Rule-based processes in generalisation and peak shift in human fear conditioning

Two experiments explored the role of verbalisable rules in generalisation of human differential fear conditioning with electric shock as the aversive stimulus. Two circles of different sizes served as conditioned stimuli (CS+ and CS–), before testing with a range of circle sizes. In Experiment 1, shock expectancy ratings followed a peak-shifted unimodal gradient, with maximum ratings at a test value further along the dimension from CS+ in the opposite direction to CS–. However, differentiable gradients were observed when participants were divided on the basis of the rules they reported using during the task (linear and similarity). Experiment 2 was designed to counter the contradictory feedback arising from extinction testing by removing the shock electrodes during the test phase. A more linear overall gradient was observed, and sub-groups defined by self-reported rules showed distinct gradients that were congruent with their rules. These results indicate that rule-based processes are influential in generalisation of conditioned fear along simple stimulus dimensions, and may help explain generalisation phenomena that have traditionally been attributed to automatic, similarity-based processes.

Sexually divergent expression of active and passive conditioned fear responses in rats

Traditional rodent models of Pavlovian fear conditioning assess the strength of learning by quantifying freezing responses. However, sole reliance on this measure includes the de facto assumption that any locomotor activity reflects an absence of fear. Consequently, alternative expressions of associative learning are rarely considered. Here we identify a novel, active fear response (‘darting’) that occurs primarily in female rats. In females, darting exhibits the characteristics of a learned fear behavior, appearing during the CS period as conditioning proceeds and disappearing from the CS period during extinction. This finding motivates a reinterpretation of rodent fear conditioning studies, particularly in females, and it suggests that conditioned fear behavior is more diverse than previously appreciated. Moreover, rats that darted during initial fear conditioning exhibited lower freezing during the second day of extinction testing, suggesting that females employ distinct and adaptive fear response strategies that improve long-term outcomes.

How training and testing histories affect generalization: a test of simple neural networks

We show that a simple network model of associative learning can reproduce three findings that arise from particular training and testing procedures in generalization experiments: the effect of (i) ‘errorless learning’, (ii) extinction testing on peak shift, and (iii) the central tendency effect. These findings provide a true test of the network model which was developed to account for other phenomena, and highlight the potential of neural networks to study the phenomena that depend on sequences of experiences with many stimuli. Our results suggest that at least some such phenomena, e.g. stimulus range effects, may derive from basic mechanisms of associative memory rather than from more complex memory processes.

Stimulus Transitions in Extinguishing a Discrimination

The number and order of stimulus transitions while extinguishing a discrimination were evaluated in order to determine what stimulus changes are essential to initiate differential responding. Following Light on vs Light off food-reinforced discrimination training in operant chambers, 40 rats were randomly assigned to five groups for extinction testing in the presence of S+ or S−. For three groups the first transition during extinction was from S+ to S− whereas for two groups the first transition was from S− to S +. There were basically no differences in responding to the external stimuli among the three S+ to S− groups; the two S− to S+ groups differed from the other three groups and from each other. It was concluded that an S+ to S− transition is essential in extinction for the appearance of differential responding.