Abstract
Adjectives are essential in how people describe, evaluate, and reason about others. They differ along meaningful semantic dimensions such as desirability (e.g., "friendly" is more positive than "rude") and breadth (e.g., "punctual" is narrower than "reliable"). Adjectival breadth has received limited empirical attention, partly because existing resources are sparse and outdated. We introduce a new database with subjective ratings from approximately 1,500 Americans for 1,214 adjectives on both breadth and desirability. Unlike existing resources, this updated database is more comprehensive and diverse, allowing for detailed analysis of adjectival use in academic and applied contexts. We validate this database with a large-scale analysis of online product reviews, showing how variation in adjective breadth is a common feature of natural language use. This database should prove valuable for research on semantic representation, social inference, and evaluative communication across various fields.