Indifferent Obliteration of Petroglyph Art

Print Friendly, PDF & Email

Petroglyphs are often found superimposed by other petroglyphs, but in some cases they have also been (partially) obliterated by hammering, rubbing or polishing of the rock’s surface. This short study investigates a number of cases in North and South America where petroglyphs may have (and in some cases definitely have) been obliterated by such grinding activities. This study therefore strongly recommend to accurately record instances where grinding activities took place, even when there are no petroglyphs visible.

By Maarten van Hoek


Indifferent Obliteration of Petroglyph Art




Maarten van Hoek – rockart@


A PDF of this paper can be downloaded at ResearchGate



Petroglyphs are a form of rock art made on natural stone surfaces (boulders and outcrops) by removing small parts of the rock surface. This action may produce markings ranging from rather deep markings to extremely superficial markings. As long as the result is a more or less recognisable image (figurative or abstract) it is no problem classifying the image as a petroglyph. However, rock surfaces at petroglyph sites (in this study only petroglyph sites will be discussed) often include several anthropic markings that cannot immediately be identified as rock art, while other clearly anthropic manifestations (like mortars, grinding holes and grinding slicks) definitely are no form of rock art.

The standard manifestations of petroglyph art are often well defined, figurative images (like an anthropomorph), abstract designs (like a spiral) and cupules. Yet, in some cases intentional obliteration by hammering or rubbing the rock’s surface by prehistoric peoples (not by ‘modern’ vandals) may – in my opinion – also be regarded as a form of ‘negative’ rock art tradition. In the Central Majes Valley of southern Peru there are many instances where especially the facial features of snakes, felines and anthropomorphs have carefully and – in my opinion – intentionally been obliterated (Figure 1), while in other instances premeditated obliteration was achieved by random hammering creating an amorphous area (Figure 2). In at least one case crude hacking obliterated a petroglyph at Toro Muerto, while its (zoomorphic?) shape is still somewhat discernable (yellow arrow in Figure 3). Especially its tail is still visible.

Figure 1: Carefully obliterated head of a large snake petroglyph at Toro Muerto, Majes, Peru. Notice that the obliteration only took place exactly within the outline of the snake’s head. Photograph © by Maarten van Hoek.


Figure 2: Crudely obliterated head of a zoomorphic petroglyph at Toro Muerto, Majes, Peru. Notice that the obliteration only destroyed part of the head. Photograph © by Maarten van Hoek.

Figure 3: Crudely obliterated petroglyphs at Toro Muerto, Majes, Peru. Photograph © by the Proyecto Arqueológico Toro Muerto.

In many areas of the world petroglyphs and grinding slicks co-occur in the same area, often very close together, like the two adjacent boulders at the rock art site of Nunyerry Creek, Pilbara, Western Australia. One boulder shows a deeply patinated grinding slick while the adjacent boulder has a petroglyph of an anthropomorph.

In more rare cases, however, grinding slicks and petroglyphs co-exist on the same panel.  Relatively frequently grinding depressions also are found being superimposed by later petroglyphs, as is the case with several examples on the Djado Plateau, Chad, and at Arkana, Ténéré, Niger, both in the very heart of the Sahara Desert. In one case a large outcrop in western Chad features more than six shallow grinding depressions three of which have been superimposed by a geometric design (possibly a snake?), while at least in four cases smaller petroglyphs of quadrupeds seem to have been superimposed upon several grinding slicks, but it seems that those small zoomorphic images were later partially obliterated by further grinding.

In many cases superficial hammering, bruising, rubbing or polishing of the rock’s surface produced amorphous worked areas that seem to have no direct relation to the petroglyphs on the rock’s surface. In my opinion such practices even resulted in obliteration of petroglyphs. But – contrary to the intentional obliteration at for instance Toro Muerto – I prefer to call those instances where images have (often partially) been erased indifferent obliteration. In other words, it was not the specific intention to erase previously manufactured petroglyphs or to eliminate their power and/or symbolism. In this paper I will discuss a selection of petroglyph sites possibly featuring instances of (possible) indifferent obliteration in the rock art of the Pacific Seaboard of the Americas.


Before starting the discussions it is good to define some terms here. In this study a grinding hole or a mortar is a deep, well defined, oblong or circular cavity rubbed out of a rock surface. A grinding slick on the other hand is only (very) superficially executed and often covers a much larger area that is usually oval in shape, often (but not necessarily) utilising natural, shallow depressions in the rock’s surface. The grinding slick can also be rather irregular (without sharply defined edges). Grinding slicks may show signs of rubbing or hammering, but often it seems that both activities have taken place at the same time. Of course it is also possible that petroglyphs on for instance vertical rock art panels have been hammered or rubbed without the purpose of grinding something. In rare cases stone surfaces even seem to have been rubbed smooth before manufacturing petroglyphs.

Obliteration in this study refers to the action that (often only partially) rubs out or hammers away an existing petroglyph. There can be distinguished two general forms of obliteration. Intentional obliteration in rock art can be compared with the wilful destruction of the names, faces and often parts of the body or the complete body of the sculptures of the Egyptian pharaohs on ancient Egyptian temple walls. Then the name (the very essence of the pharaoh) and (specific parts of) the image are the target. It thus proves that obliteration does not always involves the complete destruction of an image.

Unintentional or indifferent obliteration refers to the action where the older image is not at all the target, but still gets (partially) destroyed. A fine example is the ubiquitous ‘polissoir’ that is found on many Egyptian temple walls, as well as on many petroglyph panels across the globe. Polissoirs often indifferently destroy parts of ancient sculptures as well as petroglyphs. Although these long, often deeply abraded grooves may have been produced as a sort of ‘gestural art’ to record a visit to a site, they most probably have also been made during certain (fertility) rituals, for instance to extract the power from these ancient temples and petroglyph panels. But possibly the fine stone powder – thought to be charged with specific power – was also taken home as a relic or even swallowed on the spot, a practice that is generally called geophagy.

In this study indifferent obliteration is the process where older petroglyphs become (partially) erased because of later anthropic activities like grinding and rubbing. Indifferent obliteration is often hard to prove because erosion, weathering and patination (especially in desert environments) blur the successive layers of anthropic activities. Yet there are some instances where indifferent obliteration is evident.



