Augmented Modality Exclusivity Norms for Concrete and Abstract Italian Property Words.

Piermatteo Morucci,Davide Crepaldi,Roberto Bottini

doi:10.5334/joc.88

Abstract

How perceptual information is encoded into language and conceptual knowledge is a debated topic in cognitive (neuro)science. We present modality norms for 643 Italian adjectives, which referred to one of the five perceptual modalities or were abstract. Overall, words were rated as mostly connected to the visual modality and least connected to the olfactory and gustatory modality. We found that words associated to visual and auditory experience were more unimodal compared to words associated to other sensory modalities. A principal components analysis highlighted a strong coupling between gustatory and olfactory information in word meaning, and the tendency of words referring to tactile experience to also include information from the visual dimension. words were found to encode only marginal perceptual information, mostly from visual and auditory experience. The modality norms were augmented with corpus–based (e.g., Zipf Frequency, Orthographic Levenshtein Distance 20) and ratings–based psycholinguistic variables (Age of Acquisition, Familiarity, Contextual Availability). Split-half correlations performed for each experimental variable and comparisons with similar databases confirmed that our norms are highly reliable. This database thus provides a new important tool for investigating the interplay between language, perception and cognition.

Highlights

The meaning of words has been suggested to be grounded, at least partially, in the perceptual and motor system (Meteyard et al, 2012, Barsalou et al, 2008). This idea, usually referred to as embodied semantics (Barsalou, 1999; Glenberg & Gallese, 2012) has found support from many neuroimaging studies showing that processing word meaning involves the recruitment of modality-specific networks distributed across the cortex (Barsalou, 2008; Binder & Desai, 2011; Vigliocco et al 2009)
Words referring to concepts that can be experienced through different senses (e.g., “football”) activate multiple modality-specific networks
In order to study the link between sensorimotor experience and conceptual processing it is important to have tools that allow a precise characterization of the sensorimotor characteristics usually associated with a given word/concept

Summary

Introduction

The meaning of words has been suggested to be grounded, at least partially, in the perceptual and motor system (Meteyard et al, 2012, Barsalou et al, 2008). This idea, usually referred to as embodied semantics (Barsalou, 1999; Glenberg & Gallese, 2012) has found support from many neuroimaging studies showing that processing word meaning involves the recruitment of modality-specific networks distributed across the cortex (Barsalou, 2008; Binder & Desai, 2011; Vigliocco et al 2009). Words referring to concepts that can be experienced through different senses (e.g., “football”) activate multiple modality-specific networks (i.e., visual and action–related; Van Dam et al, 2012). Is a word heavily based on perceptual information? Is it more strongly associated to vision or touch? Is it unimodal (e.g., only visual: “red”) or multimodal (e.g., visuohaptic: “round”)?

Methods

Results

Conclusion