Relational integrativity of prime-target pairs moderates congruity effects in evaluative priming

Abstract

In evaluative priming, positive or negative primes facilitate reactions to targets that share the same valence. While this effect is commonly explained as reflecting invariant structures in semantic long-term memory or in the sensorimotor system, the present research highlights the role of integrativity in evaluative priming. Integrativity refers to the ease of integrating two concepts into a new meaningful compound representation. In extended material tests using paired comparisons from two pools of positive and negative words, we show that evaluative congruity is highly correlated with integrativity. Therefore, in most priming studies, congruity and integrativity are strongly confounded. When both aspects are disentangled by manipulating congruity and integrativity orthogonally, three priming experiments show that evaluative-priming effects were confined to integrative prime-target pairs. No facilitation of prime-congruent targets was obtained for non-integrative stimuli. These findings are discussed from a broader perspective on priming conceived as flexible, context-dependent, and serving a generative adaptation function.

The Affective Meaning of Words is Constrained by the Conceptual Meaning

“…As the pronunciation task is a relatively pure index of semantic activation (Balota and Lorch 1986), these findings suggested that the affective meaning may be parasitic on the semantic meaning. The nonsignificant affective priming effect in the low associative strength condition was consistent with the findings of Ihmels et al (2016). They found that the processing facilitation of prime-congruent targets could only appear when the prime and target words could be integrated into a meaningful concept.…”

“…In Experiment 1, the absence of the affective priming effect in the ‘low associative strength’ condition may be due to the reduced activation of the corresponding semantic representations. However, the absence of the affective priming effect in the ‘low associative strength’ condition in Experiment 1 seemed to be inconsistent with the results of Ihmels et al (2016); they found a significant affective priming effect in a LDT task when the prime and target word pairs were associatively unrelated but not when they are related (Experiment 3). The inconsistency may be due to the methodological difference.…”

“…was not controlled, it was always the case that the semantic association between primes and targets was relatively strong (e.g., happy–bride), thus resulting in a significant affective priming effect (Ferré and Sánchez-Casas 2014; Hermans et al 2002). When the category (e.g., Storbeck and Robinson 2004), the semantic associative strength (e.g., the present study) or the relational integrativity (e.g., Ihmels et al 2016) was deliberately manipulated, the affective priming effect was obtained only when the prime and target word pairs were from the same semantic category or were strongly semantically associated or integrative. In Experiment 1, the absence of the affective priming effect in the ‘low associative strength’ condition may be due to the reduced activation of the corresponding semantic representations.…”

“…The inconsistency may be due to the methodological difference. In Ihmels et al’s (2016) study, the affective priming was obtained from the comparison of affective congruent versus incongruent trials, while in the current study, the affective priming effect was calculated by comparing the emotional congruent trials with the neutral-prime control trials. As for the absence of affective priming effect for associatively related prime-target pairs, Ihmels et al (2016) contributed it to the floor-effect, as the overall reaction times for the related trials were so fast (mean = 463 ms) after further speeding up by the associative context.…”

“…These findings indicated that the affective meaning may be restricted by the conceptual meaning of a word. Moreover, a recent study (Ihmels et al 2016) disentangled the emotional congruity and integrativity (which refers to the easy of integrating two concepts into a new meaningful compound representation), and found that evaluative-priming effects were confined to integrative prime–target pairs. In sum, although these studies concerning the relationship of the semantic and affective priming provided indirect evidences on the reciprocal relationship of the conceptual and affective meaning of a word, the findings were conflicting and could not provide a clear answer to the question.…”