I am convinced that all instances of amorphous hammering and indifferent obliteration (and indeed several other forms of anthropic makings, some of which are hard to classify) are relevant in the study of rock art as well. Fortunately there are high quality inventories that include such anthropic markings other than petroglyphs as well as natural features into their illustrations. A fine example is the extensive dissertation describing much of the rock art in the South Mountains, Arizona, USA, by Aaron Michael Wright (2011). Not only natural features are included in his numerous and excellent drawings, but in several cases there is also mention of subsequent obliteration of petroglyph panels and also of vandalism. I will return to his observations later in this study.

Therefore, when surveying a rock art site, the investigation must not only include the true manifestations of rock art (in this paper only involving petroglyphs), but also all other anthropic marks (even instances of vandalism, like the bullet scars on Panels AZ T:12:307.1.A-B at South Mountains, Arizona; Wright [2011]), as well as all natural rock features that may be associated with or may have triggered the rock art production (also evident in the South Mountains of Arizona; Wright [2011]). However, it proves that anthropic manifestations other than petroglyphs are often ignored in rock art studies. Unfortunately this is also often the case with inventories or studies by professional, academic archaeologists.

Quebrada de Algarrobos

An example is the unique and enigmatic occurrence of serrated edges at several petroglyph panels at the important rock art site of Quebrada de Algarrobos – Chuquillanqui in the Chicama drainage of northern Peru. Those serrated edges (see the arrows in Figure 4) are clearly anthropic markings and definitely were of special significance to the manufactures, and although it is completely obscure why they were created and if indeed they can be classified as petroglyphs, they seem to be more the result of some special and important ritual than the desire to simply decorating the rock’s edges. Yet, all the serrated edges that I recorded at Chuquillanqui (Van Hoek 2016) were completely neglected by academic archaeologist Daniel Castillo Benítez from Trujillo, Peru, in his book published in 2006, which – up to now – offers the only (incomplete and sometimes incorrect) mainly graphical inventory (only involving black-and-white illustrations) of the rock art of the Chuquillanqui Complex (involving at least two major sites: Quebrada de Algarrobos and Cerro el Diablo).

Figure 4: Several serrated edges at a very large boulder or outcrop at Chuquillanqui, Chicama, Peru. Photograph © by Maarten van Hoek.

I know of only two other (minor) instances in South America where rock edges of petroglyph boulders have been serrated. One is found at Toro Muerto in the Majes Valley of southern Peru, while the other example is found at Calaunsa in the Codpa Valley of northern Chile. Also in North America several cases have been recorded, for instance at Upper Arrow Canyon and Grimes Point, Nevada, where boulders with deeply serrated edges occur, as well as at Long Lake, Oregon, and Chalfant, California;. However, all those distant examples may express a completely different rationale for their manufacture.


Another example is the otherwise solid survey of the rock art site of Tolón in the Jequetepeque Valley of northern Peru by archaeologists Edgar Bracamonte Lévano and Ceyra Pasapera Rojas (2008). They registered 77 images and five cupules (pocitos) on 13 boulders, whereas in 2011 I recorded a 14th boulder with petroglyphs that I reported to Bracamonte.

Yet, three rock features at Tolón – all completely differing from petroglyphs – have not been included into their survey. The first anthropic feature concerns two possible grinding holes (mortars). A short distance to the SW of the main group of petroglyph boulders is an enormous flat area of outcrop rock that has two anthropic depressions. One small (and more doubtful) depression (A in the inset of Figure 5) seems to have a short channel (C in the inset of Figure 5) to a much larger and deeper depression (B in the inset of Figure 5) that is clearly polished through usage. Those two depressions may have been created for domestic purposes (for instance grinding food), but the very short distance (53 m NNW) to an important group of very large boulders (marked 2 to 5 in Figure 5) with several ancient Formative Period (Cupisnique) petroglyphs and the existence of several boulder-lined prehistoric tombs further east in the desert may point to ritual use.

Figure 5: Anthropic depressions at the edge of a very large outcrop at Tolón, Jequetepeque, Peru. Photographs © by Maarten van Hoek.

Secondly, although Bracamonte Lévano and  Pasapera Rojas describe and illustrate the five ‘pocitos’ on their Boulder 13 (Bracamonte Lévano and  Pasapera Rojas 2008: 16) they did not notice the large and unfortunately (recently?) fragmented boulder (TOL-016) with a large number of rather small, circular markings. Those fully pecked or rather, gently hammered markings are far too superficial to be classified as cupules (pocitos), but also the fact that a number of those markings have been arranged in a straight line seems to prove that they are anthropic (Figure 6). This linear arrangement may be compared with the thirteen dots in a row on Panel AZ T:12:102.44.A in the South Mountains south of Phoenix, Arizona, USA (Wright 2011: 624), said to possibly represent thirteen moons.

Also, the group of markings on Boulder TOL-016 at Tolón may have been created during certain rituals. In fact, the markings are also similar to markings occasionally found on ringing stones; stones that produce a clear sound when struck with a small stone. However, because of the fragmented condition of the boulder the possible acoustic property cannot be tested anymore.

Figure 6: Anthropic markings on a boulder at Tolón, Jequetepeque, Peru. Photograph © by Maarten van Hoek.

Finally, NW of the main group at Tolón is a boulder with three depressions that most likely are of natural origin (Van Hoek 2012: Fig. 12). However, the three holes in the vertical surface create the impression of a face or a skull. For that reason, the three holes may have been worked on (by a manufacturer of rock art?). Although this ‘skull’ rock has no petroglyphs as far as I could see, those (partial?) natural features may still have triggered rock art production in this area.  The same may be true for the ‘calavera’ (skull) rock formation at Yonán (with petroglyphs), also in Jequetepeque (Campana and Deza 2006: 58; Fig. 3). Also a large natural skull-like rock formation (without petroglyphs, though) high up Cerro Blanco in the Virú Valley may have triggered the production of petroglyphs on several boulders at the foot of Cerro Blanco. Such natural features are therefore also relevant to rock art studies and must at least be mentioned in scientific inventories.



