Rock Art, landscape and prehistoric settlement at the High Atlas

Rock Art, landscape and prehistoric settlement at the High Atlas (Morocco).

ARPA’S team1

1. Introduction

Landscape Archaeology aims at the perception, not only physical but emotional, that human beings have about the space they are moving on and at the way they inscribe on it their own culture and experiences of being in the World in the phenomenological sense (Heidegger 2003). So, there are as many landscapes as socioeconomic organizations, and they differ from each other depending on the pattern of economy, social structure and material and mental (rational) control of their physical surroundings. This project focuses on the rock art area of Oukaïmeden, a high valley 80 km south of Marrakech, located at 2630 m. above sea level (photo 1).

Historical records from 16th century AD onwards describe its exploitation as summer pasturage exclusively by two tribes proceeding from the villages of the Rerhaya valley to the West and the Ourika valley to the East (Mahdi 1999; Franchi 2005). The hypothesis we want to test is that the prehistoric rock art of the area is connected with a seasonal but regular occupation of the valley that started at the Middle Holocene period (Sub boreal), i.e. 6000 BP, when the climatic conditions of seasonality started and differences in climate and vegetative growth justified the periodic movements between valley and mountain (see Alifriqui 2003). In that sense, the rock art could have served as a kind of mental map, conveying information both on sacred and profane subjects, as entrances, resources, rights of access and/or prohibitions. To test ours ideas we are developing a field study including GPS positioning of archaeological and rock art sites, digging, 14C and archeobotanical sampling and landscape modelling by means of GIS. We offer to discussion some first results on the GIS appliance devoted to modelling the Oukaïmeden rock art landscape.

2. Background of the archaeological knowledge on the High Atlas region

Research on Oukaïmeden rock art started in 1949 and since then works went on discontinuously until a general Corpus was produced by Malhome in 1961. Nowadays the Centre National de Patrimoine Rupestre (Salih & al, 1998) is in charge of research, being responsible for test pits in the area. At the moment it is only provided one 14C datation for the rock art area, dated to the beginning of the first millennium B. C. (Graoui & al. 2008).

3. Getting Data

To carry out all these analysis we need several kinds of cartography, in order to generate a DEM2 and archaeological data.The first problem was the absolute lack of cartographic information; there was no cartography of the area studied. How to get the topographical information needed? .We examined all the possibilities and finally we found in the Global DEM our solution to get topographical data. Examining and studying all the Global DEM available, we concluded that the most appropriate for our investigation was the ASTER-GDEM. which has 1 arc-second (30 m) grid of elevation postings. To get thematic cartography about the land cover, we used Classification of Remotely Sensed Imagery. We worked with Thematic Mapper images.Our team used GPS for positioning of archaeological sites and rock art stations and developed a database (fig. 1.)

2 Digital Elevation Model


4. Analysis


To study the mobility, first of all, we developed a friction surface based on the strong slopes which represent the resistance to travelling from a pixel to the nearest one. To create a magnitude of friction surface required assigning friction values to slope values through an empirically derived equation. Then we calculated a cost surface using an anisotropic function based on the direction of the movement and the friction surface. Those frictions that vary in strength depending on the direction of movement are known as anisotropic frictions.


Once a cost surface has been created using an anisotropic function, a pathway tool determined the least-cost route between any designated cell or group of cells.

Visibilities (Viewshed)

With the help of the DEM from a cell the visual rays are extending in all directions and find the line of sight at the height of the cells to determine if they are (or not at sight.

5. The archaeological analysis

Sunstroke layer (fig.2)

Rock art areas lie predominantly on the slopes oriented S/SE facing to the sun and to the best sheltered and warmest side of the valley. This is also the main geological orientation of the sandstones on which rock art is engraved, although there are some exceptions, as the quadruped motifs engraved on sandstones standing on both sides of the Tiferguine’ stream ford. Opposite to them are located the funerary areas (tumuli), all of them aligned to the East and looking to the North. In between the valley bottom and the natural paths where people and animals move on the valley area run. One important shortcoming we have to cope with it is the dearth of chronological information. from a chronological point of view. So, we do not know yet whether rock art were partially or totally contemporary with tumuli. Anyway we might have to speak of a complementary between rock art and tumuli extended through time, sometimes both were complementary and sometimes one of them (the tumuli), could have substituted the art as a symbolic meaning of territorial control. So the location of both of them can be interpreted in a sacred and profane way at the same time. The south orientation provides the best shelter in an area subjected to alpine climatic conditions. At the same time, the confrontation rock art/graves suggests the opposition light/darkness or life/death. . It could be said the same about the predominant East orientation of tumuli, referred to death and regeneration in a symbolic/profane organization of the landscape with the paths at the valley bottom marking the border between life and death and the cycle of self-regeneration of nature (Bradley 2000).

