1. We have created a rigorous methodology, terminology and work template for gathering, defining, and categorising over 4000 distinct plot features in different versions of the same narrative. By undertaking manual plot feature annotation alongside AI-generated plot feature generation, using Gemini-Pro, we have developed effective prompts and gathered valuable interdisciplinary insights into the strengths and weaknesses of manual and AI approaches to this task, based on the current stage of development that the AI LLM models are at.
2. We have gathered scholarship on 'Floire et Blancheflor' from across a wide range of European regions in Old French, Anglo-Norman, Middle English, Middle High German, Middle Low German, Middle Dutch, various medieval Spanish dialects, various medieval Italian dialects, Old Swedish, and Old Norse. In doing so, we have been linking up disparate scholarly knowledge from across nine different philological field of studies and been using knowledge from the Digital Humanities to provide a shared framework for categorising and comparing them.
3. We have been drawing on a range of tools form the Digital Humanities, including OCR-Kit and Transkribus to access and transcribe a range of old editions and unedited manuscripts to which there has hitherto been no online access. We have also been working with a range of translation programmes, primarily Deepl and GeminiPro, to undertake digital translations of medieval texts and have been checking and refining those translations based on our expertise in the 'Floire et Blancheflor' material. In this context, we have been overcoming the linguistic barriers that have lead to historical text versions in less widely spoken languages, such as Old Swedish or medieval Greek, being neglected in terms of comparative study.
4. We have hosted an international two-day workshop and presented papers at a variety of international and national workshops and conferences in order to publicise the findings of the project, with an emphasis on discussing our mixed-method work template and embedding it into current best practice discourse for comparative medieval study. In the context of ensuring the possibility for all researchers on the team to develop a clearly identifiable, independent research identity, while contributing to the project, we have also presented several papers focusing on comparative studies of individual versions of 'Floire et Blancheflor' in their cultural and manuscript context, with an emphasis on previously under-studied text versions.
The most significant achievements of the project to date are as follows:
1. In addition to creating digital raw text versions of 25 edited versions of 'Floire et Blancheflor', we have also transcribed 18 manuscript versions of 'Floire et Blancheflor' from all across Europe, making them accessible to a much wider readership.
2. In gathering scholarship on 'Floire et Blancheflor' from across a wide range of European, we have uncovered a total of 40 surviving text versions of the 'Floire et Blancheflor' narrative rather than the 26 that were already more widely known in the scholarly community.
3. We have created a comprehensive plot feature matrix for comparing over 4000 plot features shaping the different versions of 'Floire et Blancheflor'.
4. We have published an article on 'Fragments of Female Agency in the Early 'Floire et Blancheflor' Tradition' with the journal 'Philological Quarterly' and submitted a further two articles to medievalist journals that are currently under review.
