The grounding of concrete and abstract language: consolidated evidence, open issues, and new challenges.

Over the past few decades the idea that perception, action, and cognition are deeply entwined has gained increasing consensus within scholars across different research strands—almost making embodied perspectives on cognition the dominant scientific paradigm. One domain in which this relation has been extensively scritunised is the study of language. Converging evidence coming from behavioural as well as neurophysiological studies supports the claim that understanding language implies the recruitment of some of the same neural mechanisms underlying the capacity of interacting with the environment. In order to account for this strict relation, the notion of simulation has been proposed to underpin processes of language comprehension and acquisition. Actions, intrinsic, extrinsic, and relational properties of objects, direction as well as grammatical features have been found to be encoded in language and simulated when performing linguistic tasks. Recently, however, some of the findings bolstering EC approaches to cognition have failed to replicate in multi-lab cross-cultural studies. So, nowadays the question is not whether and how cognition is embodied, but rather when embodied aspects of cognition are necessarily engaged. A further issue faced by EC is constituted by abstract concepts representation and acquisition. Indeed, abstract concepts and words are more difficult to account for in a perspective strictly couched into sensorimotor and perceptual aspects, as they typically lack of a single bounded concrete referent to simulate. Over the years, different proposals have been put forward to unravel this problem—alternatively emphasizing the role of linguistic metaphors, situations and introspections, or emotions as scaffolding mechanisms for the acquisition and processing of abstract concepts. More recently, researchers within the EGC framework have started to acknowledge the role of both sensorimotor, perceptual aspects as well as linguistic distributional information for abstract concepts. Furthermore, recent lines of research are gradually abandoning the idea of a dichotomous distinction sharply opposing abstract and concrete concepts, favouring instead a multidimensional approach to conceptual knowledge. Two main novelties characterize these approaches. First, abstract concepts are not conceived as a single, monolithic categories; rather, scholars are beginning to study different subcategories of abstract concepts (e.g., numbers, emotions, social concepts). Second, both abstract and concrete concepts are understood as being grounded into perceptual and linguistic aspects to differing extents—also depending on the situation for which they are retrieved, or as a function of cultural and social constraints.