Behavioral and Physiological Evidence Challenges the Automatic Acquisition of Evaluations

Abstract

Dual-learning theories of evaluations posit that evaluations can be automatically (i.e., efficiently, unconsciously, uncontrollably, and involuntarily) acquired. They also often assume the existence of evaluative-learning processes that are impervious to verbal information. In this article, we explain that recent research challenges both assertions for three categories of measures: explicit evaluative measures, implicit evaluative measures, and physiological measures of fear. In doing so, we also question the widespread assumption that implicit (i.e., typically behavioral and physiological) compared with explicit (i.e., self-reported) evaluative measures are indicative of the way evaluations are acquired. In the second part of the article, we discuss the practical implications of these recent findings.

Flavors of desire: Cognitive representations of appetitive stimuli and their motivational implications.

“…Importantly, we do not claim that our findings suggest that participants’ representations of appetitive stimuli are learned or accessed unconsciously, or that they influence behavior in a purely automatic manner (Corneille & Mertens, 2020; Hofmann et al, 2008). Indeed, because we asked participants to write down features of objects, our findings show that participants have conscious access to at least parts of their cognitive representations of appetitive objects.…”

Valence generalization across nonrecurring structures.

“…According to learning theory, if a pair of unfamiliar and relatively neutral symbols (call these CS1 and CS2) are respectively correlated with positively and negatively valenced US then CS1 should be positively evaluated relative to CS2, all else remaining equal (Staats & Staats, 1958). While the phenomenon of valence generalization following CS-US correlations has been investigated for some time (Mowrer, 2013; Staats, 1996), it remains unanswered whether the representational processes underlying transfer are exclusively propositional (De Houwer et al, 2021), or whether unqualified associations are necessary for describing the acquisition and expression of CS valence (Corneille & Mertens, 2020; Gawronski & Bodenhausen, 2006; Hofmann et al, 2010).…”

Evaluative Conditioning: Past, Present, and Future

“…Further, an EC effect without contingency awareness would suggest that the mental processes that underlie the EC effect do not require awareness, are unintentional, and are not easily controlled because people cannot intentionally cause or prevent the effect of CS-US co-occurrence if they are not aware of that co-occurrence. Recent reviews on the automaticity of the EC effect have concluded that there is no robust evidence for unawareness and uncontrollability of the encoding of the CS-US co-occurrence (Corneille & Mertens 2020. Therefore, we center here on evidence for (un)awareness and (un)controllability at the later stages of the processes that lead to EC effects.…”