Rock art motifs are found in every inhabited continent. In most cases those motifs develop independently, but it is also certain that specific rock art motifs migrated from one area to another area, sometimes travelling for thousands of kilometres. This study investigates the possible long-distance diffusion of a number of abstract rock art motifs along the Pacific seaboard of the Americas.
By Maarten van Hoek
Possible Indications for Long Distance Diffusion
of Rock Art Motifs in the Americas
Maarten van Hoek – rockart @ home.nl
A PDF of this paper can be downloaded at ResearchGate
It seems to be such a simple ‘rule’ when discussing the distribution of rock art motifs: the simplest rock art motif will occur almost everywhere and in many different contexts. This ‘rule’ is undoubtedly authenticated by the cupule; the ubiquitously occurring, small hemispherical depression in rock surfaces. The cupule occurs in the rock art of all inhabited continents. It is found in extremely remote islands like for instance Rapa Nui in the Pacific Ocean, as well as right in the centre of a busy city, like Seville in Spain. However, this specific occurrence by no means implies that the cupule tradition diffused for instance from Rapa Nui to Seville or vice versa. Although there certainly are examples of ‘migrating’ cupules (for example along the Atlantic seaboard of Europe), the cupule is by far the best example of parallel evolution in many if not most cases. This ambiguity makes the cupule completely inappropriate to serve as ‘Leitmotiv’, or – in the scope of this study – as a suitable object to demonstrate diffusion.
Therefore the ‘rule’ should better read: the more common and simpler a rock art motif, the more unsuitable it is to use it in tracing migration patterns. But also more complex rock art motifs are not always suitable to offer proof of diffusion, especially when it concerns anthropomorphic figures or parts thereof. A telling example is the vulva-motif that occurs in many much differing cultural contexts (for instance on Rapa Nui, Chile, and at Chuquillanqui, northern Peru), simply because of the presence of girls and women in every inhabited continent.
ABOUT THE ILLUSTRATIONS
Because of the nature of this study I have included a relatively high number of maps. All those maps are based on Google Earth / Open Street Map and all distances (as the crow flies) mentioned in this study are based on Google Earth. Notice that the squares on the map only very approximately indicate the locations. Moreover, in some instances my drawings (especially those in the maps) will be imperfect, if not incorrect, especially the one from Issamadanen, Mali, which is based upon a rather small photograph. In most cases all other images on the rock panel have been omitted by me. Because I have not visited most of the sites mentioned in this study, I have used many sources to base my drawings upon. Therefore, the drawings – all © by Maarten van Hoek – are based on several sources. Yet, I am the only person responsible for any error in any of the drawings made by me, especially regarding those drawings that are based on illustrations of other authors. Consequently I am also responsible for possible misinterpretations based on those drawings.
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 this article).
Haga clic en cualquier ilustración de este artículo para obtener una imagen más grande (haga clic en la flecha < hacia atrás [arriba a la izquierda] o, en algunos casos, haga clic en la ilustración para volver al artículo). Haga clic en los hipervínculos (en rojo subrayado) para ir a otras páginas web, que se abrirán en una ventana nueva (haga clic en el icono X [arriba a la derecha] para volver al artículo).
PART I: DIFFUSION OR NOT?
In Part I of this study I will first discuss five specific rock art motifs or scenes that clearly demonstrate the ambiguity of the concept of diffusion. It concerns: The Camunian Rose; The Cup-And-Ring Motif; The Twin Anthropomorphs; The Minoan Labyrinth and The Outlined Cross. In discussing those examples I definitely do not claim that they all are unambiguous examples of diffusion. In most cases there is no or unreliable (informed) proof and thus those examples (and many other motifs) may well be instances of independent, parallel evolution. I am convinced that no-one has knowledge of all rock art panels in the world and its imagery. Therefore, this study pretends in no way to discuss all the known examples of a specific motif that has so far been recorded in the numerous rock art sites across the world. Also the maps included in this study only offer a selection of sites where a specific motif has been recorded. There may and will be (many) more sites. Finally, this study ignores to discuss the meaning of every specific motif, also because – when there would indeed be question of diffusion – it is always uncertain whether icon and meaning travelled as a team; the metaphorical content may have changed while the motif migrated or after its arrival in an often completely different natural and/or cultural environment. When a motif developed independently in several separate rock art regions, there will be small or no chance that the metaphorical content will be (exactly) the same. Finally, it is always possible that a specific motif both diffuses across the globe and evolves independently. This especially seems to be the case with the Outlined Cross.
The Camunian Rose
It proves that, in order to discover migration patterns of rock art motifs, it is to be better to use complex patterns. However, it also proves that the reverse of the ‘rule’ explained above has its severe limitations as well, as it is not true that all complex patterns have a limited distribution (confined to a small geographical area) and/or occur in only comparable contexts. An example is the rare rock art motif called the ‘Camunian Rose’ (Figure 1). This outlined motif typically consists of a meandering closed line that winds around nine cupules, although other patterns also occur (Farina 1997; 1998). The major concentration of the ‘Camunian Rose’ is found among the petroglyphs of Val Camonica in northern Italy, the valley after which the motif is named. It also has been recorded on an outcrop on Ilkley Moor in Yorkshire, England, where it is known by the name of Swastika Stone. Other rock art examples are found at Guifões in northern Portugal and at Askum, Bohuslän, Sweden.
Figure 1: Map of the world showing a selection of sites featuring the ‘Camunian Rose’ motif. The large square in northern Italy indicates the main concentration in Val Camonica.
Surprisingly, the ‘Camunian Rose’ also occurs at the petroglyph site of Issamadanen, Mali, in the centre of the Saharan Desert (Dupuy 2009: Fig.7) no less than 4360 km south of Bohuslän. Although Cristian Dupuy argues that the distribution of this symbol ‘bears witness to interactions over extremely long distances’, I still prefer to leave open the possibility of parallel evolution. Yet, I do not reject the possible diffusion theory regarding the ‘Camunian Rose’ as suggested by Dupuy, as unambiguous diffusion of the ‘Camunian Rose’ is evidenced by a petroglyph at the rock art site of San Felipe, Coahuila, Mexico. The only difference is that the cupules have been replaced by small circles. It is said that Christian missionaries are responsible for this – so far – unique example in the Americas (Encinas 2014), which – in that case – must have been manufactured some date after 1492.
The Cup-And-Ring Motif
Another example is the cup-and-multiple-ring motif that is found along the Atlantic Seaboard of Europe, from Galicia, via the British Isles up to Scandinavia. The distribution along the Atlantic Seaboard most likely involves diffusion, although the exact direction of the migration of the Western European cup-and-ring art is still obscure. In many other cases there is no question of diffusion of the cup-and-ring symbol, as equally complex cup-and-ring motifs have been recorded in many parts of North America, for instance, I noticed a fine example at Jug Handle Arch, Utah (about 8000 km west of Western Europe across the Atlantic Ocean). I even recorded several examples at Alto de Pitis in the Majes Valley, Peru (9000 km SW of Western Europe and 8000 km SE of Utah following the spine of the mountains). In my opinion these American examples developed independently and definitely are not cases of diffusion from Europe. However, diffusion of specific motifs like the cup-and-ring symbol within the Americas is always a possibility.
The Twin Anthropomorphs
Yet there are complex scenes that are found (more or less) repeated at completely different places, without involving diffusion in any way. Of course, the more complex a petroglyph (scene), the more details will be dissimilar. A fine example is the unique scene of two seated (?) anthropomorphic figures on a boulder at Tamentica-1 in northern Chile. Although the two figures are facing each other, the scene (especially the position of the flexed arms and legs) is much similar to the scene of two anthropomorphic petroglyphs on a boulder at Burrup Peninsula, Western Australia, that seem to be sitting (?) back to back (Figure 2). But Burrup is almost 18.000 km west of Tamentica (involving 14.000 km across the Pacific Ocean; see Figure 1) and is moreover found in a completely different cultural context. Therefore there can not be any question of diffusion of this specific configuration in whatever direction.
Figure 2: A: Petroglyph from Tamentica-1, northern Chile, compared with B: a petroglyph from the Burrup Peninsula, Western Australia. Drawings © by Maarten van Hoek; A: based on my own photos; B: based on a photograph by Jeremy Storey.
