The objective of this paper is to clarify what types of embodied linguistic mechanisms are activated when we elaborate figurative speech about COLOUR/SEEING. Embodied theories of "negativity" and "positivity" biases have been posited in relation to language. According to behavioural and evolutionary studies a 'negativity bias' modifies the way humans react and process surrounding events, and has been tested in different realms of cognition, including corpus linguistics. The positivity bias - the Pollyanna hypothesis - affirms that humans tend to talk about the bright side of life. Good "positive" words are more prevalent, more meaningful, more diversely used, and more readily learned than "negative" words. I discuss possible conceptual underpinnings that explain the impact of positivity and negativity biases in processing visual figurativity in linguistic tasks.