Recovery from blocking achieved by extinguishing the blocking CS

Abstract

Extinction-induced attenuation of single-phase and two-phase blocking was examined with rats in a conditioned lick-suppression task. In Experiment 1, which compared the effectiveness of single-and two-phase blocking, it was found that single-phase blocking was facilitated by the initiation of training with an A-US trial rather than an AX-UStrial. Single-phase (but not two-phase) blocking was attenuated as a result of 200 extinction trials with the blocking stimulus (Experiment 2). Experiment 3 revealed recovery from two-phase blocking after 800 extinction trials with the blocking stimulus. Recovery from both types of blocking was specific to the blocked CS trained in compound with the extinguished stimulus (Experiment 4). This is the first article to report that the blocking deficit can be reversed by extinguishing the blocking stimulus. These results are discussed in light of acquisition models (i.e., retrospective revaluation) and expression models (i.e., the comparator hypothesis).Kamin (1968) reported reduced conditioned responding to a cue (X) that was paired with an unconditioned stimulus (US) in the presence ofa second cue (A) as a result of A's having been previously paired with the US. This phenomenon, called "blocking," has been traditionally interpreted as an acquisition failure, in which the prior training with A prevents X from acquiring appreciable associative strength during training (e.g., Kamin, 1968;Mackintosh, 1975;Pearce, 1987;Pearce & Hall, 1980;Rescorla & Wagner, 1972). According to acquisitiondeficit models, the blocked stimulus is prevented from forming a strong association with the US; thus, because the subject has presumably not acquired the critical CS-US association, blocking should be irreversible in the absence of further training with the blocked stimulus.However, evidence for an expression-failure (e.g., retrieval-failure) interpretation ofcue-competition effects (of which blocking is one example) has appeared over the last 15 years. Most reports supporting an expressiondeficit view demonstrate reversal of cue-competition effects in the absence offurther training with the target con-

The elusive nature of the blocking effect: 15 failures to replicate.

“…The procedure of those experiments was again similar to previous reported studies that have demonstrated a blocking effect (Beckers et al, 2006;Blaisdell et al, 1999;Wheeler et al, 2008) (see Table A3 in Appendix A for a comparison with those previous studies and Table B1 and C1 in Appendix B and C, respectively, for an overview of the stimuli and procedure used in this series).…”

“…Effect size Blaisdell et al, 1999. Cohen's d = 1.79, based on estimations of means and standard deviations for the two-phase blocking and control group.…”

“…Blaisdell, Gunther, & Miller, 1999;Boddez, Baeyens, Hermans, & Beckers, 2011; M. Jones & Haselgrove, 2013;Mackintosh, 1971;Williams, 1996). Furthermore, blocking procedures have been used to distinguish between association-formation theories and rule-based or statistical theories of learning.…”

An instance theory of associative learning

“…In phase 3 of training, A is presented alone. Recovery from blocking is observed when, following all three phases of training, B behaves as a conditioned exciter of X (e.g., Blaisdell, Gunther, & Miller, 1999). The result is surprising inasmuch as learning of the B ➔ X relationship is blocked following phase 2 of training but is later expressed following extinction of the unblocked cue.…”

“…Minerva-AL has trouble with peak shift (see McLaren, Bennett, Guttman-Nahir, Kim, & Mackintosh, 1995). Minerva-AL predicts that some examples of retrospective revaluation (e.g., backward blocking) ought to be strong and reliable when, in fact, retrospective revaluation is experimentally elusive (see Blaisdell et al, 1999;Dopson, Pearce, & Haselgrove, 2009;Holland, 1999). Minerva-AL fails to acknowledge mediated conditioning due to its explanation for retrospective revaluation-a confusion recognized elsewhere (see Dwyer, 1999;Graham, Jie, Chan, McLaren, & Wills, 2011).…”