Translate This Page

Welcome to my web site.

This web site, Andean Rock Art Papers, offers three pages with Rock Art Papers, all dealing with the Andes of South America. The first page has been labelled Part 1 in the menu-bar at the top op the web page, while the second and third pages are labelled Part 2 and Part 3.  Each page may contain several papers that can be found by scrolling down the page. In the future further papers will be published and they will then be accessible by first clicking on the relevant page (Part 1, Part 2 or Part 3 in the menu-bar at the top of this page) and then by scrolling down the page. A list of all published papers will appear on each page (also providing the format how to refer to each paper).

My major web site is Rock Art Investigations

Also visit Rock Art Papers

Header photograph: The author at Alto de Pitis, Valle de Majes, Peru.

Part 1 - Paper 1: Van Hoek, M. 2013. The Horseman of Alto de Pitis: A Post-Columbian Outsider in a Pre-Columbian Landscape. Andean Rock Art Papers; Part 1 - Paper 1.


Published: 31 - 10 - 2013

The Horseman of Alto de Pitis, Peru: A Post-Columbian Outsider in a Pre-Columbian Landscape

 Maarten van Hoek

Please notice that the text and all graphical material is my copyright !

A (not-public) YouTube video gives further relevant graphical information.
The word (video) in the text refers directly to the film.
This video can be viewed by using the following link:
The Horseman

I am most interested to receive reactions and constructive comments from any reader.

So far, several people suggested to include a photograph of the Horseman in the article itself.
This indeed proves to be an improvement - Thanks !

The Horseman Petroglyph at Alto de Pitis, Peru. Photograph
© by Maarten van Hoek.

This research report discusses the small image of a horseman on one of the petroglyph boulders at Alto de Pitis, southern Peru. In an area that is well known for its rich collection of Pre-Columbian rock art, the depiction of this unique Post-Columbian horseman is an unexpected exception.

El objetivo de este trabajo es explorar la pequeña imagen de un jinete en una de las rocas grabadas en Alto de Pitis, en el sur del Perú. En un área que es conocida para su colección rica de arte rupestre Pre-Colombiana, la representación de este jinete Post-Colombiana único es una excepción inesperada.

Keywords: Petroglyph, Horseman, Colonial Period, Alto de Pitis, Andes, South America.
Palabras clave: Petroglifo, Jinete, Período Colonial, Alto de Pitis, Andes, América del Sur.

Cast a cold eye
On life, on death.
Horseman, pass by!

From: Last Poems; 1939. W. B. Yeats (1865 - 1939).

Alto de Pitis is an important archaeological complex on the east bank of the Río Majes. In fact this river valley is the downstream continuation of the well known Colca Canyon in southern Peru. Most rock art scholars only know of the otherwise very impressive site of Toro Muerto, which is situated on the west bank of the River Majes (video), and undoubtedly many will have visited (only) Toro Muerto. What many researchers do not realise is that - when driving from the city of Arequipa to the Majes Valley - one crosses the even more amazing rock art site of Alto de Pitis, directly opposite Toro Muerto.

Although Toro Muerto (with allegedly 5000 petroglyph boulders) comprises the biggest rock art concentration of the whole Andes, Alto de Pitis, covering almost the same number of square kilometres as Toro Muerto, has relatively much more to offer, even though Alto de Pitis has significantly less (boulders bearing) petroglyphs than Toro Muerto. The rock art site of Alto de Pitis measures 630 m in maximum width (E-W) and 4580 m in length (N-S). Its boulders are widely distributed on an undulating terrace with an altitude of 440 m O.D. at the extreme southern tip and 535 m O.D. in the north. On this terrace there is no vegetation at all; not even in the dry river bed of the Quebrada de Pitis that cuts through the boulder field. Only a few birds, insects, reptiles and - very rarely - bats are sometimes spotted in this extremely dry desert.

The whole area is characterised by grey sands, strewn with small, often smooth pebbles and rough rock fragments, cobbles and scattered boulders of volcanic origin (video). The petroglyph boulders vary in size from around 50 cm in width by 30 cm in height to enormous boulders almost ten times as big. Boulder AP3-183, on which the Horseman has been executed, is one of those huge boulders.

Historical Background
According to Linares Málaga (1990: 158-159) the petroglyph site of (Alto de) Pitis was (first?) reported to him in 1951, by José Medina. This is strange because in a publication as early as 1949 Walter Krickeberg (1949) mentioned a number of petroglyphs sites in Peru and his list included as well a site called ‘Pitas (Pitis), Cerro Colorado in Majes, Vitór road, Aplao in the Department of Arequipa’ (Linares Málaga 1978: 379). This description unmistakably concerns Alto de Pitis. However, according to Linares Málaga (1978: 379) Krickeberg confused Pitas with Pitis, which is not situated on Cerro Colorado (according to Linares Málaga Cerro Colorado does not exist - however, see my remarks further down). Linares Málaga continues to explain that ‘near to Sarcas in the same valley of Majes, there is another place, called Punta Colorado’ (this should read Colorada), ‘where petroglyphs do exist’ (although, according to Núñez Jiménez [1986: 527 - 53], only three petroglyph boulders have been reported at Punta Colorada). Linares Málaga explains the Krickeberg ‘errors’ by assuming that Krickeberg was not personally acquainted with these areas. However, Linares Málaga makes so many (spelling- and location-) errors himself that I doubt his accuracy and his knowledge of all the areas he described. However, I am confident that Krickeberg did not refer to the (‘insignificant’) three petroglyph boulders at Punta Colorda, but indeed to Alto de Pitis. Therefore, I still wonder how Krickeberg had been informed about the petroglyphs of Pitis, or, as he called the site, Pitas.

