Specificity ratings for Italian data

Area of Science:

• Cognitive Psychology
• Psycholinguistics
• Computational Linguistics

Background:

• Language abstraction relies on word specificity, yet human ratings are scarce.
• Existing computational metrics may not capture subjective word specificity.
• Understanding specificity is key to abstraction and conceptual representation.

Purpose of the Study:

• To create a dataset of human-annotated word specificity for Italian words.
• To compare human specificity ratings with computational metrics (WordNet).
• To investigate the relationship between word specificity, concreteness, and lexical decision-making.

Main Methods:

• Collected specificity ratings from native Italian speakers using Best-Worst Scaling.
• Correlated human ratings with WordNet-derived specificity metrics.
• Correlated specificity ratings with concreteness ratings.
• Used regression analyses to assess specificity's impact on lexical decision latencies.

Main Results:

• Human specificity ratings showed low correlation with WordNet metrics.
• Specificity ratings were distinct from concreteness ratings.
• Word specificity uniquely predicted variance in lexical decision latencies.

Conclusions:

• Human perception of word specificity is not fully captured by current computational tools like WordNet.
• Specificity and concreteness are separate dimensions of word meaning relevant to abstraction.
• Word specificity is a significant factor in cognitive processing and lexical access.