In this paper, we provide normative values for the action verb naming test used in Crepaldi et al. (2006, Neuropsychologia, 44, 73-89). This work makes the test, whose material is available upon request to the authors, effectively usable in day-to-day clinical activity.

In a simple referential game with two neural network-based agents, we analyze the object-symbol mapping trying to understand what kind of strategy was used to develop the emergent language. We see that, when the environment is uniformly distributed, the agents rely on a random subset of features to describe the objects. When we modify the objects making one feature non-uniformly distributed, the agents realize it is less informative and start to ignore it. Surprisingly, they make a better use of the remaining features. This interesting result suggests that more natural, less uniformly distributed environments might aid in spurring the emergence of better-behaved languages. [The paper is available on arXiv]

We developed a new set of modality exclusivity norms for 643 Italian concrete and abstract words, which we augmented with a number of ratings-based (e.g., familiarity, age of acquisition) and corpus-based covariates (e.g., frequency, contextual diversity). [Read more by clicking on the title.]