1This paper explores the role of rock art as communication and evaluates the challenges to decode meaning in situations where we are unfamiliar with the culture that produces de art, either in the past or in the present.

2Rock art is a visual medium used by past and present pre-literate societies as a non-verbal form of communication and information exchange (Conkey 1978, 1980; Gamble 1982 ; Smith 1992 ; Domingo 2012). During generations it has been used to illustrate stories or to graphically represent different sorts of cultural information to educate, remind, regulate, celebrate and so forth. As summarized in the seminal text by Ucko and Rosenfeld (1967:7) rock art includes both religious and secular imagery, as well as a wide range of signs that assist humans to adapt, coexist and share specific socio-cultural and natural landscapes.

3As any form of communication rock art requires a sender (the person who encodes the information, which in this case will be the artist/s), a message (the information or content to be shared), a channel (the art forms) and a receiver/s (who decode/s the message). For the transmission of information to have maximal effectiveness the interacting agents (sender and receiver/s) need to share a common set of signs, symbols or language. But what happens when there is no common cultural background among the interacting agents? It is likely that the messages would become distorted during the communication process, thus preventing effective communication.

4The depiction of figurative motifs inspired by real object sources may allow literal readings of the art (i.e. a literal identification of the motif as “a deer, a goat, a kangaroo, a human” or an action “a war scene, a dance, a ceremony, a hunting scene”). However, this in itself is problematic, as MacIntosh (1977) demonstrated in his reanalysis of the motifs at Beswick Creek Cave, in the Northern Territory, Australia, when he found that 90 % of his interpretations undertaken in 1952 without a local teacher, were incorrect.

5Moreover, even if a motif is correctly identified this does not necessarily mean that the communication is effective, since it is likely that the image encodes a symbolic message that is unreadable to an outsider who is not initiated into the knowledge system. For example, the image of a deer may represent a prey, but it may also represent a spirit or an ancestor, a desired strength or ability, the identity of an individual or a group, and so forth. The likelihood of successful communication is even more tenuous when depictions include non-figurative objects (geometric and abstract depictions), which are interpreted as a simplification or abstraction of the concept to be shared.

6This raises the question: Is it possible for outsiders to understand the symbols of a different culture?

7When travelling to a country where we are unfamiliar with the language, the signs and/or the culture, we all experience some sort of miscommunication. Images and signs can become meaningless to us, and thus the communication process is broken. The same motif or sign can embrace multiple meanings that vary according to the context or the culture. When the same animal is depicted in the imagery of several cultural groups we need to analyse the context to understand the specific meaning. Thus, for example, the image of a red rooster in an Australian sign refers to a place to eat roasted chicken: the Red Rooster restaurants. The same animal depicted in a church, like the rooster sculptured in one of the doors of the Gaudi’s famous Sagrada Familia church in Barcelona, is linked to a religious story: the rooster crowing upon Saint Peter’s third denial of Christ. Finally, the Portuguese rooster known as the Galo de Barcelos is often used as a symbol of Portugal, and thus represents a cultural identity (fig. 1). So in a contemporary setting, representations of the same animal are used in different countries and cultures to refer to very different things: food, a myth or an identity.

8This sort of miscommunication also happens when trying to interpret the past through the analysis of archaeological remains. While when travelling to a different country a local guide can inform us of the meaning of the images, when travelling to the past we can only rely on the archaeological context, which may inform us about the potential function of the art, but rarely about the symbolic meaning (Domingo and May 2008).

9In this paper we use our ethnoarchaeological research in the Arnhem Land and Barunga regions of the Northern Territory (Australia) to argue that focusing our research efforts on deducing the meaning of the art only leads the discipline to a dead end. Moreover it overshadows the potential of rock art to explore other aspects of past societies such as their natural and cultural environment, their material culture, including the identification of ephemeral material culture that is underrepresented in the archaeological record; technological developments, social boundaries and territorial behaviours, and so forth.

