The Role of Consciousness in Semantic Processing: Insights from Metaphorical Priming

The question of whether we can access a word’s meaning outside of awareness has been broadly debated: despite numerous studies have claimed that words can be processed up to their semantic level even if presented subliminally (e.g., Sklar et al., 2012), a growing body of studies has questioned this finding (e.g., Rabagliati et al., 2018). A recent study challenged the claim that automatic semantic activation originates unconscious verbal priming, proposing instead that associative patterns between lexical forms can drive the effect (Bottini et al., 2016). Words are related to each other not only based on their meaning but also on their orthographic or phonological features, and the statistical relationships established through language use. Hence, the priming effect occasionally found might not be driven by semantic similarity, but instead by predictive relationships between wordforms appearing in similar linguistic contexts. Within this framework, we hypothesized that full consciousness is necessary to access a semantic representation, and that a priming effect can occur outside of awareness only when two words are linguistically related. Given the fact that, usually, words that are semantically related also co-occur in language use, we exploited spatial metaphors of numerical operations to disentangle these two conditions. The concepts of addition and subtraction can be conceptually represented along two spatial axes, a vertical and a lateral one, but only the vertical mapping is encoded linguistically. We obtained a 2x2x2 design from the combination of the between-participants factor of Prime Visibility (subliminal/supraliminal levels) and the within-participants factors of Congruity (congruent/incongruent) and Axis (lateral/vertical). Sixty participants perceived consciously spatial prime words, while other sixty participants were presented with the same words outside of awareness, through a masked priming technique. Prime words could be represented along the two spatial axes (up-down; left-right) and they could be congruent or incongruent with the target words, which referred to numerical operations (“to add”; “to diminish”) based on the metaphorical mapping between them. Participants performed a semantic decision task, in which they judged whether the target word denoted the concept of addition or subtraction, administered through a go-no go procedure; reaction times were recorded. Participants included in the subliminal group performed a visibility task to quantitatively ensure that they were not aware of the subliminal prime. Analyses were conducted only on “go” trials. By using linear mixed-effects regression models, we found a facilitation effect of congruent prime-target pairs on both the vertical and lateral axes in the supraliminal condition, but a significant priming effect only along the vertical axis in the subliminal condition. These results, along with a significant interaction between Congruity and Prime Visibility on the lateral axis, suggest that spatial prime words can activate related numerical target terms only if their metaphorical mapping is encoded in the language, and not only in our conceptual representation. The present work supports the hypothesis that challenges a truly unconscious semantic processing occurring outside of awareness, and claims that full consciousness is necessary to access a word’s meaning.