Rock Art and Mutual Information Networks

VIVIAN SCHEINSOHN; INÉS CARIDI

Valle de Uco, Mendoza

Conferencia; 4th Southern Deserts Conference; 2014

Laboratorio de Paleoecología Humana y la Facultad de Ciencias Exactas y Naturales de la Universidad Nacional de Cuyo.

Rock art has been one of humans most ancient visual communication channels. Archaeologists have always been aware of the communicative role of rock art and its information storage capacity. The formalization of this idea was established by the information storage model, which established that rock art is an adaptive process facilitating the storage and communication of information. When studying information stored in rock art, our main problem is that we have lost the meaning of rock art designs. But Information Theory allows us to treat content without any concern for specific meaning. In his classic 1948 paper on Information Theory, Shannon defined theinformation content of a message in terms of the entropy of the message source, something not related to its meaning. So, concepts such as entropy, information content, and mutual information can play an important role in analyzing rock art information. When trying to analyze rock art as transmitting information, the lack of chronological control (when direct dating is not possible) is seen as an insurmountable obstacle. The passage of time between the inscription of a design on a rock and its reading (days, years, or centuries) introduces noise in two ways: (1) attrition or losses of designs by weathering caused by wind, precipitation, and erosion could affect the rock where the art was done, leading to loss of designs or the destruction of painted areas and (2) accretion or addition of designs, which couldbe the result of one or many painting or engraving events on the very same location. This only can be acknowledged by the existence of superposition among designs, although in the face of its absence, accretion could go unrecognized. In any case, the lack of chronological control implies that if we try to establish networks between many rock art sites the result should be a palimpsest of different links and nodes superposed over time, so that anetwork with temporal dimension should be recognized. In this paper, we propose that by using mutual information networks we can thicken the temporal dimension of an otherwise temporally-flat record. We present thetheoretical and methodological issues that our analysis suggests and construct a mutual information network for northwestern Patagonian rock art designs. This network information is interpreted in terms of its temporal dimension and in relation with proposed population models for Patagonia.