This puzzle is unravelled by an old photograph in the possession of the National Geographic Society  (Picture Id: 602411 or 785801). This old photo notably proves that the well known explorer Hiram Bingham, the discoverer of Machu Picchu, recorded rock art at Alto de Pitis for the first time, as early as 1911 (Bingham 1912, 1922). Photo 602411 notably shows Hiram Bingham standing alongside a very large petroglyph boulder (labelled AP3-044 by me in 2009) that is nowadays easily visible with Google Earth. Although one web site claims that this photo shows Hiram Bingham at Toro Muerto and another web site that it is located ‘near the ruins’ (of Machu Picchu?), it is absolutely certain that the photo was taken at Alto de Pitis, on the 7th of October 1911. The caption to photo 602411 notably reads: Location: Pitas, Majes Valley, Peru. Moreover, I recorded the same rock at Alto de Pitis in 2009; almost one hundred years after Hiram Bingham.

The use of the name Pitas makes it very likely that, much later, Walter Krickeberg either visited the same site, or just heard of (or read about) Hiram Bingham’s accounts of the 1911 Coropuna Expedition (Bingham 1912, 1922). At the time of Hiram Bingham’s expedition the site was possibly known as Pitas or perhaps misunderstood as Pitas. Interestingly, the records by Hiram Bingham (1912, 1922) also related of Cerro Colorado as an alternative name for Pitas. The most relevant part of his account notably reads: At eight o'clock in the morning, as we were wondering how long it would be before we could get down to the bottom of the valley and have some breakfast, we discovered, at a place called Pitas (or Cerro Colorado), a huge volcanic boulder covered with rude pictographs. Further search in the vicinity revealed about one hundred of these boulders, each with its quota of crude drawings.

It is very understandable that Eloy Linares Málaga did not know of the specific part of the Hiram Bingham expeditions that led to the discovery of Alto de Pitis. Only a few words appear in the records by Hiram Bingham. In those days (1953) there was no internet and in general specific information would be very scarcely available and hard to get hold on. Moreover, after more than 40 years the publication of Bingham’s accounts at Alto de Pitis may have been forgotten, or at least overshadowed by his discovery of Machu Picchu. It is most likely that Eloy Linares Málaga heard of Pitas for the first time when reading Walter Krickeberg’s (1949) account, or when José Medina informed him in 1951.

Alto de Pitis
The oldest record of the name Alto de Pitis that I could find appears in a publication by Linares Málaga (1990: 165). In this publication, a photograph shows one of the most important panels of the area (Boulder AP3-114). The caption to the photo reads: Petroglifo de Alto de Pitis. Enero 1957. Foto; Eberhard Schön.

Alto de Pitis has been described as Pitas and Cerro Colorado (Bingham 1912, 1922); Sarcas (Núñez Jiménez 1986: 311; Linares Málaga 2004: 37), or Sarcas - Hacienda las Palmas (Linares Málaga 2004: 39; although a Hacienda Las Palmas is located on the west side of the valley) and Pitis (Núñez Jiménez 1986: 323; Linares Málaga 1990; Álvarez Zeballos 2009). But also the following names have been used: Alto de Pitis (Linares Málaga 1990: 165), Pitis-Mezana (Linares Málaga 2004: 39), La Mezana and Cantas (Álvarez Zeballos 2009) and Cantas or El Creston (Linares Málaga 2004: 30). To mix up things further, Guffroy (2009: 213) describes the ‘old’ sites of Sarcas and Pitis under one name, ‘Mesana’, while Hostnig (2003: 55) groups the sites of Pitis-Cantas-La Mezana under the name ‘Alto de Pitis’, but includes Sarcas in his inventory as a separate entity (while geologically and archaeologically Sarcas is clearly part of the ‘old’ Pitis group). Regarding the name Cerro Colorado, it is interesting to read that the inventory by Rainer Hostnig (2003: 44) includes an entry by that name, said to be located ‘Frente al sitio Pitis’, stating that he got that information from Álvarez Zeballos in 2003. I wonder where Álvarez Zeballos found the name of Cerro Colorado. I hope you are still with me.

Therefore, in order to avoid confusion, I prefer to assign one single name to all rock art sites in this area and because all boulders are found at a considerable height I have decided to use the name Alto de Pitis, the Heights of Pitis, also to honour the late Eloy Linares Málaga, who was probably the first scholar to use this name in a scientific publication.

Boulder AP3-183
One specific petroglyph at Alto de Pitis - on a panel labelled AP3-183A by me - is the focus of this research paper. As far as I know, a record of this petroglyph, the Horseman, was first published in 2012, when a short description and a photograph appeared in one of my earlier publications (Van Hoek 2012: 88; Fig. 198). Also in a subsequent publication the Horseman has briefly been described (Van Hoek 2013: 92). I could not find any earlier reference to the Horseman in published works that are available to me. Also the Inventory Map of the Department of Arequipa regarding Post-Columbian rock art manifestations by Linares Málaga (1992: 225) does not record the Horseman petroglyph.

In the years that followed after the recordings by Linares Málaga, several researchers visited, surveyed or described the site, for example Eberhard Schön in 1957, Hans Disselhoff (1968), Hans Horkheimer in 1968, Antonio Núñez Jiménez (1968), Jean Guffroy (2009) and Paúl Jofrey Álvarez Zeballos (2009). However, most published work only concerns minor investigation and unsystematic recording.

The most extensive publication of the rock art at Alto de Pitis appears is the ‘inventory’ by Antonio Núñez Jiménez (1986). However, in his well-known but - especially regarding the graphic part - most controversial work (Van Hoek 2011b), Núñez Jiménez incompletely describes (often with incorrect illustrations) only two very small parts of the much larger Alto de Pitis complex (Van Hoek 2013). The southernmost section was called Sarcas by him (1986: 311), while the adjacent, northerly section, only separated by the modern road, was called Pitis (1986: 323). The section where the Horseman is found, was never visited by Núñez Jiménez.

In 2009 rock art researcher Jean Guffroy published a colourful (photo)book about the petroglyphs of coastal Peru, which included a section of the ‘Mesana’ petroglyph site, which is actually covering only an extremely small part of the ‘old’ Sarcas and Pitis sites. Also Guffroy never encountered the Horseman.