The Minoan Labyrinth
Another striking example illustrating the problems involving alleged diffusion is the labyrinth motif. It occurs on a wall of Lucca cathedral in Tuscany, Italy, as well as on a rock outcrop at Cocoraque Butte, southern Arizona, USA; (Figure 3) almost 10.000 km west of Lucca. The Lucca labyrinth is believed to date from the 12th or 13th century, but the date of the Cocoraque Butte labyrinth is uncertain and its manufacture can range from A.D. 1 to almost the present date. Also the Hopi peoples in the Southwest of the USA manufactured the labyrinth. A rectilinear Hopi labyrinth petroglyph has been recorded near Oraibi, Arizona, but its date is uncertain. Also the example on a wall of the Kom Ombo Temple, southern Egypt, is hard to date, but definitely post dates the construction of the temple. The example found engraved inside the tomb called the ‘Domus de Jana’ near Luzzanas in Sardinia, Italy, is said to date from about 4000 B. C., although this date is also disputed.
Other examples have been reported at a church wall at Sipoo, Finland; on a boulder at Padugula, India; on a boulder called El Laberinto del Dormido in Coahuila, Mexico, on a boulder near Hollywood, Ireland, and even at Machu Picchu in Peru (but I regard this to be a questionable documentation). One isolated labyrinth petroglyph is found on a vertical cliff in a remote gorge at Tintagel, Cornwall, UK (Figure 3), while another labyrinth petroglyph is found among cup-and-ring motifs at Mogor in Galicia, Spain. At Naquane at Capo di Ponte, northern Italy is another ancient labyrinth petroglyph.
Figure 3: Map of the world showing a selection of sites featuring the ‘Minoan Labyrinth’ motif. The two yellow squares indicate the location of the two drawings in the map.
The Outlined Cross
The Outlined Cross, also – more controversially, though – known as the ‘Venus Cross’, has a very remarkable distribution (Figure 4). It occurs at many rock art sites (both as painted motifs and as petroglyphs) that are – like sites featuring the ‘Camunian Rose’ and the labyrinth – incredibly far apart. Interestingly, Christian Dupuy (2010: 51) recorded two Outlined Cross petroglyphs in the centre of the Sahara Desert together with a petroglyph of the ‘Camunian Rose’ (all on the same boulder!) at Issamadanen, Mali. However, from his photographs and drawing (2009: Figs 4a and 4b, 5b; 2010: Fig. 5a) it seems that the two Outlined Crosses are rather imperfectly executed and are moreover weathered and partially flaked. It is hard to make out the exact layout of the design.
Figure 4: Map of the world showing a selection of sites featuring the ‘Outlined Cross’ motif.
When discussing those (two?; three?) petroglyphs at Issamadanen Dupuy seems to favour the idea of diffusion: ‘Leur spécificité, leur rareté et leur vaste répartition géographique permettent d’y voir les manifestations d’interactions rapides et conjuguées sur de longues distances’ (2009: 122). Unfortunately, it is uncertain whether Dupuy – with his statement – refers only to the two Outlined Crosses or to the ‘Camunian Rose’, or to all motifs, because he does not mention analogies.
Interestingly, the Outlined Cross motif recorded at Issamadanen has also been reported at Imaoun in southern Morocco, 1450 km NW of Issamadanen (Searight 1999). Even more important is that at both Issamadanen (Dupuy 2009: Fig. 3b) and at Imaoun the same complex, idiosyncratic abstract designs occur (Dupuy and Searight 2005). Therefore, there are at least two parallels: the style of decorating and the Outlined Cross.
But the Outlined Cross also occurs at the A Laxe Escrita (two examples), Carnota, Galicia, Spain, and (several examples) at Storsteinen, Alta, in the far north of Norway (5700 km NNE of Issamadanen). Although I have reservations against diffusion of the Outlined Cross also in this part of the world, the African examples might have diffused to Europe or vice versa. Therefore, in all those cases the statements about diffusion by Dupuy may be correct. Yet, there is a problem. The Outlined Cross is found at many more spots in the world. Then, how would Dupuy defend the diffusion theory when considering the following examples of the Outlined Cross motif?
Indeed, what Dupuy may have been unaware of, is that the Outlined Cross motif occurs at many rock art sites that are incredibly far apart. For instance, a painted example of the Outlined Cross motif has been recorded by me at the rock art site of Boca Onima on the island of Bonaire in the Caribbean Sea (7485 km west of Issamadanen and across the Atlantic Ocean). I also recorded Outlined Crosses at various sites in the Andes of South America (for example at Cerro Mulato, northern Peru, 9330 km SW of Issamadanen and at Banda Florida in NW Argentina, 9100 km SW of Issamadanen). However, I also recorded examples in the Rocky Mountains of North America (for example in the South Mountains of Arizona, 10.840 km WNW of Issamadanen). All those American examples are found across the vast Atlantic Ocean and moreover on the western sides of those continents.
But there is more. There is also a striking concentration of the Outlined Cross in several groups of islands in the western part of the Pacific Ocean (Monnin and Sand 2015). Numerous examples have been recorded for instance in the Bismarck Archipelago and especially in New Caledonia (where more than 850 examples have been recorded!). Importantly, New Caledonia is located no less than 18.357 km east of Issamadanen in Morocco. It is evident however that – in this part of the Pacific – the Outlined Cross diffused from one island to the other, for instance via people who had the Outlined Cross tattooed on their body or via portable objects (Monnin and Sand 2015: 232; 252). However, being relatively close to Australia, it is remarkable that the Outlined Cross has not (yet?) been recorded in the abundant rock art of that vast continent. Also important rock art concentrations in the eastern part of the Pacific – like Easter Island and Hawai’i, both located between New Caledonia and the Americas – seem to lack the Outlined Cross (Monnin and Sand 2015: 267).
Other major concentrations of the Outlined Cross are found in the Americas. Although some examples are found in the east of North America (for instance at the Zella Shelter, Cumberland Plateau, Marion County, Tennessee, and in Lawrence County, Alabama), the Caribbean Islands (for instance in the Boca Onima rock shelter on Bonaire) and the east of South America (for instance at the Arco do Covão rock shelter in Caxingó, Piauí, Brazil), the great majority occurs along the Pacific Seaboard of the Americas and its ‘hinterland’. An imaginary ‘migration route’ from the northernmost rock art site in Alaska to the southernmost site in Patagonia (or vice versa) covers an incredible distance of about 14.000 km. Although several rock art sites have only one example of the Outlined Cross, in the Study Area there are many rock art sites where remarkable concentrations of the Outlined Cross occur. For instance, Argentina has several sites with often multiple examples of the Outlined Cross, like Banda Florida, Talampaya, Palancho, Parque Diaguita and several sites in the Valle Fértil area. At the Las Papas site in the Fiambalá Valley, also in western Argentina, a boulder with at least six petroglyphs of the Outlined Cross has been recorded, while in the same valley also petroglyphs of true Pipettes have been recorded. Also at La Silla in Chile, Mollebaya Chico in southern Peru and Yonán in northern Peru relatively many examples of the Outlined Cross have been reported.
A selection of other sites where the Outlined Cross has been recorded is (often only one example; painted or as a petroglyph): Antofagasta de la Sierra in the High Andes of western Argentina; Aconcagua, Illapel, Combarballá, Cuz Cuz, Chiuchiu, Conanoxa and Chapisca in Chile; Miculla, La Caldera, Alto de Pitis, Checta, Huancor, San Antonio, Cerro Negro, Chuquillanqui, Los Boliches and Mayascon in Peru; Copakati, Chuquisaca and Julpe in Bolivia; several sites in Venezuela, like the painted example at Cerro El Cejal; at Kurosopá in Colombia; Punta del Este in Cuba, at Mount Rich in Grenada and in the Dominican Republic; La Proveedora, El Tecomate and Tepache in Mexico and Valley of Fire and Grapevine Canyon (Nevada), Petrified Forest and South Mountains (Arizona) and in New Mexico, all in the SW of the USA. At Howe’s Tank in the SW of the USA a unique Outlined Cross of which the cross is composed of small dots has been reported as well.
Also, there are also numerous variations of the Outlined Cross, especially in New Caledonia. Important in this discussion are the ‘stacked’ examples. In our study area ‘stacked’ Outlined Crosses have also been reported, for example at the North Kofa Site in Arizona and Coyote Gulch, California, USA; at Los Naranjos and Totomixtlahuaca in western Mexico; in Nicaragua; at Mollebaya Chico in southern Peru; at Chapisca, Cerro Chuño, Huancarane, Río Grande and El Altar in Chile and at Antofagasta de la Sierra in Argentina.
