Stimulus competition between a discrete cue and a training context: Cue competition does not result from the division of a limited resource

Abstract

Several associative learning theories explain cue competition as resulting from the division of a limited resource among competing cues. This leads to an assumption that behavioral control by 2 cues competing with each other should always reflect a tradeoff, resulting in apparent conservation of total reinforcer value across all competing cues. This assumption was tested in 3 conditioned lick suppression experiments with rats, investigating the effects of changing the conditioned stimulus (CS) duration (Experiment 1), administering pretraining exposures to the CS (Experiment 2), and presenting nonreinforced CSs during the intertrial interval (Experiment 3) on Pavlovian conditioned responding to both the CS and the conditioning context. Fear conditioned to the context and to the CS decreased when the CS was of longer duration, massively preexposed before being paired with the reinforcer, or presented alone during the intertrial interval. These observations are problematic for the theories that explain cue competition as the division of a limited resource and suggest that the total reinforcer value across competing cues is not always fixed for a given reinforcer.

Figure 1

The original comparator hypothesis. The three long arrows represent three hypothesized associations, the target cue–outcome association (Link 1), the target cue–comparator stimulus within-compound association (Link 2), and the comparator stimulus–outcome association (Link 3). Presentation of the target cue activates two representations of the outcome, the representation directly activated by the target cue through Link 1 and that indirectly activated through the comparator stimulus (through Links 2 and 3). The strengths of the outcome representation directly and indirectly activated are compared, and the strength of responding to the target cue is determined. Rectangles depict physical stimulus events and ovals depict representations of events not presented but activated by other stimuli. US = unconditioned stimulus.

Figure 2

Results of Experiment 1. Black bars depict mean log latencies to complete the first cumulative 5 s of drinking in the context in which conditioning treatment occurred. Open bars depict mean log latencies to complete the first cumulative 5 s of drinking in the presence of Stimulus X in a neutral context. Error bars represent the standard errors of the mean. CS = conditioned stimulus; US = unconditioned stimulus.

Figure 3

Results of Experiment 2. Black bars depict mean log latencies to complete first cumulative 5 s of drinking in the context in which conditioning treatment occurred. Open bars depict mean log latencies to complete the first cumulative 5 s of drinking in the presence of Stimulus X in a neutral context. Error bars represent the standard errors of the mean. CS = conditioned stimulus; US = unconditioned stimulus; signal-pre = signal-preexposure.

Figure 4

Results of Experiment 3. The left panel depicts mean log latencies to complete the first cumulative 5 s of drinking in the context in which conditioning treatment occurred. The right panel depicts mean log latencies to complete the first cumulative 5 s of drinking in the presence of Stimulus X in a neutral context. Open bars represent the results of groups that did not receive extinction treatment of Stimulus X; black bars represent those of groups that received extinction treatment of Stimulus X after the conditioning phase. Error bars represent the standard errors of the mean. CS = conditioned stimulus; US = unconditioned stimulus; Ext = extinction.