ai es Bollettnto del Centro Studi ¢ Museo chArte Pretstorica di Pinerelo, (tala A CONSISTENT NOMENCLATURE FOR THE ROCK ART SITES OF THE SAHARA AND AFRICA Yves & Christine Gauthier - JeanLoie Le Quellec Keywords: Messalk, Normericlatur, Rock rt Résumé: Cartes impréctses ctfau incompléles. toporymes absents aul wera bles d'un document d Mautre rendent difficile sinon ambigué toute discussion sclentifique sur ie Messale (Fezzan), La nécessité de décrire auce préctsion la localisation des sitions rupesires nous pousse @ proposer une remenclature fotdement indépendantes ete la loponymie, Celle nomenclature, balable en fete pot iain le continent afriertin, repose sur unt décenpuge en paves réguliers ba sé sur les coordonnées geographiques. tatvude vt longitude, La convention pro- poséey autarise des extensions cn eas de nouvelles découvertes, Rinssunto: Le poche mappe riguercarti ta zona det Messale (Pezcn, Libial sone generalmente vagite ¢ theotplete. Inoltre {nomi dei tueghi risuttano spes 80 memeantt o artediritura diversi da un documenta alaltre, cosicché qualsiast disctrisione scicinijica st questared & particolarmente di{fcoliosa ¢ ambigta. La necessité di deserinere ¢ ai collocare von precisione ( sitl dl arte rupestre et ho conviatéa properre una terminotegia che sia totatmente indipendente daitc fopanimia, Vatice per Cintera cantinente africana, la sostre terminalagia st hee sa strict divisione del terrtiorio mediante unc quacreteaturc regolare, usando le coordinate stanctard di tongitucline © latituctine, Quesidt conbenzione ¢ pxisst bile di esiensioni nel caso si verifichine scoperte di nuovi sitt di arte repestre, Motivation First and foremost when studying Rock Art sites: the researcher must be able to deseribe the local topography and. hopefully. localize the site relative to oihers. The question of sile name or ether designation can lead the researcher into a morass of conflicting information, This is a general problem for the Sahara and a crucial one for aur work in the Messak (Libya), which has been impeded by the paucity of accurate “site name” desipnations or taponyus. Available maps are approximate at best, Ditferent maps will place a wadi here or there. seemingly anywhere. Discrepancies of 5 km are nol uncommon. Qverlaving a map ol a local region on top of a satellite constructed map can lead to extreme confusion as to the name of any particular place. Survey 1909-1994-1995-1996 - Anmi VIEWIIEIXDC- N. HLO-11- 12 W1One finds that in most insianees our present topanynile Information stems Irom the cra when nomads wandered extensively in this region of North Alrica, Place names were collected by the military, explorers and carly resear- chers from local, regular passersby. whose experience of (he desert was cur rent and for their practical purposes precise. Today since the nomadic way of life has been sharply curtailed, these contacts are rare. Consequently. lear- ning names or other designations for any place In these remote desert regions has become more diffleult. This problem is further complicated by loss of pla- ce names, or, generalizations of the name of 4 portion of a valley to refer even- tually to an entire region, This ts the case for Ber|tj. Place names vary kecor ding to local dialect and. as well. to the linguistic background of the resear- cher, The site Yerjuj, for example, has been Wranseribed as; Berdjudsch, Berdjuci, Barjoo, Barju,), Bergig.! The inevilable conclusion | no map can be considered at present lo be definitive. A few further examples can help illustrate the general problem. Are the Geddis and Beddis wilyin, mentioned on different maps, names for different valleys or merely. differ ent names for the same sile? In the Northern Messak Mellet (the “White Plateau, because is esandy+,) the Imrawen wadi was identified by one of us (JL) through Tuareg nomads of Bab erMaknitisa who travelled there. However, on an American map the limweuen wadi is placed tens of kilometers furdher South. What should one chink’? In the Messak Set- lafet (the »Black Plateau, because of its predominant colour) there Is seme confusion between the Malian designation of the ]-n-Eiedh wadi and the -n- Ahacd wadi