Based on all these facts, I regard many global examples of the Outlined Cross to represent unambiguous examples of parallel evolution. Yet, ‘small-scale’ diffusion within a confined (but still large) area (for instance the western Pacific Ocean, the Andes or perhaps even Western Europe) is most likely to have happened. I already mentioned the fact that people in the western part of the Pacific Ocean had the Outlined Cross tattooed on their bodies. Also in Colombia people painted the outlined cross on their bodies and thus may have diffused the symbol and in this way – apart from the most acceptable migration via portable objects, like textiles and ceramics – the symbol may have travelled from one area to another. But I doubt diffusion from Mali to – for instance – New Caledonia.
Importantly, despite the graphical similarities across the globe I do not wish to claim the same symbolic content for all the manifestations of this widely distributed symbol, as is suggested by some authors (Sánchez P. 2008), especially when the symbol is referred to as the ‘Venus-Cross’. However, conveniently ignoring the symbolic content, the relatively large numbers and the rather dense concentrations of the Outlined Cross in the Study Area – the Pacific Seaboard of the Americas – may be an indication of diffusion of this symbol along the Pacific coastline. Yet, also in the Study Area parallel evolution cannot be ruled out, as – in fact – the manufacture of an Outlined Cross is easy as it starts with a simple equal-armed cross that is ultimately outlined. In conclusion, the Outlined Cross by itself is not a truly reliable element in establishing long distance diffusion. Therefore, in order to establish (possible) cases of diffusion along the Pacific Seaboard of the Americas – our Study Area – we need to consider other motifs.
PART II: DIFFUSION IN THE STUDY AREA
Because of and despite all these difficulties it will be interesting to explore the distribution patterns of some other rock art motifs along the Pacific Seaboard of the Americas and see if they may represent evidence of long distance diffusion. Of course there are many motifs and scenes that (quite understandably) developed independently in many areas without involving long distance diffusion. For instance, the icon of the ‘flute’ player is expected to occur in areas where the flute was played and therefore the icon originated independently in many parts of the Study Area. On the other hand, specific motifs may be indicative of diffusion over long distances, especially when a rock art design is exotic or unexpected within the scope of the repertoire of a rock art site or even rock art region. Such unexpected rock art motifs are classified here as ‘alien’ (but have of course nothing to do with influence from outer space). ‘Alien’ designs are more likely to have been imported from outside and may therefore represent instances of long distance diffusion.
The following (abstract) motifs or patterns will be discussed in this study (please notice that all these names are subjective interpretations): the Stepped Fret, the Pipette, the Grecian Border, the Barbed Geometric, the Row of Triangles, the Textile Pattern, the Dotted Cross and finally the Crescent Moon. Another complex figure, the ‘Saluting Anthropomorph’, will be the subject of another, separate study regarding possible diffusion along the Pacific Seaboard of the Americas.
The Stepped Fret
To demonstrate possible inconsistencies between cultural expressions and the rock art repertoire of an area I will now briefly focus on an area in Asia. Interestingly, stepped frets are a most prominent feature in the architectural art of many rock-cut tombs of Madain Saleh in the Arabian Peninsula. Towering over the impressive entrance of several tombs are truly huge, three-dimensional examples of a double unit stepped fret. Such stepped frets – called Assyrian Crow Steps – actually consist of one ‘negative’ (recessed), inverted (apex-down) double-unit stepped fret that is flanked on each side by a ‘positive’ (projecting) single stepped fret. Despite the overwhelming impression of those stepped frets on those gigantic Nabatean tombs in Saudi Arabia, I do not know of examples of the Stepped Fret in the rock art of the Arabian Peninsula (they might exist, though).
In contrast, the Stepped Fret design is a true hallmark of many cultures in the Study Area. Disregarding the many graphic manifestations of the Stepped Fret, especially the single unit and the double unit Stepped Fret are widespread in (the rock art of) the Americas (Figure 5). But there is a remarkable discrepancy. In the Southwest of North America especially the double unit of the Stepped Fret is most ubiquitous in rock art (lighter yellow drawings in Figure 5), but is absent or only sporadically present in other manifestations such as architectural art, ceramics and textiles. In contrast, in the rock art of Latin America the double unit is extremely rare, while the quadruple unit and single unit occur relatively more often (darker yellow drawings in Figure 5). Yet in the Desert Andes and the High Andes this design occurs rather frequently – sometimes abundantly – in architectural art, ceramics and textiles. Although I more fully addressed this discrepancy in an earlier publication (Van Hoek 2004), it is useful to briefly explain this difference here.
Figure 5: Map of the Americas showing a selection of sites featuring the ‘Stepped Fret’ motif.
In the Southwest of North America especially the double unit Stepped Fret is often called a cloud terrace. It is said to represent a powerful rain and fertility symbol. In rock art many cloud terraces are directly associated with lightning, rainbows and other rain-symbols. Those specific rain-elements seem to be lacking completely in Latin American rock art examples. However, in many Latin American cultures rain and water – and thus fertility – are closely associated with mountains. Especially Rain Gods, like the Mesoamerican Deity Tlaloc, dwelled at the tops of sacred mountains, and in several Latin American cultures their homes – the sacred mountains – were often imitated by enormous stepped pyramids. As the oldest known representations of the Stepped Fret date from the Cupisnique cultures of the coastal area of what is now northern Peru, it is possible (but not proven) that the Stepped Fret first originated in coastal northern Peru during the Andean Formative Period (roughly 2000 B.C to 200 B.C.) or even earlier, in the Preceramic Period (3500 B.C. to 2000 B.C.).
Therefore, I would like to suggest that Stepped Fret patterns in the rock art of the Southwest of North America originally represented mountains in the sense that they depict the ‘residences for (rain-)deities’. If the Cupisnique origin suggested by me is correct then the Stepped Fret possibly diffused from the Peruvian coast southwards, but also northwards. When travelling northwards the design ultimately reached the Southwest of the USA, but simultaneously lost its mountain-symbolism, which was replaced by rain- and fertility symbolism.
Another design with a general distribution more or less similar to the Stepped Fret is the Pipette (Figure 6). The distribution of the Pipette design in the Americas was comprehensively discussed by Will Russell and Aaron Wright in a paper published in 2008. Although the major concentrations are found in the Southwest of North America (2008: Fig. 1) they included in their discussion examples from central Chile as well (2008: Table 1). However, I question the Chilean examples they refer to. In my opinion they are not pipettes (Van Hoek 2018). Also, the area in which the Pipette has been recorded in rock art is much larger than suggested by Russell and Wright. The northernmost example of a possible Pipette petroglyph that I could find is located on Meadow Island off the coast of British Columbia, Canada (Hill and Hill 1974: 279). About 11.380 km further SE is Cueva Huenul in Neuquén, Patagonia, Argentina, where at least two rock paintings have been recorded that could be interpreted as Pipettes.
Figure 6: Map of the Americas showing a selection of sites featuring the ‘Pipette’ motif.
The biggest concentration of the Pipette design is found in the South Mountains, just south of Phoenix, Arizona, USA, where no less than 61 examples have been recorded, while Grapevine Canyon in Nevada, USA, has 24 petroglyphs of Pipettes (Russell and Wright 2008: Table 1). But also in the rock art of the Desert Andes of South America the Pipette design occurs (Van Hoek 2018). It has been reported at Socospampa in the Caravelí Valley, Chillihuay in the Ocoña Valley and at Toro Muerto in the Majes Valley, all in southern Peru. Further south, in northern Chile, examples have been recorded at Cruces de Molinos and Chapisca in the Lluta Valley, Chusmiza and at Tarapacá-47. Finally Mostny and Niemeyer illustrate Pipette designs from the Guaiquivilo region in central Chile and although some may be interpreted as stacked Outlined Crosses, a few designs seem to represent true Pipettes (1983: Figs 93 and 97). Also in western Argentina some examples have been recorded, for instance at Suripotrero and at Los Morteros (together with a most complex Stepped Fret design), both in the Fiambalá Valley.
The Grecian Border
Another ‘exotic’ design is found on a boulder AP2-052 at Alto de Pitis in the Majes Valley of southern Peru (Figure 7; location indicated in Figure 8)). This design certainly is alien within the repertoire of the Majes Style Rock Art. It features a biomorphic (?) image composed of a long rectangular, outlined body that is filled with an X and an ill-executed design that looks like the Grecian Border. However, it must be emphasised here that the patterns that are classified in this study as the Grecian Border have nothing to do with the classical Grecian Border and often do not even look like the classical Grecian Border. In this study all patterns shaped into a more or less repetitive, straight motif composed of hooks or curls will be classified as a Grecian Border, even when the repeated units are not exactly identical.
