Academic journal article Psychonomic Bulletin & Review

Is Stimulus Competition an Acquisition Deficit or a Performance Deficit?

Academic journal article Psychonomic Bulletin & Review

Is Stimulus Competition an Acquisition Deficit or a Performance Deficit?

Article excerpt

Traditionally, blocking (X-outcome, followed by XY-outcome, resulting in attenuated conditioned responding to Y, relative to XY-outcome alone) has been explained in terms of the X-outcome association's preventing the acquisition of the Y-outcome association. This view is challenged by models that view stimulus competition as a deficit in the expression of the acquired Y-outcome association. Here, we provide evidence that blocking is a performance deficit in which the Y-outcome association, the to-be-blocked stimulus, can affect behavioral control by the blocking stimulus (i.e., attenuate responding to X). The results are discussed in terms of acquisition and performance models of stimulus competition.

Early psychological theories of learning (e.g., Bush & Mosteller, 1951) assumed that spatiotemporal contiguity was both necessary and sufficient for learning to occur. That is, when two events were in a close temporal and spatial relationship, an association was presumably established between them, so that the presentation of one of the events would retrieve a representation of the other event. However, this view was challenged in the late 1960s by the analysis of stimulus competition. A frequently cited example of stimulus competition is the blocking effect (e.g., Kamin, 1968). Kamin observed that conditioned responding to a stimulus (Y) trained as a predictor of an outcome (e.g., an unconditioned stimulus [US]) was diminished (blocked) if this stimulus was presented during training in compound with another stimulus (X) that had already been established as a good predictor of the occurrence of the impeding outcome (i.e., X-US pairings, then XY-US pairings). The capacity to explain blocking and other stimulus competition effects separated early views of learning (e.g., Bush & Mosteller, 1951) from a newer generation of associative models (e.g., Mackintosh, 1975; Pearce, 1987; Rescorla & Wagner, 1972; Wagner, 1981).

These newer models of learning explain blocking on the basis of the assumption that redundant information is not encoded. Once an organism has learned that a stimulus is a reliable predictor of an outcome, this stimulus will impede or block learning about other stimuli presented together with it as predictors of the same event. This framework, which we will hereafter call the informational hypothesis, challenged the assumption that contiguity was sufficient for learning to occur. According to the informational hypothesis, learning occurs only when one event nonredundantly predicts the occurrence of another. Thus, blocking reflects a deficit in the acquisition of an association between Y and the US. Other stimulus competition effects (e.g., overshadowing [Pavlov, 1927], the US-preexposure effect [Randich & LoLordo, 1979], and the relative stimulus validity effect [Wagner, Logan, Haberlandt, & Price, 1968]) are explained in a similar way by these associative theories.

The informational hypothesis view has been challenged since the mid-1980s by the comparator hypothesis (e.g., Denniston, Savastano, & Miller, 2001; Miller & Matzel, 1988). The comparator hypothesis differentiates between the processes of acquisition and production of a response. The model assumes that associations between stimuli are learned through contiguity and that, at each opportunity to respond, associations compete with each other to determine whether the stimulus will elicit a response. That is, contiguity is considered sufficient for learning to occur, but responding in the presence of a stimulus depends on a comparison at the time of each test trial between (1) the strength of the association between the test stimulus and the outcome and (2) the associative strength between the other (discrete or contextual) stimuli that were trained together with the test stimulus and the outcome. For example, in a blocking procedure (i.e., X-US followed by XY-US, testing on Y), the previously trained stimulus (X) does not prevent the acquisition of an association between Y and the US; that is, the Y-US association is acquired through contiguity and independently of the effects of X. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.