Also in 2009 Paúl Jofrey Álvarez Zeballos published an interesting and more comprehensive survey about Alto de Pitis on the internet. His study area covered the much larger area north of the Sarcas section, but curiously he excluded from his survey the sections of Sarcas and Quebrada de Pitis (2009: Mapa 1). It is this survey by Álvarez Zeballos that to a certain extent revealed the great importance of the rock art on the east side of the Majes valley, also because he is the first to fully describe the superb petroglyph site of La Laja, located only 17 km north of Alto de Pitis. Álvarez Zeballos also was the first who claimed that the whole area, from Torán in the south to Cantas in the north, actually forms one large petroglyph field (2009: Introducción). Although he seems to have explored the section in which the Horseman is located, I doubt if he has seen this important stone; at least there is no mention in his survey of this specific boulder, or of the Horseman petroglyph.

Between 2002 and 2012 my wife Elles and I visited the Majes Valley many times and especially at Alto de Pitis we recorded numerous interesting and often alien-looking petroglyphs (Van Hoek 2013: 75 - 136). In total we surveyed 369 petroglyph boulders (584 panels) at Alto de Pitis and, together with the extra rocks/panels illustrated/described by Núñez Jiménez (1986) and Álvarez Zeballos (2009) but not seen by me (unfortunately several boulders/panels have been destroyed), the total of known petroglyph boulders at Alto de Pitis is at least 381 (596 panels). As many rocks will have escaped being noticed by me, and possibly also by Álvarez Zeballos, the overall number of petroglyph boulders at Alto de Pitis will definitely exceed 400 (resulting in more than 600 panels) and thus Alto de Pitis is now one of the very few rock art sites in Peru with more than 500 decorated panels.

Especially Alto de Pitis is most comprehensively described and discussed by me in another publication because of the many unique skeleton-anthropomorphs petroglyphs (‘Carcanchas’) at this site, labelled the ‘Petrified Cemetery of the Death Valley of the Andes’ by me. Those ‘Carcanchas’ - the Living Dead’ - are connected with a Sacred Mountain, the volcano Apu Coropuna, 85 km to the NNW, that is visible only from Alto de Pitis (Van Hoek 2013).

Location and Description of Boulder AP3-183
Boulder AP3-183 is located exactly 98 km west of the city centre of the city of Arequipa (bearing 279°), 4,5 km to the SE of the village of Corire (bearing 143°) and 2,1 km due north of the tarred main road from Arequipa (the video offers several location maps). The boulder is found on an undulating terrace at an altitude of 518 m O.D. and 470 metres east of the valley/escarpment. The large boulder is in a rather secluded location with views only to the north. From the boulder no vegetation is visible. Although Boulder AP3-183 has ‘only’ two decorated panels, especially Panel A is important for its depiction of a little Horseman.

The north facing Panel B (video) has an interesting collection of petroglyphs, all of the so called ‘Chuquibamba Rock Art Style’ that is so characteristic for this area (although ‘Majes Rock Art Style’ would have been more appropriate). The images include depictions of ‘felines’, birds’, a ‘snake’, abstract motifs and other figures, as well as two ‘dancer’ figures that are so distinguishing for the rock art of this part of the valley. Interestingly, Alto de Pitis is one of the only three rock art sites in the Andes where this type of ‘dancer’ figure has been reported (Van Hoek 2013: 58 - 61 and 103). A large part of Panel B has flaked off, leaving a distinct white area (located between Panels B and A) that clearly stands out against the brownish colour of the rest of the boulder. It is certain that as a result of this flaking (parts of) other petroglyphs have disappeared. The much flaked east and SE facing panels of the boulder have no (visible) petroglyphs.

The west facing, almost vertical Panel A (video) is roughly triangular in shape and measures roughly 260 cm from its apex to the ground floor. The maximum width of the boulder is about 165 cm, measured horizontally across the image of the Horseman. The Horseman itself is situated approximately 57 cm above ground level and about 55 cm from the left edge of the panel.

Panel A features many other petroglyphs of doubtless various epochs. Above the Post-Columbian Horseman is a group of outlined, dotted zoomorphs, so characteristic for the bulk of Pre-Columbian rock art imagery of the Majes valley, which is often referred to as the Chuquibamba Style (although I would have preferred the term ‘Majes Rock Art Style’). To the right is a large group of mainly match-stick zoomorphs (probably camelids of a later date), some of which seem to show ‘riders’ (see however an alternative explanation in Van Hoek 2012: 87 - 92). Just below the Horseman is a group of small zoomorphs, some of which might have been intended to depict horses. These images will not be discussed here, because they lack the graphical content of a rider.

Description of the Horseman (link to the photo - opens in new window)
The Horseman petroglyph is rather small. From the tail-end to the nose of the horse the image measures about 15 cm. The overall height is about 17 cm, measured between the lowermost end of the hind leg to the top of the second incised line above the spear. Despite its small size some fine details are most revealing.

The fully laterally depicted image depicts the probably (pecked and later?) abraded/recessed image of a horse and a rider. The animal does not look like a camelid at all and definitely is a horse showing two legs, two short triangular ears and a typical thick horse-tail. Only the upper body of the rider is shown which is sitting directly on the back of the horse; no human legs have been indicated below the horse’s belly. However, the otherwise rather rough and relatively deeply recessed interior of the body of the horse shows very faint features which possibly once may have depicted the human legs and other features. The front of the horse shows a small projection.

The Horseman as well is similarly (pecked and later?) abraded/recessed. The body and the circular head of the Horseman show no detail, but again very faint traces of once existing details are still visible. The (right?) hand which shows three, possibly four, splayed fingers that have been incised. The left (?) hand also may have had (incised) fingers, but these seem to have been weathered off.

Interestingly, the Horseman holds two things in his hands (I assume that the anthropomorphic figure is male because it holds a spear). Indeed, one hand holds an object what appears to be a spear (video). This object is about 11 cm in length and comprises a single straight line - pointing forwards - ending in a distinct, outlined spearhead; the whole object being rather superficially but clearly incised. Hovering above and parallel to the ‘spear’ is a second but fainter, incised line; an earlier attempt to sketch the spear? The other hand of the rider - with splayed fingers - unquestionably ‘holds’ a tack: two thin lines that run from the hand across the recessed interior of the horse (in itself a rare feature) and that finally are fixed to the mouth of the animal (video).