Figure 7: Petroglyphs on Boulder AP2-052 at Alto de Pitis, southern Peru. Photograph © by Maarten van Hoek.
The closest parallels of the Grecian Border at Alto de Pitis are found at the little known rock art site of Chiza in northern Peru, 420 km SE of Alto de Pitis. A much damaged boulder at this site features petroglyph of a row of five rectilinear spirals, framed by two parallel grooves, which give the impression of a Grecian Border design (see Figure 8; based on a photo by Carolina Rodríguez). About 40 km further SE, at the rock art site of Chillayza, is a petroglyph on a vertical outcrop cliff that may be related to the Grecian Border design.
Also north of Alto de Pitis are further parallels, but again, they are – as explained above – also of a rather differing layout. Two ‘similar’ petroglyphs are found at the rock art site of Cochineros in the Mala Valley, 600 km NW of Majes (Figure 8). Again, those two petroglyphs concern rows of spirals that only give the impression of the Grecian Border. Much further NW, on Boulder CH-ALG-1-007 at Quebrada de Algarrobos of the Chuquillanqui Complex in the Chicama Valley of northern Peru (595 km NW of Cochineros) is a rectilinear design that could be regarded as a single unit of the Grecian Border. Further NW, at the remote site of Quebrada de la Zorra in Jequetepeque (65 km NW of Chuquillanqui), is a boulder with a rather large anthropomorphic petroglyph (Figure 9). It is flanked on both sides by geometric patterns. On its left side is a row of two linked spirals and some faint traces of a second attempt. On its right side are two better developed rows of up to four linked spirals that have much in common with the Cochineros examples. Further NW still, near Cerro Mulato in Chancay, is the petroglyph site of Huaca Blanca, where one more complex, single unit of Grecian Border has been recorded by me (as well as an Outlined Cross).
Figure 8: Map of the Americas showing a selection of sites featuring the ‘Grecian Border’ motif.
Figure 9: Petroglyphs on a boulder at Quebrada de la Zorra in Jequetepeque, northern Peru. Drawing © by Maarten van Hoek, based on a photograph by Nolish Percy Rojas Palomino.
But there are much more distant parallels in the Study Area. About 5480 km NW of Quebrada de la Zorra in Peru and about 7140 km of Chiza in Chile is the site of La Proveedora in northern Mexico. At this important site are several petroglyphs that also are composed of rows of spirals. Although most examples are composed of interlocking elements (like curvilinear and rectilinear spirals), at least one petroglyph is much similar to the Cochineros examples in Peru (see Figure 8). However, some petroglyphs of rectilinear and interlocking designs at La Proveedora also look like the ill-formed Grecian Border at Alto de Pitis (Ballereau 1988: Lámina XVII-b). At Tehuelibampo in Mexico – 480 km SE of La Proveedora – is a petroglyph that may also be related to the Grecian Border design. Importantly, in the same Tehuelibampo area are also petroglyphs of the Outlined Cross (single and stacked examples) and of the Textile Pattern (still to be discussed). It may be telling that motifs that are possible contenders to demonstrate diffusion often occur at one and the same site.
Another characteristic element abundantly occurring at La Proveedora – and involving one even more appropriate contender regarding long distance diffusion – is the Barbed Geometric. This specific pattern is found at several other rock art sites (as petroglyphs and painted examples) in northern Mexico and in the SW of the USA (which actually may be considered as one rock art region). Especially Arizona – bordering northern Mexico – has several rock art sites that feature the Barbed Geometric. Moreover, it also occurs on Anasazi ceramics and therefore the Barbed Geometric style could easily have been diffused via ceramics or textiles.
However, a problem is that it is often hard to distinguish between a linear Barbed Geometric and a Row of Triangles, especially because both patterns often occur at one site, like at Parowan Gap in Utah, USA. When a linear pattern is composed of relatively large and more clearly shaped like equal-sided triangles, it will not be considered to represent the Barbed Geometric. Moreover, the Row of Triangles always involves one more or less straight line or a few straight parallel lines. The Row of Triangles will be discussed at the following section.
Indeed, many geometric patterns at La Proveedora are composed of lines that are ‘barbed’ with rows of very short projections (short, straight lines or small triangles, often shaped like shark’s teeth). Such patterns are classified here as Barbed Geometrics. There are numerous variations of the Barbed Geometric at La Proveedora. One of those abundant Barbed Geometrics is a unique nested (double) rectilinear S-shape (Figure 10-1) that is found – together with other petroglyphs – on a sloping surface of a large boulder (Ballereau 1988: 402; Lámina I-f). The open end of a third nested element – without ‘barbs’, though (unfinished?) – is pointing upslope. Immediately below that boulder is another large boulder with an even more complex Barbed Geometric (Ballereau 1988: 402; Lámina I-a).
Figure 10: Map of the Americas showing a selection of sites featuring the ‘Barbed Geometric’ motif.
Surprisingly, true Barbed Geometric patterns are rather rare in South America. And also in this continent it is often hard to distinguish between the Barbed Geometric and the Row of Triangles. Another problem is that some purportedly ‘barbed’ petroglyphs may even involve wing-symbolism, as the short lines (or small triangles in several cases) may represent the feathers of a bird’s wing in (mainly anthropomorphic) figures. For instance, bird-symbolism may also have been aimed at in a most complex, stacked and almost completely ‘barbed’ figure on Bloque 6 (featuring two other, simpler ‘barbed’ petroglyphs and one Outlined Cross) at Site Cxa-E7 at Conanoxa in the Camarones Valley of northern Chile (Van Hoek 2016: Fig. 77). At the same site is another boulder with a petroglyph of two very faint, concentric squares with all eight sides being ‘barbed’; the very small ‘barbs’ all pointing in the same direction (of course only when viewing each pair of parallel lines). The direction of the Conanoxa ‘barbs’ is an important point in this discussion.
Most importantly however, at site Cxa-E7 at Conanoxa María Francisca Fernández Donoso (2013: 22: Fig. 4 ) also reported a panel with a petroglyph that is almost exactly the same as the example at La Proveedora (7125 km NW of Conanoxa!). The resemblance is most striking, especially as there is also a third nested element without ‘barbs’. Importantly, the ‘barbs’ all point in the same direction (Figure 10-2). As the Conanoxa example may be regarded to be alien to the rock art repertoire at Conanoxa and indeed for the whole of the Desert Andes, it may well concern an imported design that diffused southwards from northern Mexico (not vice versa). However, apart from possibly representing an instance of parallel development, there may possibly be another, closer source.
About 82 km SE of Conanoxa is the impressive petroglyph site of Ariquilda in the Quebrada de Aroma with many sophisticated and idiosyncratic images. The petroglyph relevant in this discussion is – as far as I know – also unique for Ariquilda and therefore may be alien. It is found about ten meters above the valley floor on a huge, almost vertical cliff panel that has many more petroglyphs. Although there are some subtle differences, it has almost the same layout as the nested rectilinear S-shape at Conanoxa (Figure 10-3). The differences are important, however. The first discrepancy is that the Ariquilda example has three parallel lines, of which only the outer lines are barbed, while the second, more important, divergence is that the rather irregularly placed barbs are pointing in opposite directions.
There are now several possibilities. It is possible that the Ariquilda and Conanoxa Barbed Geometrics diffused from La Proveedora in northern Mexico to the Desert Andes It is also possible that the Barbed Geometric pattern only diffused from Ariquilda to Conanoxa or vice versa and that the Proveedora example is simply an instance of parallel evolution. I already noticed that designs (thus also the Barbed Geometric) could easily have migrated on a textile or ceramic. In this respect it is significant that excavations of Tomb 87 at the prehistoric cemetery of Tarapacá-40 yielded a fragment of a textile with an image of a (bicephalic?) snake that (although incomplete) is almost similar to the Barbed Geometrics at Ariquilda and Conanoxa (Figure 10-4). Importantly, Tarapacá-40 is only 40 km south of Ariquilda.
