The Effect of Valence and Arousal on Spatial Attention

Conceptual metaphor theory suggests that to mentally represent abstract concepts we use metaphorical associations to map them onto more concrete constructs (Lakoff & Johnson, 1980; 1999). Using a choice reaction time (CRT) task, Meier and Robinson (2004) found that positively valenced words primed attention to higher areas in vertical space, while negatively valenced words primed attention to lower areas in vertical space, consistent with the good is up/bad is down conceptual metaphor. Meier and Robinson (2004; 2005) suggest this provides evidence that emotional words create an automatic and obligatory metaphor-congruent shift in spatial attention, driven by a Valence x Position interaction. However, other research shows that the arousal level, not just the valence of emotional words can affect reaction time (Robinson, Storbeck, Meier & Kirkeby, 2004). This means concluding that valence alone is driving the shift in attention is premature. Furthermore, Brookshire, Irvy and Casasanto (2010) dispute Meier and Robinson’s claim that the relationship between affect and metaphor is automatic, instead suggesting that affect-metaphor associations are optional and only accessed under certain contextual conditions. The purpose of this thesis was therefore two-fold. First, it aimed to explore whether valence, arousal, or an interaction between the two was responsible for driving the metaphor-consistent shift in spatial attention observed by Meier and Robinson (2004). Second, it aimed to progress the discussion about when affective stimuli, in the form of emotional words, automatically activate the good is up/bad is down conceptual metaphor. Three CRT experiments were conducted in which (a) emotional stimulus words were differentiated by arousal level as well as valence and (b) the evaluation of stimulus words’ affective tone (pleasant/unpleasant) was manipulated. A Valence x Position interaction in relation to the good is up but not bad is down conceptual metaphor was found when the valence of priming words was evaluated, suggesting valence, rather than arousal or a combination of both, is driving the affect-metaphor relationship. No evidence for the automatic activation of affect-metaphor associations was found when the word’s affective tone was not evaluated. These findings suggest that while driven by valence, affect-metaphor associations are not fully automatic and occur only under certain contextual conditions. The implications for our understanding of how emotion impacts spatial attention are discussed, suggesting metaphors enrich, rather than monopolise our mental representation of abstract, affective concepts.