Transits Layer(fig.3)

Another regularity in the pattern of the landscape can be concluded from the analysis of the present GIS layer. Rock art is distributed following a W/E axis, spreading across the river Irini course which plays the role of main movement’s distributor. No less important is the fact that small water streams and seasonal wadis are distributed perpendicular to the Irini river like in a fishbone pattern, because they conform watersheds as natural fringes between areas. Curiously enough they correspond almost exactly with the study areas we had distinguished for practical reasons during our first field campaign in 2008. One hypothesis we are working on is that watersheds coincide with the present fringes between pasturage areas of the different factions of the two tribes currently exploiting the valley and probably, with those of the prehistoric inhabitants, so that natural features of the landscape would have been perceived and used as fringes and landmarks for people living and using their resources in a long durée way of organizing the landscape

Accessibilities (fig.4)

A third layer for discussion is the accessibility and ease of ambulation between archaeological sites. In order to measure the variable it has been considered the resistance to movement represented by the steep slopes of the territory under study which makes communications to and from Oukaïmeden, very difficult. To calculate the latter variable Naishmith’s formula was used (Poucher 1960), whereby a person moves an average of 5 km. per hour, although the higher the friction (in this case the slope), the longer the time invested in the displacement, at an average of 30 minutes per 300 meters of vertical drop, i.e. the initial time is doubled if the slope is 12%, triples if is 24%, quadruples if 36% and so on. We could add to this formula the one posed in the late 60’s by Higg and Vita-Vinzi, the so called site catchment area, designed to calculate the economic area of a human group living in a certain archaeological site. The formula was calculated on the basis of ethnographic accounts of the behaviour of mobile and sedentary populations. Higg & Vita-Finzi’s formula calculates a maximum catchment area of five km radius or approximately one hour walking from the base for sedentary farmers and two hours walking or circa ten km radius for mobile hunter-gatherers and herders (Higgs & Vita-Finzi 1972). Taking into account both formulae we can draw some conclusions from the analysis of the accessibilities layer. Firstable, rock art areas and archaeological sites are included into the area between one hour and maximum two hours walk from the centre of Oukaïmeden, what is more, the great majority of them fall into the one hour-walk area. And second, due the stress caused by friction, it seems probable that prehistoric people that used the grazing area and was responsible for the rock art did not come from very far away, but probably from the valleys below, that is, we should be talking of transterminance instead of transhumance. As we said before (vide supra), at least from the 16th century A.D. onwards, the pasturages of Oukaïmeden are exploited exclusively by people coming from Rerhaya and Ourika valleys. An ethnoarchaeological enquiry developed by our team among the Berber shepherds based on the summer villages or azib showed that herders entering the summer pastures of Oukaïmeden, spend between four hours and a day's march that, given that the entry day to the grazing area is August, 10th, it can be estimated at a maximum of 10 hours. It is true that nowadays herders going from Ourika profit from the road built in French colonial times and that the Moroccan government has recently built a track which connects Oukaïmeden with the small villages or douars

sparse on the Rerhaya valley, so that the time spent could have been slightly longer before, although, given the steep terrain, differences may not have been very significant.

Spatial organization (fig.5)

Next layer to be considered is the one devoted to visibilities. In this case, we are interested not as much where the rock art sites are visible from or the views commanded by a certain point marked by rock art, as what a person moving through the valley would have seen. Otherwise, whether there is a visible pattern in the rock art location. This is a question not easy to resolve, for weapons, idols and animals could have been depicted on the same rock surface (photo 2) or, as in some cases, weapons or other depictions are superimposed to animal engravings, so we could guess that the message conveyed could have changed through time.