The Chronological Context of the Horseman
The evolution of the horse started about 52 million years ago with the appearance of Eohippus in the Eocene Epoch of the Tertiary Period. In most continents the animal gradually evolved into the modern horse and many rock art images in especially Europe and northern Africa, give evidence of the human interest in this animal, ranging from very old (Palaeolithic) to almost recent images. At various points in time and at differing places the horse was finally domesticated and used for many purposes. One of those purposes was warfare.

However, in the Americas indigenous horses became extinct at the end of the Pleistocene, about 12000 years ago. Thus horses were absent in North and South America until the Spanish Conquistadors introduced domesticated horses from Europe, from 1492 onwards. Also the ancient Andeans who had in South America domesticated the llama, the alpaca and the dog, knew nothing of the horse until around 1528 Francisco Pizarro and Diego de Almagro made first contact with Inca Empire near the modern town of Tumbes on the northern coast of what is now the modern republic of Peru. Interestingly, there are some interesting differences between North and South America regarding the occurrence of riders in rock art.

The Rockies
Particularly in the Southwestern part of North America this contact with strangers on horses resulted in various rock art expressions. Especially the recently published book by Dennis Slifer (2013) illustrates numerous examples of Post-Columbian rock art in the Southwest, including exceptional cases of horses and equestrians and related imagery. A fine example occurs for instance on the Newspaper rock in Utah, where among the many prehistoric Anasazi and historic Ute petroglyphs is the image of a horseman who is probably hunting a deer (Figure 1). All these Southwest horse-images date from A.D. 1540 (Malotki and Weaver 2002: 181).

Figure 1: The horseman petroglyph on the Newspaper rock, Utah, USA. Photograph © by Maarten van Hoek.

Especially the Navajo produced so many images of the horse (predominantly featuring a rider) that Malotki and Weaver (2002: 180) even have been able to propose a chronological sequence for the various Navajo manifestations of the horse and rider. They moreover argue that the horse motif has undergone those major changes in the course of the historic period that appear to coincide with traumatic events in Navajo history (2002: 187). For the Southwest one of those traumatic events will definitely have been the Spanish conquest.

The Andes
In the Andes of South America the situation is quite different. First of all, in the Andes the overall numbers of rock paintings and petroglyphs of horses are significantly less compared with the numbers of horse depictions in the Rockies of North America. And although the Spanish conquest of the Andes started in the northwest (coastal Peru), especially the southern (Chile) and eastern parts (highland Peru, Bolivia and Argentina) of this mountain range have their fair share of rock art depicting Post-Columbian contact.

However, in this respect is important to realise that several of the animals in Andean ‘horse-and-rider’ imagery not necessarily depict horses. Unlike the situation in North America, in several cases existing Pre-Columbian images of other species of quadrupeds (mainly ‘camelids’) will have been changed into ‘horses’ by adding a rider onto the back of the animal (Arenas and Martínez 2009), or by completely changing the original layout of the ‘camelid’ beyond recognition (although this is almost impossible to prove). Moreover, a few depictions of Pre-Columbian camelids occasionally feature a Pre-Columbian anthropomorphic figure that is either standing or sitting on the back of the animal, thus expressing the so-called ‘Lord of the Camelids’ theme (Berenguer 1999: 30).

The Eastern Andes
In the eastern part of the Andes, rock art depicting zoomorphs ridden by humans has been reported at relatively many places. Examples of riders in the rock art of the eastern Andes occur for instance at Capillayoq, Killarumiyoq and Jaquechia (Hostnig 2003: 100; 114 and 267), at Hutumayo and Apachaco (Arenas and Martínez 2009: Figs 8 and 14) and at Irocunca, Alta Huarca and Chearaje (Tarco Sanchez 2012: Figs 38, 46, 58 and 60) in Peru (see also Hostnig 2004: Fig. 7); at Hake Kayu and Chiripaca in Bolivia (Arenas and Martínez 2009: Figs 6a and 9); and at Sapagua (Arenas and Martínez 2009: Fig. 7), Cerro Colorado (Gentile Lafaille 2011), Banda Florida (Figure 2) (Van Hoek 2011a: Figs 12, 13 and 17) and on Panel TAL-D-001 at Talampaya in Argentina (Van Hoek 2011a).

Figure 2: A ‘rider’ petroglyph (probably on a llama) on Panel BAF-004B2 at Banda Florida, Argentina. Rupestreweb. Photograph © by Maarten van Hoek.

The Western Andes - Chile
However, most remarkable is the notable dearth of definite Post-Columbian horse images in the western parts of the Andes. It is surprising to see that in the deserts to the west of the Andes (the area from the Sechura Desert in the north of Peru to the Atacama in the north of Chile) only very few examples of riders have been reported, notwithstanding the fact that numerous rock art sites have been recorded in this area. Furthermore, most examples are found in the very southern (Chilean) part of this extremely dry desert.

Indeed, in the southern part of the western Andes, several examples (of ‘llamas’ and riders) have been reported from the semi-desert south of Copiapó in central Chile, for instance in the valley of the Río Hurtado, 55 km ENE of Ovalle (Ballereau and Niemeyer 1999: Figs 39D and E; 40B) and at the petroglyph site of El Coligüe, 80 km south of Ovalle (Guerra Terra 2012). Further north and located in the heart of the Atacama Desert, examples have been reported from the Ayquina-Caspana area in the drainage of the Río Loa (Gallardo et al. 1990; Arenas and Martínez 2009).

An interesting mix of ‘riders’ and ‘camelids’ occurs at the Quinchamale site near Santa Bárbara rock art complex, also on the Río Loa in northern Chile (see also Berenguer 1999: 47). This petroglyph concentration is located alongside a Pre-Columbian trail for pack animals, llamas, and later used for mules and horses. Two panels with riders are of interest. Panel BARn-005B has possibly three quadrupeds with a ‘rider’. The animals may concern Pre-Columbian images. Panel A, however, has a row with Roman letters and two outlined images of unequivocal Post-Columbian ‘horses’ that are ridden by a human; both completely laterally depicted.