shawn (n oler sources, The site designated erWarer is often in- correctly (ranseribed as El-Aurer, suggesting an erroncous pronunciation, In general, the extent of the Messak Settafet and Messak Mellet are well known and bounced on one side by the Aramas wadi which the Americans, as im- plied above. once again place further South and the region called ‘Mestafet’ ow other maps is clearly an abbreviation - Mes[sak Set)tafet, ete, Is (here a solution to this predicament which can be-achieved in nable tinie-frame? The direet method of organizing a research team which will methodically catalogue each and every wadi and hillock using an agreed upon convention for names and with satellite verified coordinaics ty impractical, Many years of effort and an unreasonable amount of money would be requ red. With this In mind, the unedited work begun in the 40's by G, Lelubre presents a viable startlag poll and allernative.? During his time in the Mes sak Seliafet, he travelled with an salct Kel Messak. who heel an admirable knowledge of ihe countryside-* trom Ubdari to the Aghelad pass. Togelher they identined most of the widydn ane ad an thelr journey using & Pelgn pass, a watch and length of a camel's step. The method was primitive 5 1A lable whtel cam ares de transcriptions wie Eaglisth, Fre iertman ond flaln 10 the tansliterations of sile names Irum the orginal Fergani and to (he Arabic characters .ravicedd by Le Quellee 1986: 362-863. This work suggests hit researchers shank) be enreurnged lo adopt & Wanslitcration system for the price al a litile etfart ihe transliteration system: for dhe price of We effort the pubewtions would gate clarity, * Thanks to Mr Lelubre who kindly placed his unedited documentation at our disposal. ) belubee, in leit, 241% 1992, 1996 | Anm VILWIAS-X N91 11-12 112 Survey 1902-18-19markably accurate. Nevertheless, there nains much work whieh sull needs to be done in the South of the Settafet and for the entire region of the Mellet Even assuming that the name survey might be completed, there would always exist the chance for a few remaining nameless iopographie features. How shall we proceed"? Should we use arbitrary names unrelated to the local history ? Indeed we should avoid names such as ‘Tadrek n’Elizabeth» or in- congruous designations such as la Baleine (Whale). or ale Sous -Marin [Sub- marine)», Consider the -wadi Haleebs,? which we could never succeed in fin- 2 cither on a map or through the help of a local guide... and for a good rea- son: this incidental name was given to the site In the Messak where was di scovered a famous milking scene, and without doubt, this "Haleeb» refers to nulk fhalib in Arabic). li scerns a shame to us that such names should become established by common usage since one rapidly forgets that those names ha- ve been given recently. Under suck circumstances, more attractive would be stead an ancient traditionnal place name: it would have meant that Tuaregs had pafd attention to the remnants of Rock Ari and Included them in their oral traditions. Certainly it should be hoped that the names imtreduced Into the literature should be properly accredited and correctly transcribed. Finally, is there a need ta remember that, for those who understand them. the indigenous names of place are anywhere as precious reminders of lhe local environment and culture? ff the terms such as Bir (pl. Abyart ewells or Guelta [pl, Gla) splace of waters or Gareat (pl. Gur) «mesa» arc gone from common usage, others are merely less well understood: Ramiat «sands or Ta- drart, feminine for Adrar. «chain of mountains», Sulf others are often ignored. for example, fheren whic the plural of Aff «source of drippings, Antu (pl. Gnan) spermanently closed wells or Abankor (pl. Ibenkar| «temporarily drained wells, incorrectly transcribed by Frobenius as «lbanka»." In the region souiliwesl of the Ft 2, one can also cite the example of the Bab erMaknu: Bap where It is helpful to know the name «Passage swept (by the wind)», while the famous «Adrar Ektebens (or «lktebine) - correctly translated by Frison-Ro- che as «Mountain of Writings: will provide, for example, a place of interest for a researcher of Rock Art Having