Also types and techniques of representations differ from each other, what means that they probably are not contemporary, a problem difficult to solve due to the fact that there is not an easy correlation between human occupations and rock art, and 14C samples taken from our 2009 digging campaign are still in progress. In spite of it we could say, based on typological reasons, that most part of carved weapons can be dated at the Earlier Bronze Age, so and broadly speaking at the Second Millennium

B.C. and that all or at least most part of animal depictions seem to be older for typological reasons, that is, that they should be dated at the Late Neolithic or Copper Age on the base of the lithic industry sampled in the area. More problematic are the human depictions, for whom there are neither typologically homogenous nor all of them probably contemporary. Anyway we could speak of trends in the way that the pattern of the rock art is distributed. In that sense, weapons seem to be mainly concentrated at the entrances, especially although not only at the passes. Animals are spread either on fringes as happens at our zone 9, the so called Bull’s pass (Col du Toreau), or close to the main hydric resources as, for instance, the endorreic area nowadays transformed into a reservoir (our zone 8), the ford over the Tiferguine stream (our zone 10) and other wet areas located at our 10th and 11thareas (photo 3)

Regarding human depictions, they are probably not all contemporary. Anyhow, those male and female depictions with explicit sexual features lie always on marginal areas at the fringes of cliffs. Therefore it is tempting to suggesting that they could be related with the celebration of

rites of passage to the adulthood.

6. As a way of conclusion

First results of GIS application to the archaeological area of Oukaïmeden suggest the existence of some kind of structure in the rock art distribution, probably related to different chronological periods and with apparent changes among them. This internal structure is organized on a main axis along the valley, connected with watersheds and in obvious relation with the geological structure of the area, but it also reflects the human use of the space in Prehistoric times, and subsequently. There is still a huge amount of research to be done on the site, but a crucial point to be developed on the future is the global understanding of the economic and social system underlying the gravures, what means to broaden the archaeological fieldwork and GIS to the valleys below, where trasterminant shepherds expend the rest of the year.


1 Acronym of Arte, Paisaje y Poblamiento en el Alto Atlas. M Ruiz-Gálvez*; Y Bobkot**; M H. Collado***; M. ElGraoui****; Farjas*****s; E. Galán******; A Lemjidi**; C. Nieto*; P. de la Presa*****; J. de Torres*. *. Complutense University, Madrid;**Insap, Rabat;*** Polytechnic University Madrid; **** Council of Extremadura’s Heritage, Merida;***** Archaerological National Museum, Madrid;****** Centre National du Art Rupestre. Marrakech



ALIFRIQUI, M. 2002: “Résultats des études écologiques et phytohistoriques de la station de l’Oukaïmeden ». Actes des Journées d’ étude du Patrimoine et Environnement. Rabat. Ministère de la Culture et de la Communication.

BRADLEY, R. 2000: Archaeology of Natural Places. London, Routledge

FRANCHI, F. 2005: Société pastorale et société urbaine: enjeux territoriaux autour d’un agdal. étude réalisée à partir de l’exemple de l’agdal n’Oukaïmeden, Haut Atlas de Marrakech. LPED, Université de Provence.

GRAOUI EL, M.; ALIFRIQUI, M.; JUNGNER, H.; NAHID, A. & SEARIGHTMARTINET, S. 2008: “Recherche d’indices chronologiques sur le paysage des graveurs de rochers de l’Oukaïmeden (Haut Atlas, Maroc)”. Sahara, 19, pp: 105-108.

HEIDEGGER, M. 2003. Ser y tiempo. Madrid, ed. Trotta.

HIGGS, E. S. & VITA-FINZI, C. 1972. "Prehistoric economies: a territorial approach".

En E.S. Higgs (Ed.): Papers in Economic Prehistory. Cambridge. Cambridge University

Press, pp: 27-36

MAHDI, M. 1999 : Pasteur de l’Atlas. Production pastorale, droit et rituel. Casablanca, Fondation Konrad Adenauer.

MALHOMME, J. 1959-1961: Corpus des gravures rupestres du Grand Atlas. Rabat, Service des Antiquités du Maroc (Publications du Service des Antiquités du Maroc, Fascicules 13-14).

POUCHER, W.A. 1960: The Lakeland Peaks. London, Constable

SALIH, A.; OUJAA, A.; HECKENDORF, R.; NAMI, M.; EL GRAOUI, M.; LEMJIDI, A. & ZOHAL, H. 1998: “L’aire rupestre de l’Oukaïmeden, Haut Atlas, Maroc: Occupation humaine et économie pastorale”. Beiträge zur

Allgemeinen und Vergleichenden Archäologie, 18, pp: 253-295.

Créer un site gratuit avec e-monsite - Signaler un contenu illicite sur ce site