From semantic knowledge to semantic features: French semantic feature production norms for 360 concepts.

This article presents a new set of semantic feature production norms, collected from 580 young adults, for 360 French concepts across various semantic categories. Although empirically derived feature norms have been developed for several languages and have been shown to be useful for investigating semantic memory and providing assessment tools, none are currently available for native French-speaking populations. In this study, the participants performed a property generation task in which they were asked to list features to describe the characteristics of each given concept (e.g., for apple: <is a fruit>, <can be red>, <has seeds>). After having conducted a unification procedure of the participants' productions, we included a total of 4586 semantic features in the norms. We then computed various semantic indexes, ranging from semantic richness variables (e.g., number of features, semantic neighborhood density) to feature informativeness variables (e.g., relevance, cue validity, distinctiveness). In the end, the present semantic feature production norms are organized into four different openly available datasets, each describing a set of semantic measures: (1) concept-feature measures, (2) concept measures, (3) concept-concept matrix, and (4) feature-feature matrix. These semantic feature norms for French concepts should be a valuable tool for researchers interested in investigating normal and impaired semantic cognition.