The Cognate Data Bottleneck in Language Phylogenetics
🤖 AI Summary
A long-standing bottleneck in linguistic phylogenetics is the scarcity of large-scale, high-quality cognate data. Method: This study systematically evaluates the feasibility of automatically constructing large cognate datasets from multilingual lexical knowledge bases—specifically BabelNet—by designing a pipeline that maps synset definitions to character-state matrices via feature extraction and matrix encoding, then applies computational phylogenetic methods (e.g., maximum likelihood) and machine learning–assisted analysis for genealogical inference. Results: Phylogenies inferred from the automatically generated matrices exhibit significant topological discordance with established language family trees, demonstrating that current general-purpose semantic resources are insufficient for reliable, phylogenetically valid automated cognate identification. This work provides the first quantitative evidence of structural limitations in semantic knowledge bases for language evolution modeling, offering critical methodological warnings and empirical grounding for developing domain-specific cognate resources and interpretable, automated annotation frameworks.
Application Category
📝 Abstract
To fully exploit the potential of computational phylogenetic methods for cognate data one needs to leverage specific (complex) models an machine learning-based techniques. However, both approaches require datasets that are substantially larger than the manually collected cognate data currently available. To the best of our knowledge, there exists no feasible approach to automatically generate larger cognate datasets. We substantiate this claim by automatically extracting datasets from BabelNet, a large multilingual encyclopedic dictionary. We demonstrate that phylogenetic inferences on the respective character matrices yield trees that are largely inconsistent with the established gold standard ground truth trees. We also discuss why we consider it as being unlikely to be able to extract more suitable character matrices from other multilingual resources. Phylogenetic data analysis approaches that require larger datasets can therefore not be applied to cognate data. Thus, it remains an open question how, and if these computational approaches can be applied in historical linguistics.
Problem
Research questions and friction points this paper is trying to address.
Lack of large datasets for cognate-based phylogenetics
Infeasibility of automatic cognate dataset generation
Inconsistency of extracted data with gold standard trees
Innovation
Methods, ideas, or system contributions that make the work stand out.
Leverage complex models and machine learning
Automatically extract datasets from BabelNet
Analyze phylogenetic data inconsistencies
🔎 Similar Papers
No similar papers found.