understood the above arguments, 1{ should be recognised that we need determine way to precisely localize the sites under discussion. Il is im- portant that we formulate a nomenclature which is simple, anambiguous and which we will all agree tu use. Our charge From the previous discussion, we have learned: First, the importance of crealing a nomenclature thal fs lotally Indepen- dent of local place names whieh are only occ ly helpful and/or aceur: * Berthoud, 1977-78, did not name dh eauian by Jacquet 1978. § Probenius 1937) 14. Concerning qeogeaphic and hydrologic terms in arabi thara, ser Capot-Rey et al. 1963, site, The nainé Comes from an independent publi rit berber in the 1883-1904- 1995-1998 - Aran MIME ES = RO site 1Second, ft should he a universal system acaplable to the whole of Africa nol merely the Sahara, Third, il should be an open system which permits ad ditions without creating any confusion, Inevitably, there will be ur adogucd sites. A namenelature which allows, for example. a wadi name and ring number ( some combination of a geographical criterion and a chronologi- Wone ) risks problems if other sites are discovered among those that are al- ady known. Finally, this nomenclature should be precise and use a limited number of symbols to definitively determine the location, It should alsa be ea- SY to read. Proposal The only simple system, that requires no discussion and is already untversal- ly used in other fields, is (he one avigation + co-ordinates according lo longitude and fadtude. What ren + is to adapt this system to our needs, Africa is bounded by longitude 17° Wes! and 52" East and the latitudes 38° North and 34" South. Designating a place wih a complete co-ordinate set, G2. 32°1GN/12"45E, requires an unacceptably large mumber of characters, 13 i We suggest dividing Africa into 2" squares of longitude and latitude, using O° longitude at (he equator as the origin of our coordinate sysiam. The 26 letters of the alphabet would then provide 52° on cither side of the Greenwich Meridian- This is sufficient to reach the Horn of Aivies (app. SIE J and Mauritania (17°W ), In general the alphabet is sufficient to detine all African latitudes, To achieve an appropriately fine grid we propose that each 2°x2° square be subdivided inta 10° squares. lor the larger squares we will use capital letters and for the finer grid we will use small lellers, Having thus divided Atrici to small sections, we need an unambiguous method to distinguish the four points of the compass referring to our origin at the Greenwieh-equator intersection, Rather than use a minus sign ++, we pro- pase that the letters for West and South should be underlined, This conven- tinn can easily be outlined by the following: A corresponds {(q 0 to 2° East B corresponds to 2 to 4° Bast “corresponds to 56° to 52° East A corresponds to 0° tp 2° West B corresponds to 2° to 4° West a corresponds to the grid O' - 10° in the larger grid b cor sponds to the grid 10° - 20° in the larger grid feorresponds lo the grid SQ" - 1" in the larger grid la Sureny 1980-184 1945-1998 — Arm) MU-VIIEDX-K 8990-101# corresponds to the grid 1° - 1** 10° in the larger grid 1 corresponds to the grid 1°50 - 2° in the larger grid For instance: the site El-Warer (alias el Aurer } found al 25° 42° 74" N/ 12°05’ 53" E translates lo MkFa. The sile 16° 168/13"15E is IbGk ‘The site 16° 165/13"45W ts IbGk. ‘The site 16° 1GN/ 13°45W Is IbGk One can easily see thal the 13 characters used by our navigational sy- stem have been reduced Lo 4, At the Equator one minute of are is equivalent to 1852 m, therelore our finest grid refers to a square which is approximately 18 kilometers on each edge. In some areas, where there are parallel widyan whieh are very close to each other (e.g. the Messak, nol to mention the difficulty caused by innume- rable intersections}. this grid is tay loose to account for the many sites which are often less than 300m apart. Here, one can make use of a chronological or- dering, for example. IbGk-2 for the second site found in block IbGk, There is nothing particularly original about this nomenclature. Such a system is used im Canada,® but it depends. in that case, on the