Before dealing with some more unambiguous (?) instances of indifferent obliteration, it will be useful to briefly discuss some dubious examples. To start with, in some of the cases it is not clear if the obliteration has a natural cause or is indeed anthropic. Non-anthropic obliteration of rock surfaces (not necessarily with rock art) also occurs. In southern Africa there even are cases in which cattle rub against rock art panels resulting in (partial) obliteration. Also rhinoceros and hippopotamus are known to have rubbed numerous rocks in South Africa; often at rock art sites. And blowing wind could have moved branches across stones, scraping off parts of the natural surface over the years, sometimes damaging rock art.

Often it is not clear whether the ‘obliterating’ marks on the rock’s surface are anthropic or the result of erosion and/or weathering. For instance, I recorded a fine example on Boulder MP1-072 at Motocachy Pampa 1 in the Nepeña Valley of northern Peru. Unfortunately the boulder appears to have been displaced (possibly by the most unwanted anthropic disturbance in this area; see Van Hoek 2014) and moreover was in the shade at the time of photography. Therefore little was visible of the petroglyphs. Highlighting the panel in the photo revealed a partially exposed image of apparently an (inverted) bird (Figure 7). Most of the rest of the surface and the petroglyph was obliterated. However, it is uncertain whether the obliteration was anthropic or natural, as both types of obliteration occur on the wind swept pampa. A similarly doubtful example has been reported by me on Boulder MP1-016, also at Motocachy Pampa 1, which was – shamefully – also hammered and hacked by modern vandals who partially destroyed the boulder (Van Hoek 2014: Figs 7 to 11).

Figure 7: Inverted petroglyph of a bird on a (displaced?) boulder at Motocachy-1, Nepeña, Peru. Photograph © by Maarten van Hoek.

On Boulder MP1-023 – also at Motocachy Pampa 1 – is an area that has clearly been hammered (Figure 8). This has produced an amorphous group of rather clear bruising marks. This could be an instance where a petroglyph had been obliterated completely, but as there is no sign of a recognisable (partial) figurative or abstract image, I regard this as an instance of (possibly ritual?) hammering of the stone. The panel is almost vertical and this position seems to exclude domestic use like hammering food or pigments. However, also this boulder may have been disturbed by the shameful vandalism in this area and the surface may originally have been (near) horizontal.

Figure 8: Hammered boulder at Motocachy-1, Nepeña, Peru. Photograph © by Maarten van Hoek.

Further north in Peru, at Susanga in the Virú drainage, is fragmented Boulder PLN-015 (Figure 9). On its upper, slightly sloping surface are some traces of petroglyphs and a large area of hammering. I could not trace any petroglyphs that were partially (and understandably also not completely) obliterated by this hammering. The hammering on Boulder PLN-016 – also at Susanga – may represent a petroglyph or a petroglyph that was hammered afterwards. The almost vertical panel of Boulder PLN-031 has a number of very faint zoomorphic petroglyphs, while a large part of the surface (especially the left part) was covered with small and shallow depressions that apparently all are anthropic. Although some petroglyphs seem to have been hit, there was no sign of intentional obliteration.

Figure 9: Hammered boulder at Susanga, Virú, Peru. Photograph © by Maarten van Hoek.

At Mayasgo-1, also in the Virú drainage, is Boulder MYG-1-B-009 that also shows an area of hammering (Figure 10 and inset). The hammered surface is almost horizontal, but again there is no sign of a petroglyph. However, a petroglyph may have existed and may have (intentionally?) been obliterated completely. At the same site – very near a large boulder (MYG-1-C-001) that is covered with (initially natural?) cupules and some petroglyphs (Figure 11) – is a smaller boulder (MYG-1-C-002) with a large bowl-shaped and clearly anthropic depression (Figure 11; inset). I could not see any petroglyphs on the much weathered and dusty upper surface of Boulder MYG-1-C-002, but the depression most likely was used in some ritual, as was the adjacent cup marked boulder and a short distance from a major area with petroglyph boulders (MYG-1-B).

Figure 10: Hammered boulder at Mayasgo-1, Virú, Peru. Photographs © by Maarten van Hoek.

Figure 11: Large boulder with cupules at Mayasgo-1, Virú, Peru. Inset: detail of the anthropic basin near the cupule boulder. Photographs © by Maarten van Hoek.

Further north still, at the remarkable site of Quebrada del Felino in the Jequetepeque Valley of northern Peru, is Boulder FEL-006. On its SE sloping upper surface are the very faint traces of an anthropomorphic petroglyph. Just above the head is an area of possible hammering (Figure 12). It is uncertain if part of its head or a possible headgear has been obliterated. Also at Quebrada del Felino is Boulder FEL-047 that has a very large bowl-shaped central area – large and deep enough to collect a rather large quantity of water – that seems to have been rubbed smooth, possibly during domestic or ritual activities. I could not detect any petroglyphs on this boulder. A short distance NW of this boulder is another, smaller boulder with a small depression on its rather steeply sloping surface (arrow in Figure 13). It seems as if the whole surface had been hammered. No petroglyphs could be detected on its surface.

Figure 12: Anthropic markings on a boulder at Quebrada del Felino, Jequetepeque, Peru. Photograph © by Maarten van Hoek.

Figure 13: Anthropic markings on a boulder at Quebrada del Felino (arrow), Jequetepeque, Peru. Photograph © by Maarten van Hoek.



Having seen numerous petroglyph sites in the Desert Andes of South America I noticed that obliteration (whether intentional or unintentional) is very rare in this arid region west of the Continental Divide. Also, boulders with petroglyphs in the Desert Andes of South America very rarely have grinding slicks or areas that have been hammered. Yet there are some examples of indifferent obliteration.

Indifferent obliteration in Peru

Motocachy Pampa

A fine example occurs at Motocachy Pampa 1 in Nepeña, northern Peru. At this ancient site another boulder – Boulder MP1-022 – has been similarly hammered as Boulder MP1-023. This time however the boulder clearly shows the petroglyph of a bird of which the tail and the wings survived (Figure 14). The rest of the image is obliterated by hammering (and/or rubbing). The fact that the petroglyph was only partially obliterated may be a solid indication of indifferent obliteration. Moreover, there is no sign of a specific part of the bird’s image having been selected, as a large area outside the petroglyph was also hammered.

Figure 14: Hammered boulder at Motocachy-1, Nepeña, Peru. Photograph © by Maarten van Hoek.


