10.5281/zenodo.2546802
https://zenodo.org/records/2546802
oai:zenodo.org:2546802
Etienne P. van de Bijl
Etienne P. van de Bijl
Vrije Universiteit Amsterdam
Cody Kingham
Cody Kingham
Vrije Universiteit Amsterdam
Wido van Peursen
Wido van Peursen
Vrije Universiteit Amsterdam
Sandjai Bhulai
Sandjai Bhulai
Vrije Universiteit Amsterdam
A Probabilistic Approach to Syntactic Variation in Biblical Hebrew
Zenodo
2019
Hebrew bible, Markov Chains, Clustering, Statistics
2019-01-22
10.5281/zenodo.2546801
https://zenodo.org/communities/niaaproceedings1718
Creative Commons Attribution 4.0 International
There is currently a disagreement in the field of Hebrew Studies on the methods used to date individual books within the Hebrew Bible. An important question in this debate is if linguistic differences between the books are significant enough to warrant a diachronic explanation. In this project, we seek to answer whether differences in syntax between the books are large enough to merit groupings into Standard or Late Biblical Hebrew. We use a statistical tool called Markov Chains, which models transition dependency in sequences. Our method takes into account word and phrase order for parts of speech, phrase functions, and phrase types. We then cluster the books based on their statistical similarities. Our results may corroborate key claims of the diachronic approach