We use cookies to ensure that we give you the best experience on our website. You can change your cookie settings at any time. Otherwise, we'll assume you're OK to continue.

Durham University

Research & business

View Research Directory

Hand stencils in Upper Palaeolithic cave art

A research project of the Department of Archaeology.


The human hand forms one of the most ancient themes of human art. Prehistoric examples of hand prints (positive images formed by covering the hand with paint and placing it on a surface, rather like modern children create) and stencils (negative images formed by placing the hand against a surface and blowing paint around it) are known from prehistoric contexts in Latin America, the Sahara, Indonesia, Australia and Tasmania, in many cases dating back several thousand years. For decades these have been thought to be Mid Upper Palaeolithic in age (around 22-29,000 14C BP) but recent dating and critical evaluation of existing data have shown that they are among the earliest examples of European Upper Palaeolithic cave art., stretching back at least to 35,000 (calendar) years ago.

Negative stencils - far more common than prints -were created by projecting a fluid paint of either black colour (using the pigment manganese) or, more commonly, red (using haematite – ‘red ochre’) at the hand. Usually, stencils cluster in certain areas of deep caves – often areas difficult to access, or features of significance such as small ‘chimneys’ produced by geological erosion or curtains of stalactites. On the latter, the stencils stand out against the sparkly white background, made more mysterious by the flickering light of the small animal fat lamps used by Palaeolithic artists to explore the deep caves.

The experimental replication of hand stencils shows that they were best produced by blowing watered-down pigment of runny consistency through hollow tubes, and the recovery of ochre-stained shells and bird bones below stencils in several caves reveals the specific equipment used to do this. A large bivalve shell was used to contain the runny paint, and a short tube inserted into it like a straw. Holding the ensemble close to the subject hand a second tube was used to blow across the first. This created a vacuum which sucked the pigment up from the shell and out as a fine spray. This created the characteristic diffuse cloud of colour around the hand, while revealing the hand in sharp outline. The technique is difficult to master and quickly results in light-headedness. It also produces a strange, loud whirring and whistling noise. It is easy to imagine how the strange and dangerous deep cave environment, discomfort, light-headedness and whistling combined to create a magical experience, and for this reason archaeologists assume that the art held magical or ritual significance rather than being simply Palaeolithic graffiti.

Around 43 French and Spanish caves are known to contain Palaeolithic hand prints and stencils. Many of these cluster in the great centres of Pleistocene art; Asturias and Cantabria in Spain, and Ariège, Lot and Dordogne in France, although they are found as far afield as southern Spain (e.g. Ardales cave near Malaga) and southeast France (e.g. Cosquer cave near Marseilles). It is possible that the hand was a common image in Upper Palaeolithic art in general and that only those prints and stencils left on hard surfaces in the stable microclimates of deep caves have survived, whereas examples possibly created on the body, on clothing and on other perishables have disappeared. As with all of the art of small-scale societies one can assume that hand stencils functioned in some way other than as pure ‘art for art’s sake’. As with today caves were probably explored out of curiosity, although there is much evidence that the art left of their walls was magico-religious in nature. Humans interacted with cave walls in many ways during the Upper Palaeolithic, possibly because they were thought to form the thin veil which separates this world from others, a remarkably common notion across the world. In addition to painted images of the animals that were critical to the survival of these hunter-gatherers - such as wild horse, deer, ibex, bison and extinct cattle – finger tracings meander in the soft clays of the cave walls, dots were produced in red pigment with the fingertips, and in places the shiny white ‘moonmilk’ – a calcite-rich clay produced by the erosion of cave walls – was scooped away with the hand, probably part of ritual and magical acts that hand stencils arose out of.

Palaeolithic hand stencils usually occur singly or in small groups, but in a small number of caves such as Gargas in the French Pyrenees and Maltravieso in southwest Spain a few hundred examples are known, where they cluster in ‘panels’ in which the hands of several individuals are identifiable. The positioning of the stencils seems to have been done with little regard for comfort; flat areas of cave wall which are very well suited to decoration were often ignored, and stencils could be placed closed to the cave floor or high enough to require stretching. In either of these cases holding a hand still while blowing pigments becomes especially difficult. Stencils of childrens’ hands are known among the art, although adult hands are more common. Human hands can be identified to gender with around an 80% degree of confidence: in males the ring finger (the growth of which is controlled by testosterone) tends to be longer than the index finger (controlled by oestrogen), whereas the reverse is true for females; thus the resulting 2nd to 4th digit ratio is a fair indication of gender. Applying this to known Palaeolithic hand stencils it seems that both adult males and females were ‘depicted’ although the latter appear to be more common. Whether artists worked alone or in pairs is unclear: in many cases it is entirely possible for an artist to create a stencil of her/his own hand, and as most stencils appear to represent the left hand (assuming they were created palm down) this seems to have been the case. Experimental replication has show, however, that it would have been impossible for one artist/subject to create stencils in some of the positions in which they have been found, revealing the collaboration of a separate artist and subject. In some panels the work must have taken a number of hours – probably around five in the case of a structured panel in Pech Merle Cave in Lot, France - showing that the creation of stencils was a deliberate and planned ritual act rather than random marking.