Further north, at Palamenco in the Santa drainage, are a few petroglyph boulders that have an (almost) horizontal surface that seem to have been hammered, bruised or rubbed by prehistoric visitors. For instance, Boulder PAL-053 has an upper surface on which several petroglyphs are clearly recognisable (Figure 15), but the exact pattern of the whole set and several details are hard to discern because of the random hammering and/or rubbing. The Formative Period rock art site of Palamenco is also known for its many boulders the edges of which have been rubbed; some even to form petroglyphs (Van Hoek 2012: Figs 23 to 25).

Figure 15: Hammered boulder at Palamenco, Santa, Peru. Photograph © by Maarten van Hoek.


Further north in Peru, in the Virú Valley, is the ancient, Formative Period petroglyph site of Tomabal-A. On the almost vertical Panel TOMa-009B is a large MSC-Style (Cupisnique) design of which a small part in the lower right-hand corner has been obliterated by hammering (Van Hoek 2007: Figs 9 and 10). Lower down the slope of this site is Boulder TOMa-022 that has a slightly sloping upper surface that has almost completely been hammered. One petroglyph on an adjacent panel survived, but an apparently similar petroglyph was partially erased by indifferent obliteration (Figure 16). There may, of course, have been more images that are now completely invisible.

Figure 16: Hammered boulder at Tomabal-A, Virú, Peru. Photograph © by Maarten van Hoek.


At the impressive petroglyph site of Yonán, northern Peru, (with images ranging from the Formative Period to the Moche era) is Boulder YON-004. Apart from three anthropic depressions (blue arrows in Figure 17) there is also an area with light, random hammering (yellow arrow in Figure 17) that very irregularly erased some older images. As those markings are less patinated, the obliteration may be much younger than the petroglyphs.

Figure 17: Hammered boulder at Yonán, Jequetepeque, Peru. Photograph © by Maarten van Hoek.

Cerro Mulato

Much further north, at Cerro Mulato in the Chancay Valley, is Boulder CMq-225. It has a distinct petroglyph of an outlined triangular that is filled with small dots. The triangle seems to have two ‘legs’ and a few more appendages. The apex of the triangle has clearly been hammered and again, this partial obliteration may be an indication that there was no specific intention to obliterate the petroglyph (Figure 18).

Figure 18: Hammered boulder at Cerro Mulato, Chancay, Peru. Photograph © by Maarten van Hoek.

Toro Muerto

So far mainly examples from northern Peru have been discussed. There is however one very important site in the south of Peru where several forms of obliteration occur. It is Toro Muerto in the Majes Valley, 150 km NW of the fair city of Arequipa. At Toro Muerto more than 2000 petroglyph boulders have definitely been recorded. This number is a total of boulders registered during my own surveys at the site combined with the still ongoing, most extensive and professional survey by the Proyecto Arqueológico Toro Muerto. However, it is certain that there are many more decorated boulders. Some even mention a total of about 5000.

Toro Muerto has a (relatively) exceptionally large number of petroglyphs that have intentionally been obliterated. I demonstrated earlier (Van Hoek 2005 – 2006) that at Toro Muerto especially the facial features of anthropomorphs (in fact mainly ‘Dancers’), felines, snakes and some quadrupeds were obliterated in prehistoric times. Since then, several more examples have been recorded by me. Although we can only guess at the real reason behind this practice, I strongly advocate that this mainly facial-obliteration was executed to annul the power and/or the symbolism of the earlier icons of the site.

At the NE side of the boulder field – about 3.3 km from the edge of the Majes Valley and 300 m above the floodplain – are two boulders that have smooth depressions that – because of their location – most likely have been ritual grinding holes (or shallow mortars). Boulder TM-Bf-001 has at least three smooth but shallow depressions on its almost horizontal upper surface (1 to 3 in Figure 19) plus an area that has been superficially rubbed (4 in Figure 19). It seems that also the areas immediately surrounding the four depressions have also been hammered. There are several petroglyphs on this horizontal panel, but it is uncertain if images have unintentionally been obliterated.

Figure 19: Anthropic depressions on a boulder at Toro Muerto, Majes, Peru. Photograph © by Maarten van Hoek. Inset: Detail of depression 2. Photograph © by the Proyecto Arqueológico Toro Muerto.

Less than 75 m to the SSE is Boulder TM-Bf-007 that has several decorated panels. The stripes of Panel A continue onto the more or less horizontal Panel C where two shallow depressions clearly are grinding holes. Very often one or two small round boulders are found resting on those mortars as if these represent the original grinding stones (however, such round boulders are very common in the deserts of the Majes Valley). Especially the manufacturing of the western grinding hole (the left one in Figure 20) may have indifferently obliterated some petroglyph parts.

Figure 20: Anthropic depressions with (possibly prehistoric) grinding stones on a boulder at Toro Muerto, Majes, Peru. Notice the ‘hidden’ ‘dancer’ separating Panels A and B. Photograph © by Maarten van Hoek.

I already mentioned the fact that at Toro Muerto two types of intentional obliteration occur (see Figures 1 to 3). However, also two types of indifferent obliteration have been recorded, though more rarely than the intentional obliteration. The first type concerns the apparently indifferent obliteration of two (?) possible bird petroglyphs on the vertical panel of Boulder TM-Db-005B (Figure 21). It seems that rather randomly an area – including the petroglyphs – was pecked so that the exact pattern of the images became indistinguishable. It is possible that the surface was hammered without the intention to erase the images (in some kind of ritual), but equally it may have been an instance of premeditated obliteration.

Figure 21: Obliterated petroglyphs on a boulder at Toro Muerto, Majes, Peru. Notice the two possible bird petroglyphs (1: large; 2: very small). Photograph © by Maarten van Hoek.

Much further to the SE are two instances of type two. Boulder TM-Ex-008 (TMP-0252) is a large boulder (at 568 m O.D.) with no less than nine decorated panels. Panel A (I) is one of the larger panels. Within a very large ring mark is an area that seems to have been obliterated by rubbing, while in the centre is a smaller, slightly deeper area – with a small circular petroglyph in its centre – that might have been used for grinding (Figure 22). In the large rubbed area many petroglyphs are still recognisable as petroglyphs, but the fine details are often lost. It indeed seems that the petroglyphs within the ring have intentionally been erased.

