Tracing carvers on the rocks

The partners have reached important milestones with digitization, 3D technologies, Artificial Intelligence (AI), and chronological observations on rock art that allow an indepth investigation of Bronze Age carvers. This project will drive digitalization of rock art documentation, rock art panels, and granite surfaces to provide data to train AI algorithms to recognize motifs, styles, superimpositions, and carved vs. uncarved features. This work will provide new fundamental insights into rock art chronologies, the long-term use of panels, and different carvings techniques. With the work in this interdisciplinary project, new interpretations and theories about rock art carvers, their knowledge, and their social roles can be put forward.

Background

Swedish Bronze Age rock art (1700–500 BC) constitutes one of the world’s richest and bestpreserved prehistoric legacies. Over the years, the rock art has been used as evidence to verify ideas about Bronze Age ideology, religion, long distance connections, means of transportation, warfare, subsistence, landscapes, and socialorganization (Ling 2014).

Despite the rich scholarship, very little attention has been paid to the rock art carvers themselves. Currently, thereis little knowledge about the kind of techniquesthat were used when the images were carved.
Moreover, we do not know if the rock art is a result of an organized praxis carried out by an institution, or if they should be regarded as an outcome of a more individual expression.
Consequently, we lack information that could help us to forward theories about the social roles of rock art carvers in Bronze Age society.
However, recent advances in different research fields enable an in-depth investigation to identify the carvers and/or their institutions:

1. The Swedish Rock Art Research Archives’ (SHFA; www.shfa.se) digitization of rock art documentation (64,000 2D records out of a total of 105,000) provides a good starting point for quantitative and qualitative approaches to identify carvers. 
2. New chronological observations in rock art that allow us to analyse the different phases of production with greater accuracy. 
3. New digital 3D techniques to record, analyse and interpret rock art and granite structures more precisely.
4. New advances in AI that forward quantitative and qualitative approaches to rock art documentation and structural geology

With the support of these developments, it is now possible to advance questions about various carving techniques and engravings from different phases. These recent advances give us the potential to identify individual artists or groups of carvers from the Bronze Age. Thus, this project aims to digitize high-quality rock art documentation and to advance methodologies and bring to the fore new information about the dating of rock art, carving techniques, and knowledge about carvers. It is a highly collaborative project between Dep. of Historical studies/ SHFA (G.U), Vitlycke Museum (Västarvet), Gothenburg Research Infrastructure in Digital Humanities (GRIDH) (GU) and Dep. of Earth Sciences (GU).

Purpose and aims

Who were the carvers, and when and why did they carve?
The main objective with this project is to enhance our knowledge about the Bronze Age carvers.With the new opportunities that digitalization, big data, and artificial intelligence (AI) approaches provide, and by combining new and old methods, we are now able to approach this important research issue. The ultimate goal is to identify and link groups of images to carvers or groups of carvers from the Bronze Age. When numerous images are identified and combined, we expect to be able to reconstruct the historical dynamics on the rock art panels, and thus, demonstrate how various carvers used and reused different locations in the landscape. Another aim is to localize centres as well as surrounding satellites of rock art production in the landscape. This can be done by exploring chronological, stylistic, technical, and geographical variation. In doing so we aim to answer to the following research questions:
 

• Can we identify significant time differences between carving events and/ or specific phases of increased rock art production?
• Which techniques have been used to carve the rocks in Tanum? Can we identify individual carvers or workshops with the help of new digital techniques?
• Were there significant geographical shifts of rock art production in the landscape and can we detect local hierarchies in the complexity of themes?

In order to address our research questions and attain our goals we have designed an
interdisciplinary project involving archaeology, structural geology, data science and cultural heritage approaches. The project will develop new methods to accurately analyse, identify and classify rock art documentation. These data will then be used to train AI algorithms (see Methods) to recognize and quantify motif and style variations, superimpositions, and granite rock and grain structures.
We define the following analytical steps: 1) Investigate the chronology of the motifs by exploring the stylistic and typological variability of rock art, for example boats, metalwork, wagons, etc. 2) Investigate the way motifs were constructed and changed by studying superimpositions ofmlines and motifs. 3) Investigate different carving techniques by analyzing the rock and grain structures of carved and uncarved Bohus granite. To accomplish these goals, we propose three work packages, each linked to corresponding analytical goals:

• Work Package 1 (WP1) will address motif variability on rock art panels and their chronology(goal 1).
• Work package 2 (WP2) will investigate superimpositions of scenes and motifs on rock art panels (goal 2).
• Work Package 3 (WP3) will investigate the correlation between natural rock and grain structures, and carving techniques (goal 3).

With this systematic approach the project will push the boundaries of rock art research by
informing us about Bronze Age rock art carvers and the processes that triggered carving events.