The Indo-European Cognate Relationships dataset

Area of Science:

• Computational Linguistics
• Historical Linguistics
• Language Evolution Studies

Background:

• The Indo-European language family is vast, with complex historical relationships.
• Understanding cognate patterns is crucial for reconstructing proto-languages and tracing linguistic evolution.
• Existing datasets often lack comprehensive coverage or standardized formats for computational analysis.

Purpose of the Study:

• To introduce the Indo-European Cognate Relationships (IE-CoR) dataset, a novel resource for linguistic research.
• To provide a benchmark dataset for computational studies on Indo-European language evolution.
• To facilitate the study of lexical inheritance and horizontal transfer across Indo-European languages.

Main Methods:

• Compilation of a relational dataset covering 160 Indo-European languages.
• Analysis of 25,731 lexeme entries into 4,981 cognate sets based on 170 reference meanings.
• Inclusion of data on horizontal transfer, time calibration, and geographical/social metadata.
• Adherence to Cross-Linguistic Data Format (CLDF) protocols for interoperability.

Main Results:

• The IE-CoR dataset contains 25,731 lexeme entries across 160 languages, organized into 4,981 cognate sets.
• It represents all 13 main Indo-European clades and includes data on horizontal word transfer.
• The dataset incorporates time calibration, geographical, and social metadata for enhanced analysis.

Conclusions:

• The IE-CoR dataset offers a comprehensive, standardized resource for computational historical linguistics.
• It serves as a valuable benchmark for research into Indo-European language evolution.
• The dataset's design promotes interoperability and can serve as a model for other language families.