Methods

10Today, there are only a few places in the world where rock art is still part of the symbolism of a living culture and where we can still (to varying extents) access the meaning of the art guided by the knowledge of the authors of this art or their descendants. The Arnhem Land and Barunga regions of the Northern Territory (Australia) are two of these few places. This is where we will travel in this paper to illustrate our discussion.

11Using several examples from Arnhem Land and the Barunga region we will explore the different functions of rock art, how cultural information is embedded in the imagery and to what extent the information in the art can be decoded by an outsider to the culture that produced the art.

12To answer these questions we follow a three step method:

• The first step is to select the sites, record the motifs and scenes and analyse them from an archaeological perspective (i.e. through description and quantification of motifs, techniques, themes, patterns of composition and addition and so forth). A detailed example of this process can be found in Domingo (2008), May and Domingo (2010) and Domingo (2011). At this point our aim is to test the extent to which we can identify the motifs and scenes depicted from a purely archaeological analysis.

• The second step is to analyse the archaeological context, looking for evidence of human activities to deduce the potential uses of the site, and thus suggest hypothesis concerning the function of the art.

• The third step is to conduct ethnoarchaeological research to attempt to record the meaning/s that Indigenous populations assign to the motifs, themes, scenes or sites under study, to test the validity and limits of our interpretations. To do this, we combine oral interviews with Indigenous elders and artists, with a literature review including early ethnographic records.

Introduction to the research area

13The research presented in this paper was conducted in Australia. Prior to European contact Australian Aboriginal and Torres Strait Islander groups showed great cultural diversity, with different languages, cultural practices and material culture, as demonstrated in Horton’s (1999) map of Aboriginal Australia. While they were mostly semi-nomadic, they had strong connections to particular parts of the country. At the time of contact there were numerous rock art traditions in Australia, with more geometric art system in the centre and a range of different forms of geometric and figurative art in other regions (Morwood 2002). Our research focuses specifically on the Western Arnhem Land and the Barunga regions of the Northern Territory, where figurative painted rock art predominates (fig. 2), though there are also some engravings or abraded grooves. Unlike other Australian territories, where the arrival of Europeans severely impacted on the Indigenous population and their culture, both Arnhem Land and the Barunga region were declared as Aboriginal reserves in the 1930s. These declarations attenuated the impact of European invasion, allowing the continued operation and continuation of one of the longest rock art traditions in the history of mankind.

14Over a period of around 50,000 years (Roberts et al. 1990: 153 ; Roberts et al. 1998) the Arnhem Land and Barunga regions were inhabited by a variety of Indigenous populations with an economy based on hunting and gathering. These populations organized in flexible bands and practiced some sort of mobility in search of seasonal resources (see, for example, Chaloupka 1981). While their technology was relatively modest, they developed one of the most sophisticated socio-cultural lives recorded by anthropologists, with a significant investment of time in cultural practices, including art, religion and laws (Morphy 1992; Flood, 1997: 2; Smith 2002, 2004). In these societies, ceremonies played a significant role in the transmission of knowledge and cultural practices. In ceremonies religion, history and laws were melted through dance, music, stories and various forms of art (body art, rock art and portable art) to ensure the appropriate training of the new generations, the exchange of ideas and raw materials and appropriate rituals regarding the various stages on life (see, for example, Berndt and Berndt 1970; Morphy 1992; Smith 2004 and Taylor 1996).

15The arrival of Europeans in the Northern Territory in the 1880s had a significant impact on Indigenous ways of life. However, interruption of their traditional economies and a general transformation into a more sedentary lifestyle did not cause the disappearance of their beliefs and socio-cultural practices. On the contrary, Indigenous identities were reinforced in comparison to the "others" and continued to ensure the transmission of ancestral knowledge to new generations (Taylor 1996; Domingo and May, 2008; May 2008). This cultural continuity offers a unique context to test to what extent an archaeological approach to rock art is useful in interpreting the rock art of a different culture, when we lack information about their socio-cultural and symbolic traditions.

