Does Distance to Animals Change Their Appreciation?
What is the project about?
Originally, the idea behind the "Does Distance to Animals Change Their Appreciation" challenge was to investigate whether distance to animals influences how people perceive them. During the process, this shifted to studying how animals, plants and humans are connected in stories from Hunter-Gatherer communities.
References and links
What is your research question?
The challenge's original question was about measuring physical distance between humans and animals in written sources and comparing them over time. The initial discussion and the consequent choice of the data set moved the group away from physical distance towards the analysis of node distance and sentiment. The current research question is: how can the node distance between humans and animals in tales and written sources be represented to be used for sentiment analysis?
How is the project and/or case situated?
The challenge is part of the push towards improving the transcultural understanding and comparison of environmental appreciation, with particular attention for the sentiments expressed by humans for animals. In this sense it sits squarely in the field of environmental humanities, bridging research questions that bring together anthropologists, literary scholars and historians. It also works as a proof of concept of how extraction and representation of sentiment of animals and environmental features in texts can be achieved.
What methods, data sets, and tools are used?
- The data set is extracted from the Forager Folklore Database (FFDB) and consists in myths and and folktales from hunter-gatherer societies as recorded by anthropologists, linguists, missionaries, etc.
- As regards the tools, a subset of the texts (15 items) has been encoded in TEI by hand, to act as a gold standard for automated encoding of a broader subset of recorded tales. We also use BookNLP to tag humans, plants, and animals in texts we couldn't tag manually. The Gold Standard is used to evaluate this.
- In addition we apply networks and LLM-based methods.
Who is part of the team?
Arjan van Dalfsen, Jan Jokisch, Carina Zander, Nele Mantaj, Wilko Hardenberg