More revealing is nearby Boulder BARn-007. It has some Pre-Columbian petroglyphs but also a clear example of a Post-Columbian, laterally depicted petroglyph of a horseman (Figure 3). The other end of the boulder features the numbers 1875/80 (Figure 4), most likely indicating the year(s?) in which this spot was visited. Boulder BARn-007 is one of the very few instances in the Andes where a (horseman) petroglyph is ‘accompanied’ by a date (it is uncertain however if the horseman petroglyph is of the same date as the date). The date of 1923 recorded by Linares Málaga (1992: 225) on a boulder at Toro Muerto, southern Peru, is irrelevant in this case.
Figure 3: The horseman petroglyph (and other, Pre-Columbian images) on Boulder BARn-007, Quinchamale Gorge, Santa Bárbara, Río Loa, northern Chile. Photograph © by Maarten van Hoek.

Figure 4: The date petroglyph on Boulder BARn-007, Quinchamale Gorge, Santa Bárbara, Río Loa, northern Chile. Photograph © by Maarten van Hoek.

Strangely, the abundant rock art sites north of the Río Loa have only one example or - in most cases - not a single instance of Post-Columbian equestrians. For instance Tarapacá-47, a site in northern Chile also located on a major caravan trail and featuring more than 400 decorated boulders, I recorded only one simple petroglyph of a horseman (Figure 5) (Núñez and Briones 1968: 52, Fig. XIII-b).

Figure 5: The horseman petroglyph on Boulder TAR-096, Tarapacá-47, northern Chile. Photograph © by Maarten van Hoek.

The Western Andes - Peru
Then follows a large area apparently void of Post-Columbian equestrians until, at Miculla (235 km NNW of Tarapacá-47), a site in the extreme south of Peru with also more than 400 petroglyph boulders, possibly four examples of Post-Columbian riders occur. But only the scene on one boulder (MIM-051) seems to depict a Post-Columbian equestrian on a Post-Columbian horse (Figure 6).

Figure 6: The horseman petroglyph on Boulder MIM-051, Miculla, southern Peru. Photograph © by Maarten van Hoek.

Only 75 km NW of Miculla is the rock art site of Turulaca, southern Peru. Boulder 7 at Turulaca shows a petroglyph of a ‘rider’ sitting on a zoomorph (an altered ‘camelid’?); the two legs of the ‘rider’ appear below the belly of the zoomorph (Gordillo Begazo 2010 ). Linares Málaga reported rock paintings of possible riders at the rock shelter of Q’aq’ahuma (located near the volcano of Pichu Pichu, east of Arequipa and 200 km NW of Miculla), but further north in Peru no riders seem to have been recorded by him (1992: 224). The ‘last’ petroglyph of a rider (cabalgadura) south of Alto de Pitis has been reported by Augusto Cardona (n.d.) at Huasamayo near Palca, just south of the Río Vitór (245 km NW of Miculla).

Then, in the coastal strip from Alto de Pitis (305 km NW of Miculla) to Tumbes in the north of Peru (a distance of 1653 km / an area teeming with rock art sites), no rock art depictions of the horseman have been recorded as far as I am aware. The most northerly example I know of in South America occurs in the centre of Colombia at the site of Sutatausa where two rock paintings possibly depict a horseman (Martínez Celis 2013: Fig. 54). These rock paintings are located some 2340 km north of Alto de Pitis and 1220 km NE of Tumbes.

Spear Carrying Riders
Most significant is the spear carried by the Horseman at Alto de Pitis. This weapon is only very rarely depicted in the Post-Columbian rock art of the Andes and again most examples occur at the eastern side. There is a scene of a ‘rider’ pointing a stick or a spear at a bowman in front of him at the petroglyph site of Sapagua in NW Argentina (Martínez 2009: Fig. 3), more than 1000 km SE of Alto de Pitis. Post-Columbian rock paintings of riders with hats and lances have also been reported at the Chirapaca site in the Department of La Paz, Bolivia, located 425 km east of Alto de Pitis and - like Sapagua - east of the continental watershed (Taboada Téllez 1992: Figs 11, 45 and 53). However, only two riders hold an object with a distinct spear head (1992: Figs 45 and 53).

Several rock art images of riders only show a simple straight line, often held completely vertical. These objects may be spears. Examples occur for instance at Huachichocana; Los Molinos; Cerro Pircado and Inca Cueva I (Fernández Distel 1992a: Figs 4 and 5) and at San Lucas (Fernández Distel 1992b: Fig. 8); all located in the NW of Argentina. However, one of the rock paintings at Inca Cueva I shows a rider with a vertical object that ends in a well-defined spear head. Other less clear and sometimes doubtful instances occur at Vila Caima and Cachi Cachi (Querejazu Lewis 1992a: Figs 1 and 5) and at Tunari Norte I (Querejazu Lewis 1992b: Figs 1 to 4, 6 and 7); all in Bolivia.

When focussing on the western part of the Central Andes it proves that only very few examples of spear-carrying horsemen occur. At Ayquina on the Río Loa in northern Chile (798 km SE of Alto de Pitis) petroglyphs have been recorded of anthropomorphs (most likely riding camelids) that are holding linear objects that might be spears or sticks or even whips (Arenas Campos 2011: Figs 2c and 47b). Furthermore, Arenas Campos (2011: Fig. 47a) suggests (2011: 155) that a rider petroglyph from Turulaca in southern Peru (a petroglyph site about 235 km SE of Alto de Pitis with a few examples of riders (see also: Picasaweb) is also carrying a spear. However, the object at Turulaca - on Boulder 13 - is only a simple line, held upright instead of pointing at the anthropomorph standing in front of the rider. It is therefore uncertain if the object is indeed a spear; it equally may be a whip or a stick.

The Chronology of the Horseman
As far as I know the Horseman of Alto de Pitis in the Majes Valley (video) is the northernmost example of an unequivocal depiction in western Peru of a Post-Columbian equestrian that is clearly armed with a spear. But which chronological context do we have for this Horseman? Unlike Boulder BARn-007 at Santa Bárbara in northern Chile, there is no date carved on Boulder AP3-183.

