Flipbook

5 SAFETY BEHAVIOR AFTER EXTINCTION TRIGGERS A RETURN OF THREAT EXPECTANCY 135 for the CS- at the end of the Extinction phase. In the current study, participants who showed insufficient fear extinction were excluded. Although this does not eliminate the possibility that there was residual threat expectancy, at the end of the Extinction phase threat expectancy was low (see Figure 1), and comparable for A, B, and C. We therefore consider it unlikely that residual threat expectancy explains the current findings. Second, the availability of safety behavior may have signaled threat and activated alarm mechanisms (Sloan & Telch, 2002). People engage in safety behavior to avert a feared outcome, and therefore safety behavior is meaningfully linked to perceived threat (Salkovskis, Clark, & Gelder, 1996). In this study, participants learned to use safety behavior to prevent the expected US in the Safety behavior acquisition phase, and thereby safety behavior became meaningfully linked to the threat. Hence, the availability of safety behavior may have increased threat expectancy, and motivated safety behavior use after fear extinction. Third, participants may have inferred threat from their use of safety behavior. Gangemi, Mancini, and van den Hout (2012), and van den Hout et al. (2014; 2017) found that individuals may infer threat from safety behavior, “I avoid, so there must be danger” (akin to ‘emotional reasoning’; see Engelhard & Arntz, 2005). Although ‘behavior as information’ effects are mainly found in anxious individuals and not in healthy control participants, using safety behavior in an objectively safe situation may have induced cognitive dissonance. Participants may have reduced this dissonance by bringing their threat expectancy in line with their behavior (Festinger & Carlsmith, 1959). Fourth, participants may have perceived different contexts within the experimental task based on CS - US contingencies and safety behavior availability and unavailability, which may explain the findings (see Bouton, 2002; 2016; Vervliet et al., 2013). In the Pavlovian acquisition phase and the Safety behavior acquisition phase, stimulus A and C were always followed by the US, and when safety behavior was available it was effective in preventing the US. Participants may have perceived