The Visual Search Analogue of Latent Inhibition: Implications for Theories of Irrelevant Stimulus Processing in Normal and Schizophrenic Groups
Total Page:16
File Type:pdf, Size:1020Kb
Psychonomic Bulletin & Review 2005, 12 (2), 224-243 The visual search analogue of latent inhibition: Implications for theories of irrelevant stimulus processing in normal and schizophrenic groups R. E. LUBOW Tel Aviv University, Ramat Aviv, Israel and OREN KAPLAN College of Management, Rishon Lezion, Israel Latent inhibition (LI) is a robust phenomenon that is demonstrated when a previously inconsequen- tial stimulus is less effective in a new learning situation than a novel stimulus. Despite LI’s simplicity, there is considerable disagreement as to its theoretical basis. Attentional theories claim that unat- tended stimulus preexposures reduce stimulus associability. Alternatively, it has been asserted that as- sociability is unaffected and that LI is a result of competition/retrieval processes. The present article reviews a series of visual search studies, some with normal subjects, both undifferentiated and divided into low and high schizotypals, and others with pathologies that entail dysfunctional attention, such as schizophrenia, Parkinson’s disease, and anxiety. The visual search conditions were designed to model those of traditional LI experiments, while tapping attentional processes independently of the learning scores that index LI. A variety of evidence from these and other studies is used to support the involve- ment of attentional and retrieval processes in LI. A model of the mechanism of action of these pro- cesses in LI is presented, together with its application to schizophrenia. There are some conditions that make it more difficult to tensive documentation indicates that such stimulus pre- acquire or to express a new association to a familiar stim- exposure may also enhance performance on a learning ulus than one to a novel stimulus, particularly when the fa- task, as in perceptual-learning effects (for reviews, see miliar stimulus has been previously unattended. This situ- Gibson, 1969; Hall, 1991). Whether one obtains poorer ation is expressed in an apparently simple phenomenon, or better learning following stimulus preexposure de- called latent inhibition (LI). LI occurs across a variety of pends on the conditions of stimulus preexposure, the re- species, including pigeons, rats, and humans, and in many lationship between the preexposure and the test contexts, different testing paradigms, from simple classical condi- including the time between them and where that time is tioning to rule learning (for a review, see Lubow, 1989). spent, and the choice of a comparison (control) group. Given the generality and robustness of this stimulus pre- Briefly, LI in humans is favored when (1) the preexposed exposure effect, it is not surprising to find that LI has a stimulus is not relevant to (is unattended in) the preex- significant place in the history of psychological research, posure task (masking task) but is relevant to the test task. beginning with issues in classical learning theory (for (2) the preexposure and test contexts are the same, and reviews, see Hall, 1991; Lubow, 1989) and extending (3) the stimulus-preexposed (PE) group is compared through the study of the processing of irrelevant informa- with a group that is not preexposed (NPE) to the to-be- tion associated with high schizotypality and schizophrenia target stimulus but has spent the same amount of time in (for reviews, see Braunstein-Bercovitz, Rammsayer, Gib- the preexposure context. In opposition to LI, perceptual bons, & Lubow, 2002; Escobar, Oberling, & Miller, 2002; learning is promoted when (1) the preexposed stimulus is Lubow, 2005; Lubow & Gewirtz, 1995) and even of cre- relevant to (attended in) the preexposure task and con- ativity (Carson, Peterson, & Higgins, 2003). tinues to be relevant in the test task, (2) the PE stimulus Despite the overwhelming evidence that preexposure is presented in one context but the test is conducted in a to a stimulus may have a negative impact on the demon- different context, and (3) the stimulus PE group is com- stration of subsequent learning with that stimulus, ex- pared with a group that is tested with a novel stimulus in a novel context (NOV group). In short, stimulus preexposure may produce different Preparation of this article was supported by a grant from the Israel effects, some of which depend on the conditions of preex- Science Foundation, funded by the Israel Academy of Sciences and Hu- manities. Correspondence concerning this article should be addressed to posure and others on those of test. Nor do such preexpo- R. E. Lubow, Department of Psychology, Tel Aviv University, Ramat Aviv sures have consequences that are limited to impairment 69978, Israel (e-mail: [email protected]). or enhancement of performance on tests of learning. For Copyright 2005 Psychonomic Society, Inc. 224 LATENT INHIBITION AND VISUAL SEARCH 225 example, a novel item among familiar (preexposed) dis- albeit circularly, by the LI effect itself (for exceptions, see