Unfortunately, only a few dates have been carved into anthropic structures and objects near or at rock art sites in the Department of Arequipa. In the Majes Valley is a small church, located directly below Toro Muerto, and opposite Alto de Pitis. The lintel above the front door of the church is dated A.D. 1722 (Linares Málaga 2011: 206). At the hamlet of Desamparados near the rock art site of Tacar in the Vitór Valley is a large collection of huge ceramic storage jars made by the conquistadores. Many of those jars have been decorated and also show inscriptions among which are dates. The earliest date that that my wife Elles and I could trace on one of the jars is A.D. 1674 (Figure 7).

Figure 7: Detail of a storage jar at Desamparados, Vitór, Peru. Photograph © by Maarten van Hoek.

However, it is certain that the Majes Valley had been invaded by the Spanish Conquistadores much earlier than 1722 or even 1674. Notably the following lines reveal that the conquistadores founded the coastal town of Camaná at the mouth of the Río Majes as early as A.D. 1539; one year before the foundation of the city of Arequipa: En 1558 aparece como Regidor de la villa de Camaná, una población en la desembocadura del río Majes, que había tenido su fundación española un año antes que Arequipa, teniendo como fundadores muchos de los que después participarían en la fundación arequipeña. (Unzueta Echevarría 2004: 119). It is therefore most likely that the Horseman of Alto de Pitis dates from after A.D. 1539, but this still leaves a rather large time span. However, there is another interesting geographical context.

The Horseman and the Old Vitór Road
Because the River Majes is not navigable, except for small rafts at certain points, it is almost certain that in early days the Majes Valley was travelled on foot with llamas and (later) also with horses and mules. Most of the valley notably comprises a flat, often inundated floodplain, which is rather easily accessible. Getting into the rather deep canyon from the adjacent Pampas (situated at 1300 m) was and still is another story. Then, high and steep cliffs of loose sands and boulders and outcrop cliffs had to be descended or ascended (video).

Boulder AP3-183 is located at 518 m O.D. on an undulating terrace, more than 100 meters above the floodplain of the Majes (at 395 m O.D. near Alto de Pitis) and from the valley below one has to climb steep slopes of treacherously loose sands to reach the terrace (video). It is therefore most likely that the Horseman was added onto Boulder AP3-183 by people who reached the site not from the town of Camaná, but via an old route across the adjacent Pampa de Majes.

Indeed, in early days people first travelled from the city of Arequipa to the village of Vitór and then crossed the flat and monotonous Pampa de Majes using the Old Vitór Road; a dirt road that is almost completely unused today (but still indicated as an existing road in Google Earth). Parts have of the Old Vitór Road in Majes are now covered by drifting sand and the section ‘gently’ descending into the valley is almost untraceable with the naked eye (but still visible with Google Earth).

In this respect the chronicles of Hiram Bingham are also revealing. Based on the information in his accounts, it is almost certain that also Hiram Bingham approached the petroglyph boulder field of Alto de Pitis on mule-back from the old road from Arequipa to the Majes Valley, the Old Vitór Road. Unknowingly he then passed the Horseman within a stone's throw. The Pan-Americana-Sur did not exist at that time, and certainly not the modern tarred diversion from the Pan-Americana to Corire in the Majes Valley. Hiram Bingham’s expedition started on the 2nd of October 1911 when they left Arequipa by train to Vitór. The next day, the 3rd of October, the expedition set off on mule-back from Vitór and on the 7th of October they arrived at Alto de Pitis; a journey of almost six days across the desert; single way! Today this trip only takes 2.5 hours by car. His account proves that the Old Vitór Road must have been used relatively often in the old days.

It may now be significant that Boulder AP3-183 is found in a loop of this Old Vitór Road, almost in the centre of the loop (video). Yet, although the boulder is only 68 m NW, 83 m NO and 64 m SSW of the track, it is invisible from most points of the road. Of course the naturally determined position of the boulder and the Pre-Columbian petroglyphs predate the manufacturing of the Horseman petroglyph and the construction of the Old Vitór Road by many years. But it is possible that the Horseman petroglyph is of later manufacture than the Old Vitór Road, which may have been (part of) the route of a Pre-Colombian road (for instance an old Inca road).

It is certain that the first sight of horses and their armed riders during the Spanish conquest was a frightening and overwhelming event for many ancient Andeans. In this respect it is surprising to see that depictions of horsemen are relatively rare in Andean rock art and, as I have demonstrated in this paper, and that most of them are found on the eastern side of the Central Andes (the high Andes and the eastern slopes). Moreover, only very few images in Andean rock art depict a horseman unambiguously carrying a spear.

On the other hand, the dearth of Post-Columbian images in the rock art of the western Central Andes is surprising. North of the site of Turulaca in southern Peru only one petroglyph of a rider, the Horseman of Alto de Pitis, seems to have been reported so far. Moreover, other common manifestations of Post-Columbian imagery are equally rare in the Majes area. For instance, the typical ‘cross on altar’ images (the so called Tampus) that are relatively ubiquitous in other areas of the Andes, only occur on four times; two examples at Toro Muerto and two at Alto de Pitis (all four almost unpatinated), while at Toro Muerto only one other boulder with a fresh-looking Post-Columbian shield-design is known to me (just above the shield are the letters V E P - Viva El Perú - ?). So, how do we explain the exclusive occurrence of the Horseman at Alto de Pitis?

It seems much likely that someone someday after A.D. 1540 passed Boulder AP3-183, possibly while shortcutting the loop of the Old Vitór Road, noticed the prehistoric petroglyphs and added the Horseman image. The person may have been a foreigner (a conquistador perhaps) who manufactured the Horseman in order to leave his mark, perhaps to commemorate a special event like the invasion of the Majes Valley by the Spanish conquistadores. The fact that the Horseman is a spear-carrier makes this military-related hypothesis acceptable. Also the fact that the image is rather deeply patinated (certainly when compared with the fresh-looking petroglyphs of the Tampus in the area) is an argument for an early date in the Post-Columbian era.

