A standardized approach to measuring gender transparency in languages

Area of Science:

• Linguistics
• Psycholinguistics
• Computational Linguistics

Background:

• Grammatical gender systems in languages often rely on ortho-phonological regularities for noun classification.
• Gender transparency, defined by consistent noun-ending-gender associations, is crucial for language acquisition and processing.
• Existing classifications of gender transparency lack systematicity, particularly regarding complex cues and opaque nouns.

Purpose of the Study:

• To propose a systematic methodology for assessing grammatical gender transparency in languages.
• To develop an objective, quantitative measure of gender transparency.
• To investigate the role of derivational suffixes and the definition of opaque nouns in gender transparency.

Main Methods:

• Adapted Audring's (2019) standardized factors for analyzing gender system complexity.
• Integrated factors with literature on gender acquisition and processing to define transparency metrics.
• Instantiated the methodology into a quantitative model, tested using European Portuguese and Dutch.

Main Results:

• European Portuguese demonstrated a high degree of gender transparency.
• Dutch exhibited a moderately low degree of gender transparency.
• The results align with theoretical expectations, confirming the model's validity.

Conclusions:

• The proposed quantitative model provides an objective measure for grammatical gender transparency.
• This framework can help resolve inconsistencies in language classification regarding transparency.
• Future applications to other languages will enhance predictions in gender processing and acquisition studies.