Hernan I. Savastano

Cue Competition as a Retrieval Deficit

Four experiments using rats as subjects investigated the claim of Williams (1996) that cue competition results from an associative acquisition deficit, rather than a performance deficit. In Experiment 1, extinction of an overshadowing stimulus following overshadowing treatment increased responding to the overshadowed stimulus, thereby replicating prior observations with new parameters. In Experiment 2, an overshadowed stimulus failed to support second-order conditioning unless the overshadowing stimulus received prior extinction treatment. Experiment 3 replicated the recovery from overshadowing effect seen in Experiment 1 using a sensory preconditioning procedure. Most important, in Experiment 4 an overshadowed stimulus failed to block conditioned responding to a novel CS, but blocking by the overshadowed cue was observed following posttraining extinction of the overshadowing stimulus. These results, as well as those of Williams, are discussed in terms of traditional and more recent acquisition-focused models as well as an extension of the comparator hypothesis (Denniston, Savastano, & Miller, 2001).

Temporal context and conditioned reinforcement value.

The effectiveness of a stimulus as a conditioned reinforcer depends on the temporal context of reinforcement, that is, the overall rate of reinforcement in the situation. The dominant view has been that context determines the learned value of a stimulus directly, according to delay-reduction theory. By contrast, the contextual choice model (CCM) maintains that value is independent of context and incorporates the effects of context on choice in the framework of the matching law. The authors report 2 experiments with pigeons as subjects that use transfer tests to assess the value of stimuli in the concurrent-chains procedure. Results strongly support the assumption of CCM that pigeons learn the temporal relations between events independently of context but that context modulates the expression of that learning as choice.

Overshadowing of explicitly unpaired conditioned inhibition is disrupted by preexposure to the overshadowed inhibitor

In two conditioned lick-suppression experiments with rats, the interaction of preexposure to a target stimulus and subsequent overshadowing of conditioned inhibition treatment was examined. Blaisdell, Bristol, Gunther, and Miller (1998) have demonstrated that stimulus preexposure and overshadowing counteract each other in their effects on Pavlovian excitation, producing strong excitatory responding to the target stimulus. In the present experiments, a conditioned inhibition analogue of this effect was examined. Overshadowing of conditioned inhibition (i.e., attenuation of inhibitory control of behavior by the target stimulus) was demonstrated when a more salient stimulus was compounded with the target inhibitor during conditioned inhibition training. However, the overshadowing deficit was attenuated by preexposing the target stimulus prior to overshadowing treatment. To account for this phenomenon, we contrast the comparator hypothesis (Miller & Matzel, 1988) with contemporary models that focus on associative acquisition.

Temporal coding in Pavlovian conditioning: Hall-Pearce negative transfer.

The Hall-Pearce (1979) negative transfer effect in rats was used to examine whether temporal relationships are coded as part of the informational content of associations that result from CS-US pairings. The transfer effect consists of adeficit in conditioned responding following CS-USstrong pairings in Phase 2 that results from prior CS-USweak pairings in Phase 1. Using conditioned bar-press suppression, we found that gaps of different duration between CS termination and US onset in the two training phases resulted in less of a Hall-Pearce negative transfer effect than did an equivalent gap in the twotraining phases. The results are discussed with respect tothe temporal coding hypothesis (Matzel, Held, & Miller, 1988), the Pearce and Hall (1980) model, and Boutons (1993) interference model.

Interaction between preexposure and overshadowing: further analysis of the extended comparator hypothesis.