However, another explanation is that one of the local Majes people manufactured the Horseman to express his or her feelings about the Spanish conquest of his or her homeland. The Majes notably have a long history of their native soil being invaded. This record probably started long ago when the Nasca peoples infiltrated the valley around A.D. 100. And later the conquest by the military Wari certainly caused a lot of turmoil and stress (Tung 2007a; 2007b) when they invaded the Majes Valley around A.D. 560. As a result, the invasion of the elitist Wari influenced everyday life in the Majes Valley and, to a certain extent, the local ideology and cosmology. Some images of the Wari iconography may have inspired the local population, or perhaps the awesome appearance of the Wari warriors and officials ‘petrified’ the local population and as a reaction (a symbolic rebelliousness, perhaps?) they repeated (elements of) images derived from Wari iconography on the rocks at their most sacred spot; Alto de Pitis; the only site in the Majes Valley where convincing Wari Style rock art images have been reported; so far exclusively (Van Hoek 2013: Fig. 86). However, I argue (Van Hoek 2013: 87 - 91, 158 - 159) that their influence on rock art production in the Majes Valley has been only superficial, as only very few convincing Wari images have been recorded. Moreover, to me it is more likely that those Wari images may well have been manufactured by chief members of the local population; like Majes shamans.

Much later the Inca occupied the area, again with force. During Inca rule rock art production in the Majes area seems to have become almost passé. When, not much later, the Spanish warriors on their frightening horses arrived, the Majes people seem to have lost interest in rock art production completely. Therefore, the Horseman of Alto de Pitis may be one of the very few pieces of evidence of this historic event.

The ancient Majes, who did not know any kind of writing, could only graphically express their feelings and ritually left images on the rocks that were located at their sacred spots, separated from daily life in the valley. They probably produced rock art from the time that the Nasca made their first contact or even earlier. Since then huge numbers of petroglyphs were manufactured in Pre-Columbian times. The fact that so very few Post-Columbian rock art images exist in the Majes area may be explained by the possibility that the Majes had become apathetic to being invaded and conquered. After the Nasca, the Wari and the Inca, this final military invasion by the Spanish conquistadores was too much. The Majes felt powerless and did not care any longer to express their feelings graphically onto the rocks. The Horseman of Alto de Pitis may therefore prove to represent a unique exception. The fact that further north no depictions of horsemen seem to occur in rock art remains unexplained.


Álvarez Zeballos, Paúl Jofrey (2009). Petroglifos de Cantas, Pitis, La Mezana y La Laja; Valle de Majes. In: Arqueología de Perú.

Arenas Campos, Marco Antonio and Martínez, José Luis (2009). Constryendo nuevas imáginessobre los otros en el arte rupestre andino colonial. Revista Chilena de Antropología Visual. Vol. 13 pp 17 - 36. Santiago de Chile.

Arenas Campos, Marco Antonio (2011). Representatciones Rupestres en los Andes Coloniales. Un Mirada al Sitio Rupestre Toro Muerto (Comuna de la Higuera, IV Región de Coquimbo - Chile). Tesis para optar al título de antropológo. Universidad Academia. Santiago de Chile.

Ballereau, Dominique and Niemeyer Fernández, Hans (1999). Los sitios rupestres del valle del Río Hurtado superior (norte Chico, Chile). Chungara; Revista Antropología Chilena. Vol. 31 - 2, pp 229 - 292.

Berenguer, José (1999). The vanishing language of rock art in the Andes of Atacama. In: Rock art in the Andes of Capricorn; pp 9 - 56. Museo Chileno de Arte Precolombino. Santiago de Chile.

Bingham. Hiram 1912. The Ascent of Coropuna. Harper's New Monthly Magazine. Vol. 125; Number 741-45: pp 489–502. Henry Mills Alden, editor.

Bingham. Hiram 1922. Inca Land, Explorations in the Highlands of Peru. The Riverside Press Cambridge. Boston and New York.

Cardona Rosas, Augusto (n.d.). In: Panoramio.

Disselhoff, Hans Dieter (1968). Oasenstädte und Zaubersteine im Land der Inka. Archäologische Forschungsreisen in Peru. Safari Verlag. Berlin.

Fernández Distel, Alicia A. (1992a). Investigación sobre el Arte Rupestre Hispano –Indígena del N. O. de la República Argentina. In: Arte rupestre colonial y republicano de Boliovia y paises vecinos. Contribuciones al Estudio del Arte Rupestre Sudamericano. Editor: Roy Quejerazu Lewis. Vol. 3; pp 172-198. SIARB. La Paz, Bolivia.

Fernández Distel, Alicia A. (1992b). Pinturas Rupestres Posteriores a la Conquista Española en Jujuy, San Lucas, Dep. Valle Grande, Argentina. In: Arte rupestre colonial y republicano de Boliovia y paises vecinos. Contribuciones al Estudio del Arte Rupestre Sudamericano. Editor: Roy Quejerazu Lewis. Vol. 3; pp 199-209. SIARB, La Paz, Bolivia.

Gallardo, Francisco  et al. (1990). Jinetes sagrados en el desierto de Atacama: un estudio de arte rupestre Andino. Boletín del Museo Chileno de Arte Precolumbino. Vol. 4; pp 27 - 56. Santiago de Chile.

Gentile Lafaille, Margarita E. (2011). El Alero de los Jinetes: Iconografía e historia de sus representaciones rupestres (Cerro Colorado, Córdoba, República Argentina). In: Rupestreweb.

Guerra Terra, Alex (2012). Arte rupestre de El Coligüe (IV Región-Chile 2004). In: Aprender viajando.
Gordillo Begazo, Jesús (2010). Los petroglifos de Turulaca en el contexto del arte rupestre regional de Tacna. Boletín APAR. Vol. 1 - 3; pp 33 - 39. Lima. Perú.

Guffroy, Jean 2009. Imagénes y paisajes; rupestres del Perú. Editions IRD, Marseille, France; Universidad de San Martín de Porres, Lima, Perú.