tractors can be localized more quickly than a novel item Braunstein-Bercovitz, Hen, & Lubow, 2004; Braunstein- among novel distractors (novel pop-out [NPO] effect; Bercovitz & Lubow, 1998a). However, this situation can see, e.g., Johnston & Hawley, 1994). Other examples in- be corrected, if one can demonstrate that the same set of clude the once popular stimulus familiarization effect stimulus PE conditions that produces degraded perfor- (Cantor, 1969), in which simple stimulus preexposure, mance on a learning task (traditional LI) also leads to as compared with no such preexposure, produces slower relatively poor performance on a task that assesses at- responding in a choice response time (RT) task. Indeed, tention but is independent of the learning-based LI tests. any two-phase experimental design can be construed as To this end, we have used visual search tasks that do involving preexposure and test stages—as for example, not engage memory retrieval processes. The target is de- studies of extinction—in which a paired conditioned fined as a unique figure among a group of similar dis- stimulus (CS) and unconditioned stimulus (US) are “pre- tractors, rather than as a function of historical experien- exposed” in the first phase and, in the second, test phase, tial factors. By manipulating preexposure variables that the CS is presented by itself. Among the various animal- normally produce LI and assessing performance on an learning paradigms, blocking, overshadowing, sensory attentionally modulated visual search task, one can at- preconditioning, and many others also conform to a tain a better comprehension of the processes governing preexposure-and-test design. LI, as well as of visual search itself. In order to under- Note that all of these paradigms employ multitrial pre- stand the relationship between the visual search ana- exposure and test stages, in which the trials that define logue of LI and standard learning-based LI, it is first each stage are blocked. As opposed to these designs, necessary to describe a representative LI procedure of there are paradigms, typically used with human subjects, the latter type. in which the preexposed stimuli and the test stimuli are presented in multitrial pairs, as in negative priming, Eliciting Standard LI in Humans Stroop tasks, flanker tasks, and cost–benefit procedures. Braunstein-Bercovitz and Lubow (1998a, 1998b) used The blocked and paired procedures have different his- a typical procedure for inducing LI in adult humans. The torical roots, as well as different underlying assump- experiments were composed of two phases: preexposure tions. Curiously, there have been few attempts at a theo- and test. In Phase 1, the stimulus-preexposed subjects retical integration of the data generated from the two (PE group) were presented with 256 trials. Each trial types of paradigms. For our purposes, it is sufficient to contained the same pair of identical meaningless shapes. mention that the blocked trials procedures are commonly Concurrently, a pair of letters from the set TT, LL, TL, found in animal experiments that address theoretical is- or LT was displayed in the space between the two shapes. sues centered on learning and that may or may not appeal The subjects were required to judge whether the letter to attentional explanations. On the other hand, the paired pairs were identical (TT, LL) or different (TL, LT). This trial procedures are most commonly found in experiments masking task served to divert attention from the shape with human subjects that center on problems of attention that later, in the test phase, would be the stimulus to be and that may or may not appeal to learning explanations. learned. The nonpreexposed subjects (NPE group) re- However, as has long been recognized, attention and ceived the same masking task, but without the shapes. In learning are intimately related. Indeed, any theory that the test phase, the letter pairs from the masking task con- posits that stimulus preexposure modulates attention will tinued to be present in the same position as in the preex- predict multiple preexposure effects, from perceptual to posure phase, but they were now irrelevant. On test tri- cognitive. A reduction of attention to a tone should affect als, the shape that was preexposed was presented to both the perceived intensity of the tone (salience) and, as a the PE and the NPE groups. The subjects had to learn consequence, its subsequent associability. Similarly, a vi- that a change in the numerical value on a counter was as- sual shape that has been presented as an irrelevant stimu- sociated with the presence of the shape pair that was pre- lus not only will be weakened in its ability to enter into, or viously irrelevant for the PE group (when the preexposed reflect performance of, new associations (e.g., Braunstein- shape pair was present, a barpress decreased the number Bercovitz & Lubow, 1998a, 1998b), but also should be on the counter; a failure to press increased it). The PE less effective as a target in a visual search task (Lubow group reached the learning criterion (five consecutive & Kaplan, 1997).