Figure 22: Obliterated petroglyphs on a boulder at Toro Muerto, Majes, Peru. Photograph © by the Proyecto Arqueológico Toro Muerto.

About 188 m to the WSW of Boulder TM-Ex-008 and at 577 m O.D. is Boulder TM-Ex-015 (TMP-0275) with only one panel with petroglyphs. About one quarter of the slightly sloping upper surface has clearly been hammered and/or rubbed, creating a smooth, slightly lower area, simultaneously obliterating several images, some of which are still recognisable near the edges of the obliterated area. There seem to remain three types of surface: A) the original, natural surface; naturally patinated and un-worked; B) a slightly polished surface surrounding the more deeply and more intensely abraded central part of the obliteration (C).

On this panel there are several quadrupeds and possibly three bird petroglyphs (numbered 1 to 3 in Figure 23). The images mainly belong to the Majes Style Rock Art repertoire, while some images – especially the match-stick quadrupeds – clearly are later additions. One of the petroglyphs is Majes Style Bird 1 of which the tail – originally in area C – has been completely obliterated (yellow frame in Figures 23 and 24). The beak (B in Figure 24) and the leg (L in Figure 24) of Bird 1 have survived. Majes Style Birds 2 and 3 have just escaped being obliterated and only their heads have slightly been affected by rubbing, so it seems. Also at least two outlined, Majes Style quadruped-petroglyphs (2 and 3 in Figure 24) escaped obliteration, but also their legs have slightly been rubbed. Importantly, Quadruped 2 was consequently superimposed by a (much) later match-stick quadruped (in Zone A), while Quadruped 3 is touched by an even more crudely executed match-stick quadruped that was manufactured in Zone B. Finally, another match-stick quadruped was manufactured on the edge of Zone C. All those later petroglyphs (their relative age is also confirmed by their style) confirm the prehistoric nature of the obliteration, which I consider to be unintentional, especially because intentional obliteration at Toro Muerto mainly focussed on facial features and/or whole figures and not on tails of birds.

Figure 23: Obliterated petroglyphs on a boulder at Toro Muerto, Majes, Peru. Photograph © by the Proyecto Arqueológico Toro Muerto, courtesy of Karolina Juszczyk.

Figure 24: Obliterated petroglyphs on a boulder at Toro Muerto, Majes, Peru. Photograph © by the Proyecto Arqueológico Toro Muerto, courtesy of Karolina Juszczyk.



It is understandable that both intentional obliteration and of course indifferent obliteration of rock art images occurred in other parts of the Americas. Boulders with traces of obliteration occur in many parts of the Pacific Seaboard of North America and are mainly explained by grinding and rubbing activities. Often such traces – mainly including grinding slicks – occur on boulders within or (very) near a boulder field with petroglyphs. And yet, even when many undecorated boulders – yet suitable for grinding material – are available, in several cases it was a petroglyph boulder that was selected.

Below follows a selection of instances of possible indifferent obliteration at sites in the Pacific Seaboard of North America. Except for South Mountains, Arizona, I have visited none of those sites and therefore the information about those sites will no doubt be incomplete (but it is not the intention of this study to be complete). I am grateful to several people who helped me with contextual information and photographic material on which my observations are based. However, every bit of information in this paper and every opinion expressed by me is only my responsibility.

Indifferent Obliteration in California

Rainbow Canyon

In North America some fine examples of grinding slicks combined with petroglyphs are found at rock art sites in California, USA (about 7500 km NW of Toro Muerto in Peru). For instance, a rock art site in Rainbow Canyon, located in the wastelands about 10 km west of Death Valley, has at least four boulders that clearly show signs of grinding. One boulder just possibly once has had petroglyphs, but any possibly once existing images will have been obliterated (Figure 25). Interestingly, the deeper (brownish) grooves (possibly anthropic?) on the nearly horizontal surface have a different patination than the clearly rubbed (bluish) surface. Those grooves may represent areas with older patination. Yet, also the obliterated (bluish) surface seems to have patinated considerably, possibly pointing to great antiquity.

Figure 25: Boulder with grinding slicks at Rainbow Canyon, California, USA. Photograph © by Ron Wolf.

Another boulder at the same site clearly has petroglyphs, as at least three patterns (all geometric) are still visible (Figure 26). At the right side of the nearly horizontal surface is a large area that apparently has been used to grind some material. This area has been more intensively used, has patinated rather strongly and now has a blue colour. To the left is an area that has been less intensively used. It seems to have patinated more deeply and has a grey colour. This may again point to great antiquity of the primary grinding practices. It is not certain if any petroglyphs have been erased by the grinding activities.

Figure 26: Boulder with grinding slicks at Rainbow Canyon, California, USA. Photograph © by Ron Wolf.

Two other adjacent boulders (possibly once one solid boulder that cracked into several pieces) at the same site clearly have petroglyphs. Again, it seems that any possible obliteration was unintentional, but I emphasise again that the grinding slicks not necessarily must have destroyed any petroglyph. But it is a possibility.

Yellow Jacket

Only about 150 km further NNW of Rainbow Canyon is the Yellow Jacket Site (official name: CA-MNO-2189), Volcanic Tablelands, north of Bishop, California. The site includes over 200 petroglyphs believed to have been created between 2950 B.C. and 1850 A.D. by the Paiute-Shoshone who inhabited this rugged and harsh area. The site also includes remnants of human habitation, including at least one boulder with two distinct grinding slicks (but no visible petroglyphs).

However, another boulder at this site clearly shows two different signs of anthropic activity: it has petroglyphs and two grinding (?) slicks (Figure 27). The images that are still visible – dominated by a vertically orientated rectangular zigzag line – all seem to be geometric of nature. On its smooth, almost horizontal surface there are two areas of obliteration, separated by the zigzag. Moreover, both obliterated areas seem to have patinated only a little. This may point to rather recent re-use of the boulder. Judging by the scratches the obliteration seems to have been executed from left to right (or vice versa). The lower area seems to have obliterated part of a geometric pattern, while the upper area definitely has obliterated some (abstract?) petroglyphs that are still faintly visible. It seems that the obliteration did not specifically aim at eliminating the petroglyphs and therefore the action probably was unintentional. According to Ron Wolf, who took the picture, the lower, flaked part could be a sign that someone tried (and succeeded?) to steal some of the petroglyphs on this panel (Ron Wolf, pers. comm. 2018).