Functions and interpretations of rock art in Arnhem Land

16In Arnhem Land, rock art has been used in different contexts to convey ideas, beliefs, ancestral knowledge, traditions, identities, stories, social behaviour and laws. As local Elder Thompson Yulidjirri explained to one of us, the art is not an end, but a gateway (May, field diary 2001, May 2006: 43).

17The meaning of this art is mostly related to the so-called ‘Dreaming’, an English term for the creation era in which ancestral beings travelled the earth producing its topographic features (Morphy 1992). Much rock art of these regions is the graphic expression of beliefs, sacred values and rules the creation beings passed on to humans (Smith 2000). Thus, to interpret the art it is absolutely necessary to be familiar with the Indigenous culture and beliefs system.

18The imagery of the Arnhem Land and Barunga regions include Ancestral Beings as well as evil spirits, usually full of sacred symbolism. In Arnhem Land, Yingarna (the creation mother) is one name for the ancestor whose journey resulted in the creation and dissemination of Aboriginal people and language groups in northern Australia (fig. 3). She is said to have emerged from the sea to the north and travelled inland, teaching people language and clan. She can adopt different physical forms, either a female figure or a rainbow serpent. Thus, both images would evoke the same story to a knowledgeable audience, whereas to an outsider they would be identified as completely different motifs with no evidence of a link between them, or to the Yingarna narrative.

19Similarly, in some parts of this country Milingi can be portrayed as either good or bad spirits. For Rembarrnga people they are good, but at Barunga they are often thought of as mischievous. For example, they can trick you and can turn the landscape around so you get lost (fig. 4). Visually speaking, they are human like forms. For someone familiar with this culture they are easily identifiable since their rib bones are shown, whereas they are not shown for people. Once again, without ethnographic information all an outsider can tell is that they are different to other human figures. We could easily analyse their spatial distribution and their place in the artistic sequence to explore questions of cultural change and territorial behaviour, but we will be completely unable to obtain meaning per se.

20In these territories, there are other secular representations that reflect Indigenous concerns (love magic, commemoration of events, educational purposes) or representing daily life (hunting or fishing scenes). These sorts of images are said to be produced to fill the time or for story telling in public contexts (Chaloupka 1993), thought it is possible that they also have other non-public levels of meaning.

21Sorcery figures are usually human beings depicted in strange or inverted postures, sometimes with distorted genitalia (fig. 5). The aim of these representations is to cause illness or death to another person for a range of reasons (Chaloupka 1993). It would be difficult for an outsider to deduce this magical purpose by simply visualizing the painting, though they could record stylistic differences of these images in comparison to other human-like figures.

22Rock art in the Northern Territory also has numerous examples of plant and animal resources that have been used for generations as food or medicine. At Injalak Hill in Arnhem Land, representations of fish, either in pieces or complete, are used to teach how to process the meat and which parts are the most valued when distributing the food (fig. 6). But fish are more than food, they are powerful symbol of life. Fish can be tied to maps showing the territories of various clans, or they can be used to tell stories of the creation time, and so forth (Taçon 1994: 124).

23Sometimes the motifs depicted in these regions represent objects we are not familiar with (fig. 7). In such a situation only Indigenous people can help us to interpret them.

24We are in a similar situation when we move from the interpretation of individual motifs, to the interpretation of compositions and scenes, and sometimes even when we are interested in knowing the function of some sites. An Indigenous teacher is necessary to understand the themes depicted or the function of some sites, as exemplified in our discussion below.

