Multi-variate coding for possession: methodology and preliminary results

Area of Science:

• Linguistic typology
• Cross-linguistic semantics
• Computational linguistics

Background:

• Possession is a complex linguistic phenomenon with varied expressions across languages.
• Understanding differential possession requires analyzing noun classes and possessive constructions.
• Existing frameworks may not fully capture the independent yet linked nature of these elements.

Purpose of the Study:

• To develop a novel database structure for encoding differential noun possession.
• To investigate semantic, valence, and constructional dimensions of possession.
• To analyze cross-linguistic patterns in noun possession.

Main Methods:

• Database design for representing noun possession phenomena.
• Comparative analysis of noun possession across a global sample of 120 languages.
• Survey methodology to gather data on possessive constructions and noun classes.

Main Results:

• A database structure capable of encoding differential possession is presented.
• Preliminary survey results from 120 languages are discussed.
• A universal semantic core was identified for inalienable and non-possessible noun classes.

Conclusions:

• The proposed database structure effectively models differential possession.
• Inalienable nouns universally center on body parts and kinship.
• Non-possessible nouns universally center on animals, humans, and natural elements.