Three experiments with rats used conditioned suppression of barpress to test predictions of the extended comparator hypothesis, which assumes that the effectiveness of (first-order) comparator stimuli in modulating responding to a target conditioned stimulus (CS) is itself modulated by other (second-order) comparator stimuli. Experiment 1 demonstrated that both pretraining exposure to the target CS alone (i.e., CS-preexposure effect, also known as latent inhibition) and pretraining exposure to a compound of the target CS and nontarget CS (i.e., compound-CS-preexposure effect) counteract overshadowing, and that posttraining deflation (i.e., extinction) of the overshadowing stimulus attenuates responding to the target CS when overshadowing is preceded by a CS-preexposure treatment (i.e., yields a CS-preexposure effect), but not when overshadowing is preceded by a compound-CS-preexposure treatment. Experiment 2 examined the consequences of posttraining associative inflation of the overshadowing stimulus or the preexposure companion stimulus following conjoint compound-CS-preexposure and overshadowing treatment. Experiment 3 examined the consequences of posttraining inflation of the overshadowing stimulus or the context following conjoint CS-alone preexposure and overshadowing treatment. The results support the expression-focused comparator view in contrast to recent acquisition-focused models of retrospective revaluation.

Counterconditioning of an Overshadowed Cue Attenuates Overshadowing

In 3 Pavlovian conditioned lick-suppression experiments, rats received overshadowing treatment with a footshock unconditioned stimulus such that Conditioned Stimulus (CS) A overshadowed CS X. Subjects that subsequently received CS X paired with an established signal for saccharin (CS B) exhibited less overshadowing of the X-footshock association than subjects that did not receive the X-B pairings (Experiment 1). Experiment 2 replicated this effect and controlled for some additional alternative accounts of the phenomenon. In Experiment 3, this recovery from overshadowing produced by counterconditioning CS X was attenuated if CS B was massively extinguished prior to counterconditioning. These results are more compatible with models of cue competition that emphasize differences in the expression of associations than those that emphasize differences in associative acquisition.

Humans' responses to novel stimulus compounds and the effects of training.

Twenty-four college students participated in an experiment with stimulus compounds. Subjects learned to discriminate color stimuli that correlated with varying probabilities of reinforcement. Reinforcement consisted of points. For all subjects, two colors signaled a .80 reinforcement probability, and two others signaled a .20 probability. For compound-trained subjects, a fifth compound stimulus (composed of a high-probability color and a low-probability color) was correlated with a .10 reinforcement probability. During testing, interspersed probe trials required subjects to choose between two alternatives: a compound stimulus and either one of its constituent stimuli. Compound-untrained subjects preferred the compound over either individual stimulus, thus showing response summation. However, compound-trained subjects, having had experience with an exemplar compound, showed significantly lower choice proportions for a test compound, indicating that subjects’ responding to novel stimulus compounds is modifiable by experience with a single similar compound.

RETRACTED: Biological significance and posttraining changes in conditioned responding☆

This article has been retracted at the request of the Editor-in-Chief and Author. Please see http://www.elsevier.com/locate/withdrawalpolicy . Reason: Reanalysis of extensive published data collected in the animal learning and cognition laboratory at Binghamton University has detected errors in two data sets. Suppression ratios were incorrectly transcribed when the data were decoded. The corrected data in each instance maintains the reported ordinal relationships between groups, but fails to provide the inferential statistics necessary to support the asserted results. These reanalyses invalidate the conclusions of this article. Consequently, this article is hereby retracted. Author Miller greatly regrets the inconvenience that publication of erroneous results might have caused other researchers. He also wishes to express his gratitude to Escobar and Urcelay for their assistance in the reanalysis of the data. Ralph R. Miller, Binghamton University ; Francisco Arcediano, Auburn University ; Steven C. Stout, Jacksonville State University ; Gonzalo P. Urcelay, Binghamton University ; Martha Escobar, Auburn University

The relationship between value and temporal context is an empirical question: a reply to Fantino (2000).

E. Fantino (2000) argued that R. C. Grace and H. I. Savastanos (2000) experiments fail to elucidate the relationship between stimulus value and temporal context. His reasoning is that predictions for R. C. Grace and H. I. Savastanos probe tests based on delay-reduction theory (DRT) and the contextual choice model (CCM) are indistinguishable. However, his method of applying DRT to the probes ensures that temporal context will have no effect on which stimulus is preferred, contrary to the core principle of that theory. The only basis for differential responding in the probes is baseline training, and R. C. Grace and H. I. Savastanos data clearly show that the terminal-link schedules, independent of temporal context, control choice in the probes, as predicted by CCM.