe, New Mexico.

Van Hoek, M. 2005. Toro Muerto, Peru: Possible Prehistoric Deletion of Petroglyph Details. Adoranten. Vol. 2005; pp. 73 – 80. Journal of the Scandinavian Prehistoric Society. Underslös, Sweden.

Van Hoek, M. 2006. Toro Muerto, Perú. Posibles alteraciones prehistóricas en detalles de petroglifos. In: Rupestreweb.

Van Hoek, M. 2014. The Felines of Foum Chenna. In: TRACCE – on line Rock Art Bulletin.

Zainabi, A. T. 2004. Gravures, Peintures et Tumuli d’un Énigmatique Passé (pp. 33 – 41). In: Trésors et Merveilles de la Vallée du Drâa. Eds: Almakari, A. et al.. Marsam, Rabat, Morocco.

TRACCE index

TRACCE Online Rock Art Bulletin #37, —- 2015

Comments are closed.