Perhaps the most mysterious aspect of Palaeolithic hand stencils is that on many examples one or more fingers are missing. Early theories saw this as reflecting fingers that really were missing from their subject’s hands, either through frostbite, accidents during flintkapping, or through ritual mutilation. The former have been rejected: these societies were certainly tailoring sophisticated clothing in order to survive the harsh Ice Age climate and mittens would not be beyond their technological ken, and no flintkapper, however unskilled, would sever an entire finger. Deliberate mutilation, perhaps as part of the rituals that accompanied a rite of passage or as punishment, remains a possibility. In many cases, however, two, three or four fingers are missing from stencils, and the deliberate removal of this number – which would render the hand useless – would be suicidal to small-scale hunter societies in such hostile environments. It seems far more probable that such stencils with attenuated fingers resulted from the deliberate bending back of fingers, which have been created experimentally. In some cases too, little fingers seem to have been deliberately painted over, adding strength to this observation. In such cases unskilled work results in the blurring of finger outlines, although most Palaeolithic examples are sharp, suggesting perhaps a degree of artistic familiarity and resulting skill.

It seems sensible therefore to regard finger attenuation as a form of symbolic communication, and therefore that individual hand stencils could be distinguished and meaning could be read from them. What this meaning was one can only guess: did it reflect individual names or concepts? Was it recording ritual acts? In some cases the stencils were placed in clear association with images of animals – two near life-size dappled horses in the case of Pech Merle – and thus seem to be part of wider symbolic stories. Hunter-gatherer societies operating in the last few centuries have employed hand symbols in which fingers are similarly bent; perhaps it is no coincidence that in such systems animals form a large part of the vocabulary, as hand signals would allow silent communication during the hunt. Perhaps in the Palaeolithic such communication systems extended into the ritual world, a blurring of art and belief in the world of survival. In this sense hand stencils formed some of the earliest known artistic forms of human communication.


Most research on hand stencils for the last half century has focussed on the gender and handedness of those stencilled, and why fingers appear to have been bent or missing in a few caves. To look past these questions our research focussed on the context of the hand stencils in several caves in France and Spain. Were they randomly placed, or were decisions being made about where these should be placed? If so, with what associations? The results inform about decision making during the earliest period of cave art production in Europe. Our research was funded by the Leverhulme Trust.

We have also engaged in a considerable programme of public outreach – particularly at Creswell Crags – where members of the public are shown how to produce stencils. Examples of experimentally produced stencils also featured in a recent exhibition at The Royal Society.


We took a number of observations on all visible hand stencils in the caves of El Castillo, La Garma, Tito Bustillo, Ardales (Spain) and Pech-Merle (France). Measurements of finger length and palm width were taken as these inform about the identify of those individuals whose hand outlines have been preserved in the caves, as were details of any associations with features of the cave walls that might inform about decision making.


Using morphometric data we were able to show that in all cases only a small number of individuals – 3 or 4 – left prints in each cave, and the size and finger ratios of these were consistent with these being female. We were able to show that in all caves studied the greater majority of hand stencils were associated with specific features of the cave walls. An association with small bosses on the walls which the hands appear to be ‘gripping’ and with slight concavities in the wall atop of which the palm seems to have been ‘fitted’ were particularly common. Associations with cracks were very evident too; perhaps enforcing the notion that cracks in cave walls form points at which this world meets another. In a number of cases these associations were made irrespective of how comfortable or uncomfortable the position necessary for their production was, further supporting the notion that there were very deliberate reasons for the placement of stencils above and beyond simple ‘grafitti’. It seems that some of the ‘rules’ for the production of cave art emerged at its earliest stages in Europe.

Some hand stencils were created in positions difficult to access. Here, Paul Pettitt’s hand marks the position of a black stencil created atop stalactite in Ardales Cave (near Malaga, Spain) Photo courtesy Pedro Cantalejo Duarte and Ardales Cave.

Some hand stencils were created in positions difficult to access. Here, Paul Pettitt’s hand marks the position of a black stencil created atop stalactite in Ardales Cave (near Malaga, Spain) Photo courtesy Pedro Cantalejo Duarte and Ardales Cave.

Positioning of a stencil in El Castillo (Cantabria, Spain). Note the importance of shadow of the stenciled person’s arm. Photo Becky Harrison and courtesy Gobierno de Cantabria.

Positioning of a stencil in El Castillo (Cantabria, Spain). Note the importance of shadow of the stenciled person’s arm. Photo Becky Harrison and courtesy Gobierno de Cantabria.

A hand stencil associated with natural cracks on the wall of El Castillo cave (Cantabria, Spain). Photo Paul Pettitt and courtesy Gobierno de Cantabria.

A hand stencil associated with natural cracks on the wall of El Castillo cave (Cantabria, Spain). Photo Paul Pettitt and courtesy Gobierno de Cantabria.

Published Results

Journal Article

  • Pike, A. W. G., Hoffman, D. L., García-Diez, M., Pettitt, P. B., Alcolea, J., De Balbín, R. González-Sainz, C., de las Heras, C., Lasheras, J. A., Montez, R. & Zilhão, J. (2012). Uranium-series dating of Upper Palaeolithic art in Spanish caves. U-Series Dating of Paleolithic Art in 11 Caves in Spain. Science 336(6087): 1409-1413.


From the Department of Archaeology

Related links