At this stage the direction of the ‘barbs’ indeed becomes an issue. The Tarapacá textile design is clearly a snake, but importantly, its ‘barbs’ are pointing in opposite directions. Therefore it could be more related to the Ariquilda petroglyph, which, although it does not feature a snake’s head, also has its ‘barbs’ projecting in opposite directions (which is quite common in snake imagery in the Desert Andes). The Ariquilda example therefore could well symbolise a snake. A bicephalic snake petroglyph displaying almost the same S-shape (without ‘barbs’, though) as the Conanoxa Barbed geometric is found on Panel YON-099 at Yonán in northern Peru (Figure 10-5) (650 km NW of Conanoxa). In general this S-shape, even when rectilinearly depicted in rock art, may represent the curling of a snake when it moves across the soil.
On the other hand, the Conanoxa Barbed Geometric (and also the barbed square petroglyph at the same site) have all the ‘barbs’ pointing in the same direction. Moreover, the example at Conanoxa is broken off at both ends and it is most uncertain if the (once complete?) image ever depicted a snake. In this respect the Conanoxa Barbed Geometrics have more in common with the Barbed Geometrics at La Proveedora in Mexico. Therefore, regarding the Barbed Geometric there still is a possibility of diffusion between the two Americas.
Rows of Triangles
Possibly related to the Barbed Geometric is the Row of Triangles, which has been reported at many sites in both North and South America (Figure 11). This pattern always comprises one more or less straight or slightly undulating base-line (or a few straight, parallel base-lines) that features a row of (mainly) fully pecked or fully painted-in, equal-sided triangles. In some case this base-line is missing. The great majority of the Row of Triangles is horizontally orientated, but vertical and diagonal positions also occur. The apexes of the triangles may point upwards or downwards. The number of triangles may vary from two to more than ten. In this respect Boulder PAL-159 at Palamenco in the north of Peru is exceptional as it has a petroglyph of a Row of Triangles with more than twenty downward pointing triangles, while the line almost completely encircles the boulder (Figure 12). Moreover, at Palamenco there is a remarkable concentration of the Row of Triangles pattern.
Figure 11: Map of the Americas showing a selection of sites featuring the ‘Row of Triangles’ motif.
Figure 12: Boulder PAL-159 at Palamenco, northern Peru. Photograph © by Maarten van Hoek.
Often Rows of Triangles are found in groups that are parallel arranged. These groups are mainly horizontally orientated, like an example from Utah, USA, where even up to four painted parallel rows have been reported. In some cases a cluster of Rows of Triangles has deliberately so closely been configured that a complex pattern of stacked, fully pecked and un-pecked triangles emerges. Fine (painted) examples have been recorded in southern Arizona, USA. The possibly northernmost example of the complex Row of Triangles is found at Wees Bar, Idaho, USA. It is a complex petroglyph of four stacked Rows of Triangles. On the same panel are several other pecked (rows of) triangles. Also on Panel CH-ALG-2-025 at Quebrada de Algarrobos of the Chuquillanqui Complex in the Chicama Valley of northern Peru (6800 km SE of Wees Bar) is an unexpected example of four horizontally arranged Rows of (pecked and un-pecked) Triangles (Figure 13). Although there are more configurations of lines and pecked triangles on this panel, it seems to be the only true example of the complex Row of Triangle pattern in the Chicama Valley, although at the site of Cerro Negro in the same valley are four Rows of Triangles on a large boulder, two of which are opposed. The southernmost example of the (complex) Row of Triangles in the Study Area is the petroglyph of six stacked Rows of (three) Triangles at the rock art site of Cajón de Calabozos in southern Chile (9900 km SE of Wees Bar as the crow flies; 11.240 km when following the coastline).
Figure 13: Panel CH-ALG-2-025 at Quebrada de Algarrobos, the Chuquillanqui Complex, Chicama, northern Peru. Photograph © by Maarten van Hoek.
In some cases the triangles are outlined. At La Proveedora in northern Mexico is a Row of Triangles composed of five outlined (concentric) triangles and three outlined triangles filled with a partially pecked triangle. At Picacho in Arizona, USA, is at least one panel with an outlined petroglyph of the Row of Triangles. Another panel in this area has six petroglyphs of the Outlined Cross. In Chalfant Valley in California, USA, is a Row of (eight, double lined) Triangles, each with a dot in the centre. In the Mojave Desert of California is a rock art site with a much sloping panel with two opposed Rows of (five) Triangles in a vertical position; all apexes pointing outwards. At Nine Mile Canyon in the NE of Utah, USA, are at least one fully pecked petroglyph of the Row of Triangles and one outlined example comprising three triangles. The petroglyphs at Badger Creek in Oregon include the northernmost examples of the Row of Triangles that I am aware of (as the crow flies 10.300 km NW of Cajón de Calabozos in southern Chile, the southernmost site in the Study Area; 11.730 km when following the coast).
The southernmost example of the outlined Row of Triangles is the vertically arranged petroglyph of a Row of (six) Triangles on a boulder at Quebrada de Tambores in northern Chile (Figure 14). Much further north, at Toro Muerto in southern Peru is Boulder TM-Nw-052 with – on Panel B – a horizontally arranged Row of Triangles (Figure 15). All eight, downward pointing, outlined triangles have a centrally placed dot (compare this with the Chalfant petroglyphs). On Panel A of this boulder is a configuration of two interlocking Rows of Triangles. An arrangement – almost similar to the design on TM-Nw-052 – Panel B – is found on Panel PAL-113B at Palamenco in northern Peru (1040 km NW of Toro Muerto). Next to this design is an isolated, outlined triangle that seems to have been transformed into a bird. This possibly bird related transformation brings me to discuss some petroglyphs at Cochineros in southern Peru.
Figure 14: Boulder at Quebrada de Tambores, northern Chile. Drawing © by Maarten van Hoek, based on a photo by Mario Giorgetta.
Figure 15: Boulder TM-Nw-052 at Toro Muerto, Majes, southern Peru. Photograph © by Maarten van Hoek.
At Cochineros in the Mala Valley of southern Peru are several bird petroglyphs that are composed of pecked, equal-sided triangles. In many cases those bird images are found among other petroglyphs. In a few cases however (rows of) fully pecked triangles seem to have bird attributes added resulting in weird-looking configurations (for instance on Boulders COC-005, 007 and 014). One petroglyph (on Boulder COC-001) stands out in this respect (Figure 16). Two Rows of (each three) Triangles are joined at a more or less right angle, while the triangle at one end features a downward pointing, curled tail and a distinct bird’s neck and head with open beak that is pointing upwards against the slope of the huge boulder. I am convinced that the bird properties have been added at a later stage. Surprisingly, the same arrangement of two Rows of (each three) Triangles – joined at a sharp angle, but without bird attributes – occurs on a panel at La Proveedora in northern Mexico (Ballereau 1988: Lámina IV-d), 6140 km NW of Cochineros.
Figure 16: Detail of the petroglyph on Boulder COC-001 at Cochineros, Mala, southern Peru (slope indicated). Drawing and photograph © by Maarten van Hoek.
Finally, other sites in the Study Area where the Row of Triangles pattern (painted or pecked) is found, are (this is just a selection): Tamentica, Tarapacá, Chusmiza, Ariquilda, Chillayza in Chile; Miculla, Quebrada de Paypay, Cerro Mulato, Los Boliches in coastal Peru (and outside the Study Area at Achaypiña and Anccomayo in highland Peru); Pacheco Creek (Chihuahua) in Mexico; Quail Point (Figure 17) and Snake Canyon in Arizona; Richfield, Fool Creek, Natural Bridges, Parowan Gap, Cub Creek, Nine Mile Canyon and Cedar Mesa in Utah; Snake Canyon, Hiko Spring, Amphitheater Site, Mouse’s Tank and Grapevine Canyon in Nevada; Gila Cliffs and Three Rivers in New Mexico and Badger Creek in Oregon, USA.
Figure 17: Panel at Quail Point, Arizona. Notice the positive (pecked) and negative (unpecked) triangles. Photograph © by Guy Starbuck.
Finally and surprisingly, petroglyphs of Rows of Triangles have also been reported outside the Study Area on Ilha do Campeche and Ilha dos Corais, both small islands that are located a short distance off the Atlantic Seaboard of SE Brazil. Interestingly, also petroglyphs looking like the Textile Pattern have been recorded on Ilha do Campeche. Those Textile Pattern petroglyphs on Ilha do Campeche are comparable, but they are rectangular, not square. This brings me to the next motif to be discussed.