Figure 27: Obliterated petroglyphs on a boulder at Yellow Jacket, California, USA. Photograph © by Ron Wolf.


Another rock art site in California is Cottonwoods, located about 25 km from Rainbow Canyon and east of Death Valley. It is not the original name of the site, but invented to protect the art. There are many petroglyph boulders in this boulder field, as well as some boulders with grinding slicks without petroglyphs. One boulder has a curvilinear petroglyph near the edge while a large area of the upper, nearly horizontal surface has been obliterated (Figure 28). A small round boulder features in the picture. It may be the original grinding stone. The petroglyph may continue underneath the obliterated area, but if so, it is most likely that it does not concern an instance of intentional obliteration.

Figure 28: Obliterated petroglyphs on a boulder at Cottonwoods, California, USA. Photograph © by Guy Starbuck.


Also a large boulder with a slightly sloping but somewhat concave upper surface at the Terese Petroglyph Site in California (CA-KER-6188; located in the eastern portion of the El Paso Mountain Range) has a large area used for grinding (Figure 29). On the internet I have seen another photograph of this boulder featuring a small stone that was resting on the grinding slick; it is uncertain however if this was the original grinding stone. Near the grinding slick are at least two geometric petroglyphs (one partially within the area with hammer marks) and therefore it is possible that other petroglyphs on this panel have been obliterated. Another boulder at this site has a deep (possibly naturally formed) grinding hollow next to a circular petroglyph.

Figure 29: Obliterated petroglyphs on a boulder at Terese Petroglyph Site, California, USA. Photograph © by Guy Starbuck.

Spectre Wash

A large boulder with a smooth, horizontal upper surface at the Spectre Wash site, also located in the wilderness west of Death Valley, California, has a large abstract pattern that covers almost the entire surface (Figure 30). On the panel is also an area that has clearly been hammered or ground. The boulder may have been selected for grinding purposes because of the flat horizontal top, while the presence of the older petroglyph – possibly once having charged the boulder with specific power – may have been an extra reason to select this stone. But it is clear that this possible grinding slick obliterates part of the older petroglyph and in my opinion this obliteration was unintentional. Perhaps stone surfaces suitable for grinding are scarce in this area. There also seems to be much pecking or hammering outside the possible grinding slick, as well as a circular petroglyph that is less patinated. This may be a more recent addition.

Figure 30: Partially obliterated petroglyphs on a boulder at Spectre Wash, California, USA. Drawing © by Maarten van Hoek, based on a photograph by Andrew Dunning.

Greenwater Canyon

An extensive rock art site at Greenwater Canyon, California, contains many rock paintings, some petroglyph boulders and boulders with areas of hammering or grinding. At least one of those boulders (with two areas of hammering or grinding) has at least two abstract petroglyphs (a solar symbol, a small ring-mark and a double serpentine groove). However, the petroglyphs are well separated from the rather small hammered or ground areas.

Indifferent Obliteration in Arizona

In the central part of the State of Arizona many rock art sites have been recorded. One of the major petroglyph sites is Painted Rocks near Gila Bend, where I know of at least one grinding stone (without petroglyphs, though). Another dense concentration of petroglyph art is found at South Mountains, just south of Phoenix (about 600 km SE of Death Valley). Interestingly, the extensive dissertation describing several petroglyph locations within the South Mountains by Aaron Michael Wright (2011) mentions several cases of obliteration of petroglyph art and especially the captions of their drawings are often most interesting. Unfortunately, my visit to the South Mountains did not include the areas surveyed by Wright and for that reason I base my observation on his drawings.

South Mountains

On Panel AZ T:12:240.104.F (Wright 2011: 668) several petroglyphs have been hammered, but also areas outside the images have been pecked. It is uncertain if this hammering was intentional (e.g. specifically targeted at the petroglyphs). Interestingly, the caption of Panels AZ T:12:307.1.A-B reads: ‘Stippled region denotes an area of prehistoric abrasion over existing petroglyphs, seemingly in a manner to “erasepre-existing images’ (2011: 718; in all cases the emphasis in bold is mine). As the otherwise large areas of abrasion on this panel do not erase specific petroglyphs or parts thereof, this example from South Mountains may well be regarded as indifferent obliteration. The caption of Panel AZ T:12:354.4.B reads: the stippled region over the lizard/anthropomorph in the upper left portion of the panel denotes pecking and abrasion over the existing image, seemingly in a manner toerase it’ (2011: 793). This may be an instance of intentional obliteration, although the hammering continues outside the petroglyph as well. The caption of Panels AZ T:12:390.9.A-B reads: ‘Stippling denotes intense, unpatterned prehistoric pecking across the panel faces’ (2011: 839). Almost the whole rock surface seems to have been pecked over, covering (erasing?) four circular petroglyphs (on purpose or indifferently?). The caption of Panel AZ T:12:390.29.I reads: ‘Images have been abraded, possibly in an effort toerasethem.’ (2011: 854). It is uncertain if this action was intentional. A motif on a broken-off panel (AZ U:9:280.40.A) is said to ‘have been heavily repecked prehistorically.’ (2011: 891). Again, it is unknown if this has been done to obliterate the motif on purpose or not. Finally, the main panel at the “Sunstruck Site” at the entrance to Pima Canyon [AZ U:9:19(ASM)] is said to ‘exhibit a considerable degree of superimpositioning, which has obscured many images; [whereas] the areas of dense pecking are regions of heavy abrasion and/or obliterated petroglyphs.’ (2011: 940). The areas of dense pecking do not seem to aim at specific petroglyphs (or parts thereof) and therefore this may be an example of indifferent obliteration. All those examples prove that in the South Mountains obliteration of previously manufactured petroglyphs (whether intentional or indifferent) occurred relatively often. Interestingly, not only Wright uses the terms ‘erasure’ and ‘obliteration’ to describe panels at South Mountains.

