Rock Art Analysis Package (RAAP), v01: Data, Schema and Scripts for the Formal Analysis of Rock Art Data

Rock Art Analysis Package (RAAP), v01: Data, Schema and Scripts for the Formal Analysis of Rock Art Data

Author: Dr Mick Morrison, University of New England.

Last major update: 14 July 2024

This repository contains data, schema and scripts for analysing rock art motifs identified as part of the Kuuku I'yu Rock Art Project, within the Kaanju Ngaachi Indigenous Protected Area, Cape York Peninsula, Australia. The project was conducted in collaboration with the Chuulangun Aboriginal Corporation.

The goal of this repository is to enable others to reproduce the tables and figures used in our research, or to use our data, schema or scripts in other work. Please acknowledge this. You can cite the RAAP as:

Morrison, M., Marshall, N., & D. Claudie (2024) Rock Art Analysis Package (RAAP): Data, Schema and Scripts for the Formal Analysis of Rock Art Data [Data set]. In Australian Archaeology #### v.01. https://doi.org/10.5281/zenodo.12730323

The RAAP was developed via the research project presented in the following paper. If you use these data or schema, please cite and acknowledge this work:

Morrison, M., Claudie, D., Marshall, N., and D. McNaughton (Under Review) Storyscapes: Rock art, personhood and social interaction in the northern Cape York Peninsula, Australia. Australian Archaeology.

Abstract

This study analyses rock art assemblages and late Holocene social interaction in Cape York Peninsula’s (CYP) northern highlands, north-eastern Australia, from the vantage point of Kuuku I’yu (Northern Kaanju) Pama (Aboriginal People from CYP) homelands. Previous research has documented complex social networks across northeastern Australia and southern Papua New Guinea, highlighting long-distance alliances, exchanges, and kinship relations that connected island, coastal, and inland communities. However, local and interregional interactions remain under-researched, especially between coastal and inland communities on CYP’s mainland. Drawing on analysis of ethnographic materials, it is shown that Pama personhood, social identity, and social interaction in the past were underpinned by totemic, moiety, and kinship relations, as well as Stories—shared local and regional cosmogonies and cosmologies. Using digital methods centred on photogrammetry, here we analysed nearly 300 rock art motifs. Results reveal place-based variations in motifs indicating historical social distinctions. The study identifies potential links between rock art motifs, personhood, totemic beings, and major Stories. We draw on the concept of Storyscape as a culturally informed theoretical framework for understanding histories of social interaction and varied regional and local cultural influences on rock art. This reflects Indigenous ontologies of embodied relatedness, which potentially underpin recent and potentially late Holocene social interaction and material exchange spheres on the CYP mainland.

Ethics and Indigenous Cultural and Intellectual Property

This research was approved through the University of New England Human Research Ethics Committee, Approval number HE20_219, as a component of the Australian Research Council-funded Sugarbag and Shellfish: Indigenous Foodways in Colonial Cape York Peninsula (LP170100050) project (2019-2024). It formerly had ethics approval at Flinders University before the project and investigator moved to UNE.

This repository does not include images of rock art or Indigenous heritage places, as these remain the cultural property of Kuuku I'yu (Northern Kaanju) Custodians and are restricted. 

Approved images are available in the published paper or via the Chuulangun Aboriginal Corporation associated media releases.

Requirements

This is a basic to intermediate package of tools and you will need:

some experience with R;

to be comfortable modifying the scripts provided to suit your system and needs;

R (version 4.0.0 or later);

RStudio (not essential, and another R IDE should be fine)

R packages: data.table, dplyr, tidyverse, officer, readxl, pivottabler, openxlsx, basictabler, flextable, ggplot2, FactoMineR, factoextra, DescTools, corrplot, ca, dummy, viridis, kableExtra These are listed in the relevant scripts and are only required if you are trying to run a script from this release. Some can take time to get working correctly, as dependencies will need to be installed either as an R package, or onto your system generally.

Scripts

script01_datacleaning.R

This script cleans the original dataset created during Marshall's Masters thesis research (2018, Flinders University). It performs various transformations to the data and saves the cleaned data for further analysis as an RDS binary file.

Use this script if you want the cleaned data as a spreadsheet. You'll need to add a line of code to do so at the end of the script, as the intention here is reproducibility via the use of R. The line write.csv(data, "path") will likely do it. 

Data outputs (RDS files) are in `/data`

script02_tables.R

This script loads the cleaned data produced in script01 and produces the tables summarising datasets in a *.docx file. If the script is unmodified from the source, it will generate the published tables and appendices of the published paper in sequence to /outputs within your working directory.

script03_figures.R

This script produces a series of figures to accompany the manuscript. It includes statistical analyses such as chi-square tests and correspondence analysis, with results saved as image files to /outputs

Data

motifdata2018v03

The data provided is a modified version produced by co-authors Marshall and Morrison during the desktop analysis of digital models and field images of Kuuku I'yu rock art, as described in the published study. These are uncleaned so are not consistent with published data. Run the scripts to reproduce that dataset.

Classification Schema

Marshall (2018) produced the following tables of values used to classify identified motifs. These can be used to create a simple database if required. See the published paper or Marshall (2018) for further details.

Usage 

 Clone the RAAP repo to your computer. Cloning or forking from Github is recommended, and for the simplest method use Github Desktop. You can also download the latest release as a Zip.

 Create a new project within RStudio, and save to the RAAP directory.

 Open all scripts, and modify the path and working directory in the first few lines to suit your system. You'll need the full path. Save

 Open Script 1, and run (select all and run). Verify that it has worked correctly by checking that your /data directory now includes a series of binary files. No warnings or error messages should appear in R Console.

 Open Script 2, run and if it runs without error, check the outputs directory. It will now include tables.docx generated by the script.

 Open Script 3, run and if it runs without error, check the outputs directory. It will now include three figures generated by the script.

This is just a basic overview, the scripts can be modified to use other data.