Automated identification of borrowings in multilingual wordlists

Area of Science:

• Computational linguistics
• Historical linguistics
• Language evolution

Background:

• Lexical borrowing is key to language evolution.
• Automated identification of borrowings in lexical data is underdeveloped.
• Existing methods do not identify cross-linguistic borrowings.

Purpose of the Study:

• To propose a novel automated method for identifying lexical borrowings.
• To address the limitation of existing methods in cross-linguistic analysis.
• To facilitate computer-assisted studies on lexical borrowing in language contact areas.

Main Methods:

• Development of a new computational method for lexical borrowing identification.
• Compilation of a large comparative dataset of 48 South-East Asian languages.
• Testing the proposed method on the compiled dataset.

Main Results:

• The proposed method demonstrates very promising results in identifying lexical borrowings.
• The method is conceptually straightforward and easy to apply.
• Successful application on a large, multi-lingual dataset.

Conclusions:

• The new method effectively automates lexical borrowing identification across multiple languages.
• The approach is practical for computer-assisted exploratory studies.
• It advances research in understanding language contact and evolution in South-East Asia.