Unfortunately, the extensive study by Aaron Michael Wright (2011) did not cover the Hieroglyphic Canyon area of the South Mountains. However, in this specific area another interesting panel has been recorded and fortunately, the book by Todd Bostwick and Peter Krocek features a photo of that panel, the caption of which reads (2002: unnumbered; my emphasis): ‘Most of the face of this outcrop has been rubbed by a stone, with older designs obliterated and later designs added over some of the rubbed areas,’. Importantly, this example of obliteration occurs on a vertical surface of a large outcrop and therefore it could never have been the intention to grind something. This example of petroglyph manufacture, followed by a layer of indifferent obliteration of large parts of the panel (also including petroglyphs), then followed by secondary petroglyph production, reminds us of the indifferent obliteration of petroglyphs on Boulder TM-Ex-015 (TMP-0275) at Toro Muerto and the later addition of other images on its rubbed area.

Indifferent Obliteration in Oregon

In the south of the State of Oregon several boulders with grinding slicks have been recorded. Especially rock art sites in Yocum Valley and around Mortar Point have an abundance of deep mortars and shallow grinding slicks. A selection of sites with boulders featuring petroglyphs and grinding slicks from Lake County  will briefly be discussed now, but I am sure that there are more examples.

Browsing the internet I came across a web page solely dedicated to petroglyphs and grinding slicks. As I have not seen any of the Oregon sites myself, the following observations are based on those (unnumbered) internet photos and on the literature that I have available. Unfortunately, the photos published on the internet – though useful – are far too small to see details properly. Therefore my interpretations – that are based on those photos (numbered by me in the following paragraphs; not counting the banner-photo) – may be incorrect.

According to the web page all sites described here are found in Lake County in the south of Oregon (the web site does not provide site names or any location detail). Most decorated boulders in this area are of very hard basalt and for that reason petroglyphs are often very superficially executed and can easily be erased by grinding activities. Sites 1, 2 and 3 are found in the neighbourhood of Warner Valley (about 690 km NNW of Death Valley), while Site 4 is located about 40 km north of the Warner Highway (140) through Warner Valley and Sites 5 and 6 roughly 50 km east of Warner Valley.

Lake County Sites

Site 1 (photos 2 and 3) has a boulder with a sloping surface with petroglyphs, some of which seem to continue to the small horizontal flat surface at the top, where also a small grinding slick can be seen. Site 2 (Warner Valley: photos 11 and 12) features a flat boulder that has a large grinding slick that touches a petroglyph of a simple circle (a most common motif in Oregon rock art). Some hammer markings partially seem to cover this ring. It is therefore possible that other petroglyphs have been obliterated by the grinding slick as well. Site 3 (photo 14) also features a boulder with petroglyphs and at least one clearly visible grinding slick plus one poorly visible grinding slick. From the photo that is available it could not be determined if petroglyphs had been obliterated, but especially the poorly visible grinding slick might cover some (parts of) petroglyphs, while it is also possible that the faintly visible petroglyph in the poorly visible grinding slick is a later addition.

Sheep Creek Spring

Site 4, Sheep Creek Spring (photos 6 and 7), includes a number of petroglyph boulders with grinding slicks. Interestingly, some of those ancient grinding slicks seem to have petroglyphs superimposed upon them. The internet photos of Site 4 show one of those boulders (Figure 31). This small basalt boulder has two panels with petroglyphs. The smaller, SE-sloping panel features a fully pecked lizard petroglyph and a few faintly pecked (possibly) abstract images. The upper surface is more interesting in the context of this study. Its slightly concave, smooth upper surface has one, possibly two joined grinding slicks (both only roughly indicated by the orange colour in Figure 31). There also are four petroglyphs on this upper panel, including a small image shaped like a dragonfly. Especially the eastern grinding slick may have obliterated earlier petroglyphs (now invisible) and possibly layers of the still visible purported ‘dragonfly’. Moreover, it seems that especially the six parallel pecked lines have been superimposed upon the eastern grinding slick. Interestingly, the upper panel has also an area of red mineral paint, most of which is visible on the western grinding slick (paint not indicated in Figure 31).

Figure 31: Possibly obliterated petroglyph(s) on a boulder at Sheep Creek Spring, Oregon, USA. Drawing © by Maarten van Hoek, based on photo 7 by Douglas Beauchamp.

Corral Lake

At Site 5, Corral Lake (photos 1 and 4), the top of a large boulder – overlooking the dry lake (Playa) from the south – shows six grinding slicks, all separated by cracks, as well as some smaller petroglyphs. From the photos it seems as if some petroglyphs near the grinding slicks might have partially been obliterated by the grinding activities. I regret not having seen high-resolution photos of this boulder. Therefore it is impossible to judge whether indeed indifferent obliteration took place.

Moon Lake Rims

Site 6, Moon Lake, actually is formed by a conjunction of two lakes, Moon Lake and Weed Lake. The site is located only 3 km NNW of Corral Lake and thus offers more or less the same view to the north. The site we are interested in (which may be located on the west side of Weed Lake just south of the spot where it joins Moon Lake) has at least two boulders with grinding slicks. One boulder (photos 8 and 9) has a small grinding slick near its upper end. Interestingly, a few (zoomorphic?) petroglyphs are surrounding the grinding slick and one even might have been partially obliterated by the grinding activities. There are a few more petroglyphs on this fractured boulder, while to its west a boulder or outcrop that almost touches the boulder with the grinding slick features an almost vertical panel that is completely covered with mainly abstract motifs of varying patination.

The second boulder at this site (photo 10) is also fractured and has two (possibly three) areas with grinding slicks. In one corner are one or two abstract petroglyphs, while near the most distinctly ground surface may be a petroglyph part of which may have been obliterated by grinding. It seems unlikely that other images have been obliterated by the grinding activities (but it is possible). Again, I regret that high-resolution photos of this boulder were not available. Therefore it is impossible to judge whether indifferent obliteration took place.



Although there remains doubt in many instances, this concise study demonstrates that after the manufacturing of apparently important images on natural stone surfaces, subsequent visitors often (partially) erased those images for several reasons. This may have been done by ‘indifferent’ superimposition or by highly intentional obliteration of the whole image or of specific details (in case of the Majes Style Rock Art repertoire often facial details). In other cases however the subsequent visitor indifferently obliterated the previously manufactured petroglyphs by hammering or rubbing (part of) a panel with petroglyphs. In doing this, she or he had no specific intention to erase the ancient images and/or to annul their power. This indifference is demonstrated by the fact that in several instances (parts of) the petroglyph(s) is/are still visible.