Hostnig, Rainer 2003. Arte rupestre del Perú. Inventario Nacional. CONCYTEC, Lima, Perú.

Hostnig, Rainer (2004). Arte Rupestre Postcolombino de la Provincia Espinar. Cuso, Perú. Boletín de SIARB. Vol. 18; pp 40 - 64. La Paz, Bolivia.

Krickeberg, Walter (1949). Felsplastik und Felsbilder bei den Kulturvölkern Altamerikas mit besonderer Berücksichtigung Mexicos. Palmen Verlag, Berlin.
Linares Málaga, Eloy (1978). Prehistory and Petroglyphs in Southern Peru. In: Advances in Andean Archaeology; pp 371 - 392. Editor: David L. Browman. Cambridge University Press.

Linares Málaga, Eloy (1990). Prehistoria de Arequipa. Vol. 1. CONCYTEC UNSA, Arequipa, Perú.

Linares Málaga, Eloy (1992). Cuatro Modalidades de Arte Rupestre Post-Colombino en el Perú, en especial en la Región de Arequipa. In: Arte rupestre colonial y republicano de Boliovia y paises vecinos. Contribuciones al Estudio del Arte Rupestre Sudamericano. Editor: Roy Quejerazu Lewis. Vol. 3; pp 220 - 226. SIARB, La Paz - Bolivia.

Linares Málaga, Eloy (2004). Arte rupestre en Arequipa y el sur del Perú. Editorial Nuevo Milenio, Arequipa.

Linares Málaga, Eloy (2011). Memorias del Arqueólogo Eloy Linares Málaga. 80 Años de Edad y 60 de investigador. Castilla - Camaná. Universidad Alas Peruanas. Lima.

Malotki, Ekkehart and Weaver, Donald (2002). Stone chisel and yucca brush. Colorado Plateau rock art. Kiva Publishing. Walnut, CA, USA.

Martínez Celis, Diego (2013). Inventario participativo y caracterización de sitios con arte rupestre en Sutatausa (Cundinamarca). Una experiencia de apropiación social del patrimonio cultural. In: Rupestreweb.

Martínez, José Luis (2009). Registros Andinos al Margen de la Escritura: El Arte Rupestre Colonial. Boletín del Museo Chileno de Arte Precolumbino. Vol. 14 - 1. pp 9 - 35. Santiago de Chile.

Núñez Jiménez, Antonio (1986). Petroglifos del Perú. Panorama mundial del arte rupestre. 2da. Ed. PNUD-UNESCO - Proyecto Regional de Patrimonio Cultural y Desarrollo, La Habana, Cuba.

Núñez, Lautaro and Briones, Luis (1968). Petroglifos del sitio Tarapacá-47.  Estudios Arqueológicos. Vol. 3 - 4; pp 43 - 75. Universidad de Chile. Antofagasta, Chile.

Querejazu Lewis, Roy (1992a). Introducción. In: Arte rupestre colonial y republicano de Boliovia y paises vecinos. Contribuciones al Estudio del Arte Rupestre Sudamericano. Editor: Roy Quejerazu Lewis. Vol. 3; pp 6 - 12. SIARB, La Paz - Bolivia.

Querejazu Lewis, Roy (1992a). El Tunari: Montaña Sagrada. In: Arte rupestre colonial y republicano de Boliovia y paises vecinos. Contribuciones al Estudio del Arte Rupestre Sudamericano. Editor: Roy Quejerazu Lewis. Vol. 3; pp 52 - 66. SIARB, La Paz - Bolivia.

Slifer, Dennis (2013). Marking Change. Post-Contact Rock Art in the Southwest. Privately published using the BLURB Creative Publishing Service. Lexington, Virginia, USA.

Taboada Téllez, Fredy (1992). El Arte Rupestre Indígena de Chirapaca, Depto. de La Paz, Bolivia. In: Arte rupestre colonial y republicano de Boliovia y paises vecinos. Contribuciones al Estudio del Arte Rupestre Sudamericano. Editor: Roy Quejerazu Lewis. Vol. 3; pp 111-167. SIARB, La Paz, Bolivia.

Tarco Sánchez, Rául (2012). Prospección arqueológica y quilcas en la Provincia de Espinar, Cusco. Boletín de APAR. Vol. 13-14, pp 513 - 530. Lima, Perú.
Tung, Tiffiny (2007a). Trauma and violence in the Wari Empire of the Peruvian Andes: warfare, raids, and ritual fights. American Journal of Physical Anthropology. Vol. 133; pp. 941 - 956.

Tung, Tiffiny (2007b). The village of Beringa at the periphery of the Wari Empire: a site overview and new radiocarbon dates. Andean Past. Vol. 8; pp. 253 - 286.

Unzueta Echevarría, Antonio (2004). Los vascos de Arequipa y la Cofradía y Capilla de Nuestra Señora de Aránzazu. Donostia-San Sebastián : Eusko Ikaskuntza, Vol. 28. pp 115-137. Archivo Curia Provincial. Carmelitas. Manuel Iradier, 2. – B. 01005 Vitoria/Gasteiz.

Van Hoek, Maarten (2011a). Banda Florida. An overview of a rock art site in La Rioja, Argentina. In: Rupestreweb.

Van Hoek, Maarten (2011b). Petroglyphs of Peru - Following the Footsteps of Antonio Núñez Jiménez. Privately published using the BLURB Creative Publishing Service ( Oisterwijk, The Netherlands.

Van Hoek, Maarten (2012). Rumimantam Llaqllasaq Wirpuykita - The ‘Cycle of Life’ in the Rock Art of the Desert Andes. Privately Published using the BLURB Creative Publishing Service (  Oisterwijk, The Netherlands.

Van Hoek, Maarten (2013). The Carcancha and the Apu - Rock art of the ‘Death Valley of the Andes’. Privately Published using the BLURB Creative Publishing Service. Oisterwijk, The Netherlands.

Video: The Horseman of Alto de Pitis
Video fragments, photographs and music © by Maarten van Hoek - 2013.
All maps in this video are based on Google Earth (Maps) and Shaded (web page no longer available).