The Textile Pattern
There is one specific petroglyph in the Study Area that I regard to be a very good example of an ‘alien’ rock art motif within the geographical context of the Desert Andes. Having seen almost every image at the most extensive petroglyph site of Alto de Pitis in the Majes Valley of southern Peru (50 km inland), I was struck by one specific, square motif that in my opinion is unique not only for Alto de Pitis, but also for the whole of the Majes Style Rock Art repertoire and possibly for the whole of the Desert Andes as well.
In general, outlined, square motifs are rare in the Majes Style Rock Art repertoire and indeed in the whole of the Desert Andes. Some squares in Desert Andes rock art are empty, while the majority has a modest interior pattern, usually comprising an X-cross plus sometimes some dots or other simple elements. However, the square at Alto de Pitis (Figure 18) is filled with a most complex pattern, which, in the literature about the rock art of the Southwest of the USA, is called the Textile Pattern. Of course it is by no means certain that the ‘Textile Pattern’ of Alto de Pitis also symbolises a textile. It is not even certain if the pattern diffused from another area to the Majes Valley, but – in the context of this study – it is worthwhile to more profoundly investigate the distribution of the Textile Pattern in the Study Area, keeping in mind that comparable (but different) patterns occur elsewhere, for instance on Ilha do Campeche off the coast of SE Brazil (2800 km SE of Alto de Pitis) and even at New Caledonia (12.500 km west of Alto de Pitis).
Figure 18: Panel AP-066A at Alto de Pitis, Valle de Majes, southern Peru. Photographs © by Maarten van Hoek.
Of course it is most unlikely to find an exact copy of the Alto de Pitis Textile Pattern, but there are several examples in the rock art repertoire of the Pacific Seaboard of the Americas that are almost identical (Figure 19). Interestingly, the petroglyph at Alto de Pitis is – to my knowledge – one of the southernmost examples of the square Textile Pattern in the Study Area. I only know of two similar patterns filling two more rectangular frames on a boulder in the NW of the province of San Juan in Argentina, roughly 1500 km SSE of Alto de Pitis. However, several other examples have been recorded north of Alto de Pitis, but in many cases they are not exactly square as well.
Figure 19: Map of the Americas showing a selection of sites featuring the ‘Textile Pattern’ motif.
The other parallels that I know of will be discussed from south to north. Yet, it takes travelling about 2450 km north from Alto de Pitis before we encounter the first example (skipping numerous rock art sites), which is located at near Támesis in Colombia, 170 km inland. Although there are more internal lines than with the Alto de Pitis example, the design is almost similar (1 in Figure 19). In Colombia a few more examples have been recorded, including a similar, painted example at Cerro Azul, Chiribiquete, with its internal pattern composed of small dots however and the petroglyph at Cerro Quinini; the Piedra del Gritadero. Also outside the Study Area, in neighbouring country of Venezuela some petroglyphs of the Textile Pattern have been reported, like the couple at La Acequia, Barinas. Further NW, in Nicaragua, at least two similar petroglyphs have been reported at the rock art site of La Cueva de la Bruja, Carazo (1340 km NW from Támesis). One is a rather badly executed, more rectangular and irregular example (2 in Figure 19), while the other is completely fringed, resembling a solar motif.
But the major concentration of the square Textile Pattern is found distributed along the Pacific Seaboard of mainland Mexico (thus excluding the Baja California peninsula here). Two Textile Patterns are known from the rock art site of Cerro de Coamiles, Nayarit (2300 km NW from Carazo; Site 3 in Figure 19). One of the at least two Textile Pattern petroglyphs at Cerro de Coamiles is almost identical to the example at Alto de Pitis (located 5540 km further SE). The other – still much similar example – has a row of short, parallel lines from the bottom line of the square (Figure 20). At the ‘nearby’ rock art site of Cerro en Otates, Nayarit, there is at least one petroglyph of a possibly anthropomorphised Textile Pattern. On the same panel are several comparable motifs.
Figure 20: Map of Cerro de Coamiles, showing two petroglyphs of the ‘Textile Pattern’ motif. Drawings © by Maarten van Hoek, based on several sources.
Further NW along the coastal strip the Textile Pattern occurs at the rock art sites of La Noria, Las Pintadas and at Cerro de Chivo (site 4 in Figure 19), in Sinaloa (all about 230 km NW of Coamiles). At Los Naranjos, Sinaloa (170 km NNW of La Noria), is a Textile Pattern that seems to have been anthropomorphised by the addition of a semi-circle with internal dot and external ‘rays’, which possibly represents the head (5 in Figure 19). At the site of Cerro de la Máscara just north of the town of El Fuerte, Sinaloa (220 km NW of Los Naranjos), are two petroglyphs of the Textile Patterns; one – at the site of La Zorra, Group 1, Panel 1 – has curls emerging from the corners (6 in Figure 19) and one without curls. At this important site (with many more non-square textile patterns) also four petroglyphs of the Outlined Cross have been recorded. Importantly, the Textile Patterns of Cerro de la Máscara (and of several other sites in this coastal area) completely mingle with the local and regional imagery. Thus, the Textile Pattern certainly is, unlike at Alto de Pitis, no alien element in the rock art of western Mexico. This may mean that this is the area – and possibly areas further NW – where the Textile Pattern originated.
Further NW, in Tehuelibampo area, Sonora (110 km NW of Cerro de la Máscara), are at least two petroglyphs of the Textile Pattern (one with possibly facial features: 7 in Figure 19), which, again, do not look alien within the context of the rock art repertoire of this area. Also in Sonora (480 km NW of Tehuelibampo) is the – by now – well known rock art site of La Proveedora. Also at this extensive site are at least eight comparable petroglyphs of the Textile Pattern, of which one in particular (8 in Figure 19) looks like the example at Alto de Pitis (6720 km SE). About 240 km inland from and east of La Proveedora is a rock art site at the Rancho Mababi, still in Sonora, where a fine petroglyph of the Textile Pattern seems to have been recorded. Further east still at the Arroyo de los Monos, 200 km ESE of Rancho Mababi and 35 km south of the town of Nuevo Casas Grandes, Chihuahua, Mexico, is a simple petroglyph of the Textile Pattern (9 in Figure 19).
Travelling north into the Southwest of the USA the Textile Pattern emerges again, for instance at South Mountains in Arizona, 300 km north of La Proveedora. In this mountain range at least two petroglyphs of the Textile Pattern have been recorded, one of which (10 in Figure 19) is almost similar to the example at Alto de Pitis in Peru (about 7000 km SE). The other example is rectangular and has only partially the same interior design. Further NE, still in Arizona, in the valley of the Little Colorado River (approximate location: Site 11 in Figure 19), is a rock panel with four simple petroglyphs of the Textile Pattern (somewhat resembling the example at Arroyo de las Monos in Chihuahua, northern Mexico), while unfortunately a fifth and possibly more resembling design has partially flaked off and therefore cannot be checked properly (indicated by the blue arrow in Figure 21). Also at Three Rivers, New Mexico, a petroglyph of a simple Textile Pattern has been recorded. The northernmost example of the Textile Pattern that I know of is a petroglyph at Waterflow, New Mexico, USA, (12 in Figure 19) 500 km NE of South Mountains in Arizona and about 7000 km from Alto de Pitis in Peru.
Figure 21: Petroglyphs of ‘Textile Patterns’ recorded somewhere in Little Colorado Canyon, Arizona, USA. Drawing © (and distorted) by Maarten van Hoek, based on a photograph by Dennis Roshay.
There is one final, but slightly different ‘Textile Pattern’ design that possibly fits in this specific discussion as well. It concerns an outlined, circular design with almost the same internal pattern as its square relative. I recorded one (and possibly unique example for the Desert Andes?) petroglyph of such a circular Textile Pattern at Los Boliches in northern Peru (13 in Figure 19), while 4160 km further NW, at the rock art site of Coamiles in western Mexico is a petroglyph with an almost similar, yet larger design (3 in Figure 19), while there are some more circular but less similar examples at Coamiles.
The Dotted Cross
In rock art randomly and linearly arranged dots (not cupules) occur more often than groups of dots that form specific patterns. At Cerro Mulato, an important petroglyph site in northern Peru, are several examples of geometric figures composed by dots; especially triangular arrangements. One geometric arrangement on Boulder CMn-227 concerns a large but irregular cross-shape (an X-shape, rather). It has four arms; two arms are composed of three rows of small dots, while the other two arms have three rows of dots (Figure 22). On the same panel is a triangular shape composed of dots (a hallmark of Cerro Mulato) and some other petroglyphs. Only a few meters away is Boulder CMn-420 that has a similar dotted cross, also composed of four arms, but two arms only have two rows of dots. I know of only one other clear parallel in the Study Area. It is located roughly 4250 km NW of Cerro Mulato.
