Abstract

Omission and Overexpectation are two paradigms that lead to a reduction in the associative relationship between events. In Omission, this is achieved through absence of the delivery of an expected outcome. In Overexpectation, two individually trained cues are presented together and are met with an outcome of less intensity then the sum of the two expected outcomes. The reduction in the conditioned response following Omission or Overexpectation training reflects a reduction in the associative relationship between the target cue and the conditioned response. However, this reduction is easily disrupted. Testing outside of the extinction context, letting time pass, or giving unsignalled exposure to the unconditioned stimulus, each lead to an increase in the conditioned response from extinction levels. These effects can be explained by a view that the original association is not erased but rather is inhibited by extinction learning and this extinction learning is more vulnerable to disruption. Insight in to this disruption might come from a biological perspective where Omission and Overexpectation have been tracked by unique and overlapping neuronal populations in the central nucleus of the amygdala (Iordanova et al., 2016). Given evidence that unique and overlapping neuronal populations are involved, one might expect the influence of two these processes to be behaviourally additive. This raises the possibility that targeting the unique and common populations associated with Omission and Overexpectation through combining both forms of extinction training might lead to lasting reduction in behaviour as one that is more resistance to relapse. To test this prediction, following conditioning of two different cues with a positive reinforcement, Sprague Dawley rats (Rattus norvegicus) were assigned to either a constantly reinforced Control, an Overexpectation-only, Omission-only, or a Mixed group consisting of blocks of Overexpectation sessions followed by blocks of Omission sessions. Subsequent Renewal, and Reinstatement tests show higher responding in the Control group compared to the three groups that had undergone extinction or Overexpectation training, with the Mixed extinction group showing the lowest rate of response. This effect was consistent across different dependant variables. However a Spontaneous Recovery test, though yielding a similar decrease, did not result in a substantive effect-size statistic or a statistically significant null hypothesis test. Taken together these findings have potential implications for deepening reductions of maladaptive associations on behaviour, such as those present in addiction and anxiety disorders.