When colours split from objects: The disconnection of colour perception from colour language and colour knowledge

Abstract

We investigated object-colour knowledge in RDS, a patient with impaired colour naming after a left occipito-temporal stroke. RDS's colour perception, object naming and verbal colour-knowledge (the ability to verbally say the typical colour of an object) were relatively spared. RDS was also able to state if an object was appropriately coloured or not. However, he could neither match colour names to coloured objects, nor match colour patches to greyscale objects. Thus, RDS's colour-naming deficit was associated with an impaired ability to conceptually relate visually presented object shapes and colours. These results suggest that objects in their typical colour are processed holistically in the visual modality, and that abilities important for colour naming may also be involved in abstracting colours from visual objects. We discuss these findings in the context of developmental psychology and linguistic anthropology, and propose a model of neuro-functional organization of object-colour knowledge.

Color Space Geometry Uncovered with Magnetoencephalography

“…The results presented here are consistent with the idea that the brain processes information in a directional way, 57 with perceptual representations capable of driving semantic or cognitive representations, but not the reverse-semantic representations seem incapable of instantiating perceptual representations. 58 The plots of cross-temporal generalization of spiral color (Figure 5) show a pattern of bands parallel to the identity diagonal, which provides a rare example (the first, to our knowledge) of a theoretical possibility-the reactivation pattern of King and Dehaene. 15 This pattern suggests that the perceptual representation of color is encoded by a unique sequence of brain states that is reactivated with a slight delay.…”

Towards Strong Inference in Research on Embodiment – Possibilities and Limitations of Causal Paradigms

“…What all of the impaired tasks have in common is that it is necessary to abstract color information from objects. Thus, for fully-fledged color knowledge perceptual representations might not be enough and language (i.e., intact connectivity pathways with anterior-temporal regions) may be required for an abstract representation of color that has been suggested to otherwise be bound to holistic object representations in the visual system (Siuda-Krzywicka et al, 2019b). This very interesting result suggests that certain levels of knowledge cannot be based solely onto sensorimotor simulations, and potentially opens a new branch on our logical tree, asking what exactly constitutes the non-simulation part of concepts (Figure 4).…”

What Cognitive Neurology Teaches Us about Our Experience of Color

“…In another study on the same patient, we explored the relation between color naming and color knowledge (Siuda-Krzywicka and others 2019c). RDS could state whether an object was appropriately colored or not, but could not choose the color patch appropriate to a greyscale object.…”