Figure 22: Boulder at Cerro Mulato, northern Peru. Photograph © by Maarten van Hoek. Inset: Petroglyph of a Dotted Cross near Mina, Nuevo Leon, Mexico. Drawing © by Maarten van Hoek, based on a photograph by an unknown author.
A boulder (or outcrop?) near Mina, Nuevo Leon, Mexico, features an almost similar cross-shape that has four arms, each composed of three rows of small dots. However, it has more the shape of a Christian Cross (Figure 22: inset). On the same boulder is a linear band composed of four rows of dots, while other panels in the area also feature arrangements composed of small dots. Despite the significant resemblance between the three crosses, it is more likely that they represent instances of parallel development, rather than a case of diffusion.
The Crescent Moon
Another rare rock art motif that has been recorded at most distant sites in the Study Area is the Crescent Moon (Figure 23). However, it is most uncertain if this motif indeed has anything to do with the moon, but it is possible, especially in the coastal area of northern Peru where the moon was worshipped for many generations. In general the Crescent Moon motif is horizontally arranged with its ends pointing upwards. This motif mainly occurs fully pecked, but very rarely outlined examples occur as well. The Crescent Moon is relatively abundant at Cerro Mulato in northern Peru (see Figure 23.1), where at least seven examples have been recorded on the more than 500 petroglyph boulders. One of those boulders also has an Outlined Cross petroglyph (Figure 24). At neighbouring Huaca Blanco I recorded two examples; one fully pecked petroglyph and the other outlined.
Further south, in the valley of the River Jequetepeque, is the site of Montegrande (unfortunately completely destroyed recently) where once one outlined and four fully pecked examples were reported. In the same valley are Yonán – with at least three fully pecked examples (see Figure 23.2) – and a panel reported at a site (location unknown to me) that has two small Crescent Moon petroglyphs as well as a small anthropomorph with a crescent moon on its head (a Tumi?) (see Figure 23.3). Further south still, in the Chicama Valley, is Cerro Negro with two fully pecked examples (Figure 25) and at Chuquillanqui – further east in the same valley – is a large outlined example on Panel CH-ALG-2-034, superimposing a large anthropomorphic figure. In the same valley but nearer the coast and east of the town of Ascope is a boulder with a similarly large, outlined Crescent Moon petroglyph. Possible Crescent Moon petroglyphs have also been reported at Sausal in the Chicama Valley, at Alto de la Guitarra in Virú and at Palamenco in Santa. An outlined crescent is topping an anthropomorphic figure at the rock art site of Antapucro, Lurín, Lima. It proves that the main concentration of the Crescent Moon is in northern Peru.
Figure 23: Map of the Americas showing a selection of sites featuring the ‘Crescent Moon’ motif.
Figure 24: Boulder at Cerro Mulato, Chancay, northern Peru. Photograph © by Maarten van Hoek.
Figure 25: Detail of some of the many petroglyphs on a large boulder at Cerro Negro, Chicama, northern Peru. Photograph © by Maarten van Hoek.
However, the Crescent Moon motif has also been recorded in Chile. At the rock art site of Chusmiza in northern Chile (1820 km SE of Cerro Mulato) is a boulder with two inverted Crescent Moon petroglyphs (symbolising the waxing and/or waning moon?), a fully pecked disc (full moon?) and a solar petroglyph (see Figure 23.4). Possibly the total of the Crescent Moon motif in Desert Andes is no more than 25. The fully pecked disc motif (the full moon?) – reported at Chusmiza in northern Chile – also occurs at Cerro Moquegua and Yonán in Peru and – surprisingly in the context of this study – also at La Proveedora in Mexico.
Again, the Crescent Moon motif is also found in North America. At the Rabbit Site in Arizona, USA, is a fully pecked Crescent Moon petroglyph in a vertical position. It convincingly looks like a crescent moon. But more importantly, about 5420 km NW of Cerro Mulato in Peru and 7230 km NW of Chusmiza in Chile is La Proveedora in the far north of Mexico. This most extensive site has no less than fourteen fully pecked examples of the Crescent Moon motif (Figure 26) and some other, possibly related ‘arched’ petroglyphs. A difference is that not all examples at La Proveedora are horizontally arranged (with their ends pointing upwards). But the resemblance is still striking. It is also striking that – repeatedly – La Proveedora features several rock art motifs that are also found in most distant sites in the Study Area.
Figure 26: Petroglyph on a boulder at La Proveedora, Sonara, Mexico. Photograph © by Earl Maynard.
The Case of the Aguada Diffusion
The following discussion proves how differently rock art researchers think about diffusion of rock art images and raise arguments against it. In a well reasoned paper Professor Andrés Troncoso Mélendez and Dr Donald Jackson Squella analyse and discuss two biomorphic petroglyphs in central Chile (2010: Figs 4 and 5) that – so they argue – possibly are testimonies of ‘Aguada’, a ‘culture’ that once occupied northwest Argentina on the western side of the Continental Divide that runs north-south across the Andes of South America. Troncoso & Jackson argue that those two images ‘travelled’ (or diffused) across the Andes.
This study convincingly demonstrates that especially specific complex rock art images, like the ‘Camunian Rose’, sometimes are (possible) indications of contacts between distant areas. Therefore the comments on Troncoso and Jackson’s article by Robert Bednarik, the editor of Rock Art Research, are surprising and surprisingly negative. Bednarik uses (only!) two petroglyphs from the ancient (Formative Period) rock art site of Alto de la Guitarra in northern Peru and some petroglyphs from British Colombia, Canada, to demonstrate a purported ‘flaw’ in the reasoning by Troncoso & Jackson (Troncoso & Jackson 2010: 53-56). However, in my opinion the ‘flaw’ is on Bednarik, not on Troncoso and Jackson.
Notably, Bednarik offers only two examples from the enormous rock art repertoire of the Andes. It is moreover most unlikely that Bednarik has ever surveyed the two petroglyphs at Alto de la Guitarra himself, as he uses two illustrations by Núñez Jiménez (1986: Figs 625 and 624 respectively) to comment on Troncoso and Jackson. In 2009 I have informed Bednarik about the fact that the illustrations of Núñez Jiménez are often highly unreliable or even demonstrably completely incorrect (Van Hoek 2011). I am also pretty sure that Bednarik never visited Alto de la Guitarra. Therefore Bednarik should have mentioned in the caption of his Fig. 1 that he has not seen these petroglyphs in the field at Alto de la Guitarra and, moreover, he should have said that the renderings by Núñez Jiménez are often questionable or incorrect. Strangely, in this respect (!) Bednarik does not seem to have read my 2007-publication in ‘his’ high-quality Journal – Rock Art Research (!) – in which I recommend that authors mention in the text or caption when they themselves have not seen a specific rock art image that they present, in situ, whether textual or graphical (Van Hoek 2007: 257). This recommendation applies to every author and thus also includes Bednarik, the editor of Rock Art Research, who failed to mention that fact.
Yet, Bednarik still based his arguments ‘against’ the diffusion theory by Troncoso & Jackson also on two inaccurate illustrations by Núñez Jiménez. It is therefore impossible for Bednarik to firmly establish a graphical similarity between the petroglyphs at Alto de la Guitarra and the Argentinean/Chilean images. In this respect (!) I personally find it rather petty of Bednarik when he continues to write that ‘… perhaps there is more to the Andean connection than Troncoso & Jackson suspected; perhaps the ‘Aguada people’ migrated from Peru? Or vice versa?’. In this way Bednarik belittles the legitimate attempts of Troncoso & Jackson to establish a case of diffusion from the ‘Aguada’ heartland in Argentina to central Chile. In this way Bednarik also ignores the possibility that not the ‘Aguada people’ migrated from Peru (this I find a typical sneer of Bednarik, who often uses sarcasm to ‘prove’ his point), but that the graphic content of an ancient image diffused across the Andes (from ‘Peru’? – why not?; in several cases ancient imagery did travel across the Andes) to ultimately appear in a different format in ‘Aguada’ iconography and then to diffuse across the Andes to emerge again in central Chile. The scenario presented by Troncoso & Jackson is not at all unthinkable or far fetched, especially as in northern Chile more graphical data have been reported (both textiles and rock art) that seem to substantiate the diffusion hypothesis by Troncoso & Jackson.
