The problem of switching in conditional behavior.

Abstract

This article reviews studies of various authors on the phenomenon of "switching," which is observed in both classical and instrumental conditioning and consists in elicitation of different responses to the same conditional stimulus (CS) when it is applied in an environment different than the original one. The different responses include a decrease or an absence of the previously trained conditional response (CR), elicitation of an appetitive response instead a defensive one, or vice versa, as well as elicitation of two different instrumental CRs in the same trial. The studies suggest that, due to the repeated occurrence of CS in the same environment (E), also called "situation" or "context," associations are formed between CS and E. Consequently, the CR is elicited to a compound CS+E rather than to CS alone. When the CS is applied alone in a different E than the original one, the previously formed associations are inactive and the CR cannot be elicited; this leads to switching. Studies also suggest that E plays a dominant role in conditioning compared with that of CS alone, which often appears to be only a trigger for eliciting the response. However, CS tested in a different E may still produce some components of the previously acquired CR, such as a general fear behavior to an originally defensive CS or an approach behavior to an originally alimentary CS. The environmental stimuli can be considered the "determining" stimuli that determine the kind of reaction to be elicited, or "tonic" stimuli that increase the tonus in the brain but do not elicit the CR. The "determining" or "tonic" stimuli do not seem to be a special class of stimuli. Instead, they are stimuli that initially can produce the CR (e.g., intertrial CRs), but by being not reinforced they become partly inhibited; nevertheless, due to associations with the reinforcement, they still can produce some excitement related to it, thus facilitating the CR.