grid defined for their national map. It would also be possible to rely on the grids defined by the IGN (Institut Géographique National) for maps with scales 1/100 000 and 1/200 000 ex- cept that then one would need more characters for the larger squares [e.g.: NG32 for Djanet). The edges of a block [4".x 6°) would make using the alpha- bet difficult if not impossible, since for the same resolution achieved above, one would require 36 characters for longitude. A fine grid At the expense of adding a few characters {6 instead of 4 in contrast ta the full 18) it is possible te consider a grid of 1.8 x 1.8 km by retaining the ciphers for the minutes. This requires that we have a convention to distingui- sh the degrees (since each letier represents 2°), This can casily be accompli- shed by using capital and small letters, where small letters refer to even de- grees and capilals to odd numbered degrees, More coneretely: for {he site at 16°16N/13°45 E-1:i 16G45-1 (1st site of grid) for the site at 17°16N/13"45E-3: 116G45-3 (3rd sile of grid) for the site at 17°16N/12°451 11645-2 (2nd site of grid) & Wainwright 1985. Survey WKS 1M 1908 1900. Anni VIDMILDOX Nod tD-t bez 115Thus the code is more easily readable and the final grid has better reso~ fution. It ts this last system ziflich wwe propose for use. Remarks With regards te the last code, one should strive to respect the geographical position wherever possible. If for very short distances, the chronological orcer lakes precedence (some sites being discovered lat this should pose little In- convenience. ‘This system is sufficiently open ended and free from the con- straints and hazards of teponymic research, Besides. nothing stands in the way of adding a loponym to this nomen clature. For instance, when one describes an entire wadi, this would permit reduction of the text and guarantees its homogeneity. In any case, with or without reference to a toponym, the sites are perfectly situated, The grid, thus delined, seems to us to be fine cnough for most cases. Certainly. we have seen that many sites can co-exist in (he same small grid, but to be more precise we would have to acd more characters to consider a 10° gricl, for example. At which point one reaches the limits of the navigatic- nal system. Speaking of which, one should remember that Rock Art sites are allen 2s much as 1OOm long and thal the co-ordinates are often only deiermi- ned ‘near’ the art-decorated cliffs (note that the navigation system is generally attached to one's vehicle), Hence, a resolution as fine as 1" of are would be diffecult to use in many instance: The authors who intend to use this system in their own work ( see Gauthier- Le Quellec, 1994) wish to submit this prepesal to the community al large and look forward to suggestions ftom the communily Jor improvements to its precision ( if possible 4 reduction im the number of characters) and/ar improved readability { permitung easier translation to physical co-ordinates |. We are indebted to F. Hutson for his help in translating the text Y. & C. Gauthier : 264 rue de la Balme. F-38950 Saint Martin le Vinoux JL, Le Quellee : Brenessard. F-85546 Saint Benoist sur Mer References Bernioup §,, 1977-1978, Gravures repestres inéaites au Messak Selqfet (Fexean, Linch Geneve-Aivique, Acta Africana vol, 16.n° 1, p. 109-117. Caron Rev R,, Conner A. er Braypin ot THE B., 1963, Glossaire des principaux ter mes ydogrophiques et hydrogéologiques saliacens, Alger, Institut de Rerhecches Saha rienne: i sowits, Gacrimar ¥. & C.. Lin Qoniuec. 1994, Quelqies personages particiiters qravés au Messuk (Peezan, Libye), Survey, same issue. (HENIUS L.1997, Ekle Ektab, die Felsbilder Fezzans, Letpvig, Otto Harvas- HG Skiivey 1994 E90 199561996 Awor MIEVILIN XRT 12Le Quetiec JLL1987, Liart Rupestre air Fezzén occidental (Libye): widyan Zrecda el Tara (twdelt est Sheti), Camb. Mono, in Afri, Archacolagy 16, BAR Intern. series: 365, n°22. Jacnuer 1978, Au coeur du Sahara thyen, d'éranges gravures rupestres, Av cheologia 123, pe $0-51 Warnwricht LN.M., 1985, Rock art conservation research in Canada, Gollettino del Centro Camuno di Studi Pretsterict 22. p. 15-46, Buvey 6 AMM NILVIN XNA Le uw?