Finally, I personally think that in his comments Bednarik confuses two levels of reasoning. His ‘broader perspective’ in his Comments is well underpinned and I am convinced that – in global (rock) art – many examples of ‘similar’ feline-looking creatures can indeed be traced back to some sort of ‘fear-generated’ archetype. But I personally find it far too far fetched of Bednarik to even suggest that the ‘Canadian petroglyphs are perfect further candidates for ‘Aguada status’ (those Canadian petroglyphs he is referring to are found more than 10.200 km NW of the Aguada heartland on Vancouver Island). Even if those Canadian examples would have diffused to Argentina (which I doubt very much), I personally find that this way of reasoning belittles the initial and sincere intention of Troncoso & Jackson.
The ‘broader perspective’ monologue by Bednarik would have been most useful if only he had omitted his ‘Comments’ on the diffusion-issue raised by Troncoso &Jackson (but then his monologue would not be applicable as ‘Comments’). In other words, the archetype-issue is completely different to the issue of ‘Images that travel’ across the Andes. Both issues are legitimate, but as they operate on two completely different levels, I am of the opinion that the macro may not be (ab)used in this respect (!) by Bednarik to ‘comment’ the micro. Therefore, the idea postulated by Troncoso & Jackson that the two Chilean petroglyphs may be related to the ‘Aguada’ iconography and diffused across the Andes is very legitimate and deserves further (regional) investigation.
This study demonstrates that specific symbols or motifs in rock art can ‘travel’ considerable distances across the globe. Some rock art motifs even have an unexpected and unexplained enormously wide distribution, especially the Outlined Cross. The problem is however that such a ubiquitously distributed motif may have developed autonomously in several areas and then may have diffused regionally. Other motifs – like the Barbed Geometric – are found in very limited numbers but still thousands of kilometres apart and may therefore demonstrate diffusion (or not). In this respect I am aware that, especially regarding rock art motifs in the Desert Andes manufactured in pre-Columbian times, there actually is no informed evidence available. Only graphical parallels are available.
Yet, this paper demonstrates that long-distance diffusion of certain rock art motifs may also have happened along the Pacific Seaboard of the Americas. For instance, the Stepped Fret design and the Outlined Cross motif are often found in the ‘Aguada heartland’ (for instance at the Parque Diaguita site and at Talampaya in La Rioja, both in western Argentina), in the rock art and in the geoglyph art of Chile (Van Hoek 2001) and at several sites in Peru (for instance at Cerro Mulato in northern Peru), but this study also revealed that the same motifs are ubiquitously present in the rock art of the Southwest of the U.S.A. (Van Hoek 2004). It would be unscientific to regard all the Stepped Fret patterns and Outlined Crosses of each specific region in the Americas to be the result of independent and parallel development. It is, especially in view of the well established long-distance pre-Columbian contacts in the Americas, more logical that the wide distribution is also explained by diffusion.
This study also discussed other specific motifs, like the Pipette, the Grecian Border, the Barbed Geometric, the Row of Triangles, the Textile Pattern, the Dotted Cross and the Crescent Moon. All those motifs may have diffused along (parts of) the Pacific Seaboard of the Americas. Although in many cases there is no factual proof, it seems quite well possible that those rock art motifs travelled from one area to another area (although establishing the birthplace of each motif as well as the direction of the possible diffusion of each motif will remain problematic).
I am indebted to Guy Starbuck and Earl Maynard for their permission to use and publish their photographs, two of which have been included in this study (Figures 17 and 26 respectively). I am also grateful to Mario Giorgetta for his permission to use and publish his photos, one of which has been used to base my Figure 14 upon. The copyright of those photographs remains with the authors.
Ballereau, D. 1988. El arte rupestre en Sonora: petroglifos en Caborca. Trace. Vol. 14; pp. 5 – 72; CEMCA. Mexico.
Dupuy, C. 2001. Les gravures de chars et d’entrelacs dans l’art rupestre de l’Adrar des Iforas et leurs homologues sahariens. Cahiers de l’AARS . Vol. 6; pp. 25 – 29.
Dupuy, C. 2009. Une gravure rupestre dans l’Adrar des Iforas (Mali) identique aux «roses camuniennes» du val Camonica (Italie). Cahiers de l’AARS . Vol. 13; pp. 117 – 126.
Dupuy, C. and S. Searight. 2005. Les signes gravés à Issamadanen (Mali) et à Imaoun (Maroc), au regard de l’art abstrait ibérique. In : Roches ornées, roches dressées. Colloque en hommage à Jean Abelanet. Université de Perpignan. Pp. 97 – 108.
Dupuy, C. 2010. Trois époques de gravure rupestre en Adrar des Iforas (Mali). International Colloquium: ‘The Signs of Which Times?’ Chronological and Palaeoenvironmental Issues in the Rock Art of Northern Africa. pp. 47 – 69. Royal Academy for Overseas Sciences. Brussels.
Encinas, L. 2014. Las manifestaciones gáfico-rupestres del encuentro de dos mundos en el norte de México. In: Rupestreweb
Farina, P. 1997. The motif of the “Camunnian Rose” in the Rock Art of Valcamonica (Italy). In: TRACCE – Online Rock Art Bulletin, Italy.
Farina, P. 1998. The motif of the “Camunnian Rose” in the Rock Art of Valcamonica (Italy). In: TRACCE – Online Rock Art Bulletin, Italy.
Fernández Donoso, M. F. 2013. Caracterización del Arte Rupestre de las Terrazas de Conanoxa, Valle de Camarones, extremo norte de Chile, Andes Centro Sur. Práctica profesional. Universidad de Chile. Santiago de Chile.
Hill, B. and R. Hill. 1974. Indian Petroglyphs of the Pacific Northwest. University of Washington Press.
Monnin, J. and C. Sand. 2015. Essai de synthèse sur les pétroglyphes calédoniens. Kibo, le serment gravé. Archeologia Pasifika. Vol. 5.
Mostny Glaser, G. & H. Niemeyer Fernández. 1983. Arte Rupestre Chileno. Serie El Patrimonio Cultural Chileno. Colección Historia del Arte Chileno. Publicación del Departamento de Extensión Cultural del Ministerio de Educación.
Sánchez P., D. 2008. El símbolo de Venus en el arte rupestre de Perú, Chile y norte de Argentina. In: Rupestreweb.
Searight, S. 1999, Imaoun: un site de gravures rupestres du sud-marocain. Cahiers de l’AARS, Vol. 5; pp. 15 – 26. France.
Núñez Jiménez, A. 1986. Petroglifos del Perú. Panorama mundial del arte rupestre. 2da. Ed. PNUD-UNESCO – Proyecto Regional de Patrimonio Cultural y Desarrollo, La Habana.
Russell, W. G and A. M. Wright. 2008. How far is a pipette? Iconographic Evidence for Inter-Regional Connectivity in the Prehistoric Southwest and Beyond. The Artifact. Vol 46; pp. 19 – 45. El Paso, Texas.
Troncoso, A. and D. Jackson. 2010. Images that travel: ‘Aguada’ rock art in north-central Chile. Rock Art Research. Vol. 27-1; pp. 43 – 60. Melbourne, Australia. With RAR Comments by Natalia Carden, Dánae Fiore and Robert G. Bednarik; with RAR Reply by the authors. PDF – Read my comments on Bednarik’s Comments.
Van Hoek, M. 2001. The stepped diamond in Andean rock art. Rock Art Research. Vol. 18-2; pp. 96 – 100. Melbourne, Australia.
Van Hoek, M. 2004. The stepped-fret motif in American rock art: an attempt at tracing origin and meaning. The Artifact. Vol. 42; pp 75 – 91. El Paso Archaeological Society. El Paso, Texas.
Van Hoek, M. 2007. A re-evaluation of the ‘monkey’ petroglyph at the Quebrada de San Juan, Peru. Rock Art Research. Vol. 24-2; pp. 255 – 257. Melbourne, Australia.
Van Hoek, M. 2011. Petroglyphs of Peru – Following the Footsteps of Antonio Núñez Jiménez. Oisterwijk, The Netherlands.
Van Hoek, M. 2016. The Avian Staff Bearer. Upgrading a Controversial Icon in Atacama Rock Art. In: TRACCE – On-Line Rock Art Bulletin.
Van Hoek, M. 2018. The pipette-design in Desert Andes rock art. Adoranten. Vol. 2017; pp. 103 – 115. Underslös, Sweden.