25At the site known as Djulirri (Anuru Bay) there is a scene featuring various white human figures, with stick like bodies and no formal differences among them (fig. 8). Analysing the distribution of the motifs and their actions, we could guess of a possible fight (a complex battle). However, Indigenous elders informed us that the interpretation of this scene is far more complex. In fact, it is a ceremony related to a funeral. Thus, ethnographic information is key to interpreting the activity depicted in this scene as a ceremony instead of a war. Similarly, only with ethnographic information could we interpret the circle depicted in front of the face of the two leading individuals in the top right confrontation. In Arnhem Land, ‘biting bags’ (also known as power bags or spirit bags) filled with feathers or vegetable fibres and stitched closed were worn by men during certain ceremonies (Hamby, 2011: 227). It would be placed between a man’s teeth when he was fighting to give him power and courage. Since most of the rituals relating to this ceremony are restricted, further details were not discussed with Indigenous teachers. Moreover, such details were not necessary for our research.

26Another scene at the Djulirri site includes several stick like human figures, with differences in size, action and in the objects they are holding (fig. 9). Individuals on the left are holding an object on top of their heads and seem to be performing some sort of dancing. Thanks to our previous knowledge of Indigenous traditions we were able to recognise some of the objects being held by some of the individuals on the right as musical instruments. The second individual from right to left is holding a didgeridoo (a musical wind instrument traditional to this Australian region). The fourth individual is holding clap-sticks (a musical percussion instrument that men strike to accompany the didgeridoo during their ceremonies). So from these two instruments, and analysing the distribution of the motifs and their action we deduced that the group on the right were musicians and that the scene was depicting some sort of ceremony. However, we were not familiar with either the objects held by the group on the left, nor with the ceremony depicted. From our archaeological analysis we could not really tell much about this scene, apart from guessing that there might be social differences between the biggest and the smallest individuals depicted, perhaps due to age or social status.

27Through ethnography we were informed that the ceremony depicted was the Ubarr ceremony, although further details about this ceremony were not discussed.

28Far from this region, and travelling to the south, in the land near the Aboriginal communities of Barunga and Wugularr there is another interesting site that provides an example of the role of ethnography in interpreting both the meaning of the art, and the function of the site. The Mook Mook site depicts a creation story: the creation of owls, which are being ‘vomited’ by the rainbow serpent (fig. 10). The motifs at this site are quite naturalistic, and while an outsider familiar with these animals may be able to identify them, in the past they have been erroneously identified (Davidson 1981). However, it is only with ethnographic information that it is possible to identify the narrative represented. Furthermore, ethnography also is necessary to understand the function of this site. An archaeozoologist could easily deduce that the bones deposited in the base of the painted panel are kangaroo bones, and all of them correspond to the same anatomical part. An archaeologist might surmise that this is a ceremonial site from the arrangement of kangaroo femurs and the fact that they are coated in red ochre. However, there is a big step from these observations to the ethnographic knowledge that the site is related to an initiation ceremony for young boys, in which the boy’s mother has collected the femurs of the kangaroos hunted by the son as he proves himself to be a good hunter and ready to start a family.

Results and discussion

29In this paper we have used ethnoarchaeology to argue that rock art research should go beyond the search for meaning in rock art, to explore more quantitative and qualitative aspects of this form of archaeological remains.

30Through our research with Indigenous elders and artists in the Arnhem Land and Barunga regions of the Northern Territory we have learnt that visual arts are a mean of communication that can be used for a wide range of purposes: to illustrate a story, to mark a place, to provide memory of an event, to depict rules for how to behave and to communicate ownership or identity. The meaning and function of the art change according to the context or the culture. Thus, to obtain meaning from rock art it is necessary to be familiar with the cultural beliefs and traditions in which the art was conceived.

31Through different case studies we have demonstrated that while some of the information encoded in rock art can be perceived from an archaeological analysis of motifs or scenes and the archaeological context, a full comprehension of the beings, traditions and activities represented in rock art and the complex social rules and regulations depicted can only be achieved through a full understanding of the cultures who produced the art. In our view, focusing on finding the meaning of prehistoric art masks the potential of rock art to provide information on other quantitative aspects of past societies, such as territorial behaviour, technological practices, material culture, natural environment and so forth.