Yet there are problems in establishing instances of indifferent obliteration with certainty. First of all, petroglyphs and other anthropic markings are almost impossible to date. Only relative dating is possible (for instance via differences in patination and by clear cases of superimposition) and even then mistakes are easily made. Therefore, when petroglyphs and (possibly) erasing anthropic markings – like a grinding slick – are present on one panel, but clearly separated, then several sequences will be possible. First of all, the petroglyph and the grinding slick may be contemporary. But of course the petroglyph may also be the initial marking; while the grinding slick may be a later feature. But also the grinding slick may be the initial marking; and the petroglyph may be a later addition. However, in all those three cases the grinding slick might have obliterated another petroglyph.

More problematic are the instances where petroglyphs are found very near or even touching the grinding slick. Then the petroglyph may continue underneath the grinding slick. However, there are also instances where petroglyphs are clearly superimposed upon grinding slicks. Several fine instances have been reported from the central Sahara, but in that area there are also instances where petroglyphs are partially obliterated by grinding slicks and there are even instances where a grinding slick is superimposed by a petroglyph, which in turn was again ground. All those possibilities make it very difficult (if not impossible in many cases) to determine what has actually happened.

Yet, besides doubtful examples this study does offer a number of instances of indifferent obliteration where apparently petroglyphs were partially obliterated by grinding slicks or areas of hammering, for instance Motocachy, Boulder MP1-022 (see Figure 14); Tomabal-A, Boulder TOMa-009B (see Figure 16); Cerro Mulato, Boulder CMq-225 (see Figure 18); Toro Muerto, Boulder TM-Ex-008 (see Figure 22) (all in Peru) and Spectre Wash in California, USA (see Figure 30). The best example however is Boulder TM-Ex-015 at Toro Muerto (see Figures 23 and 24) in southern Peru. In that case the rubbed area partially obliterated older petroglyphs and was even superimposed later by small and simple zoomorphic petroglyphs. This example is clear proof that rubbed areas or grinding slicks may (!) have obliterated ancient petroglyphs, even when there is no sign of any petroglyph on the remainder of the panel.

A similar procedure seems to have happened at the outcrop in Hieroglyphic Canyon, South Mountains, Arizona, USA, and just possibly the same is true for the boulder at Sheep Creek Spring, Oregon, USA (see Figure 31), where the ‘dragonfly’ petroglyph might have been superimposed by the grinding slick, whereas the six parallel lines seem to have been superimposed upon the grinding slick (much?) later. However, lacking good quality photographs of all Oregon cases – at least those mentioned in this study – my observations regarding the Oregon examples are by no means certain. For that reason I strongly advocate to use high-resolution / large-format photos – taken at different angels and at different times of the day – in order to establish possible sequences in anthropic activities. Therefore I am grateful to Karoline Juszczyk of the Proyecto Arqueológico Toro Muerto, and to Ron Wolf and Guy Starbuck who generously granted me to use their high-quality photographs of several petroglyph panels.

Finally I would like to emphasise that – in general – every secondary anthropic use of natural stone surfaces with rock art (like additions and superimpositions of petroglyphs, the creation of domestic or ritual depressions and of course of instances of obliteration no matter how dubious) should be meticulously recorded because all this information is of crucial essence to understand the history of the rock art site. This procedure is especially essential in order to establish instances of potential intentional and indifferent obliteration.



I am grateful to Karoline Juszczyk of the Proyecto Arqueológico Toro Muerto who permitted me to use all the (graphical) information that the team of the Proyecto Arqueológico Toro Muerto has available. She also supplied me with high-quality photographs of Boulder TM-Ex-015 (TMP-0275). Therefore, all the photos (except for the inset in Figure 19) of the Proyecto Arqueológico Toro Muerto published in this paper have their logo. I am also much indebted to Ron Wolf and Guy Starbuck who both generously permitted me to use their photographic material and who, moreover, provided me with useful information, some of which has been used in this study. I emphasise however that the observations and opinions expressed in this study are only my responsibility.



BOSTWICK, T. W. and P. CROCEK. 2002. Landscape of the spirits. Hohokam rock art at South Mountain Park. University of Arizona Press, Tucson.

BRACAMONTE LÉVANO, E. and C. A. PASAPERA ROJAS. 2008. Los petroglifos de Cafetal de Tolón y su contexto cultural en el valle bajo del Jequetepeque. Paper presented at the III Simposio Nacional de Arte Rupestre. Huaraz, Perú.

CAMPANA, C. and C. DEZA. 2006. Petroglifos de Yonán. In: ARKEOS, Revista Electrónica de Arqueologia PUCP. Vol. 1; No. 2; pp. 55 – 77.

CASTILLO BENITEZ, D. S. 2006. Arte Rupestre en la Cuenca del Río Chicama. Ediciones SIAN, Arqueología / 4. Trujillo, Perú.

VAN HOEK, M. 2005. Toro Muerto, Peru: Possible Prehistoric Deletion of Petroglyph Details. ADORANTEN 2005. pp. 73 – 80. Underslös, Sweden. 2006: Toro Muerto, Perú. Posibles alteraciones prehistóricas en detalles de petroglifos. In: Rupestreweb.

VAN HOEK, M. 2007. Petroglifos Chavinoides cerca de Tomabal, Valle de Virú, Perú. Boletín de SIARB. Vol. 21; pp. 76 – 88. La Paz, Bolivia.

VAN HOEK, M. 2012. Andean rock art and apophenia: from the macro to the micro. In: Rupestreweb.

VAN HOEK, M. 2014. The Motocachy Pampa Disaster, Peru. In: TRACCE – On-line Rock Art Bulletin, Italy.

VAN HOEK, M. 2016. Los Bordes Aserrados de las Rocas de Chuquillanqui, Perú / The Serrated Rock-edges of Chuquillanqui, Peru. In: TRACCE – On-line Rock Art Bulletin, Italy.

WRIGHT, A. M. 2011. Hohokam Rock Art, Ritual Practice, and Social Transformation in the Phoenix Basin.  Dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy. Washington State University, Department of Anthropology.

Leave a Reply

Your email address will not be published. Required fields are marked *