Conditions of recovery after unlearning.

Abstract

The conditions and characteristics of the recovery of first-list associations following the end of interpolated learning (IL) were examined in five experiments on retroactive inhibition. In Exp. I-IV, recovery was studied as a function of the paradigm of transfer and the temporal arrangement of original learning (OL), IL, and the test of recall (MMFR). In the experimental treatments, -S"s learned three successive paired-associate lists conforming to either the A-B, A-C, A-D or the A-B, C-D, E-F paradigm. In one experimental arrangement the IL-test interval was varied, and in the other the temporal point of interpolation (TPI) was varied. Both manipulations yielded evidence of first-list recovery: first-list recall increased as a function of the IL-test interval and was greater when IL occurred early in the retention interval than when it immediately preceded the retention test. Recovery was found to depend critically on an adequate degree of OL. The increases in first-list recall were comparable for the two paradigms. Greater changes in performance were produced by manipulation of the TPI than of the IL-test interval. In Exp. V, recovery of backward associations was investigated under the A-B, C-B and A-B, B-C paradigms, and a trend toward recovery was found only for the latter. The results are viewed as consistent with the hypothesis that recovery reflects primarily the dissipation of response set interference, i.e., of the tendency established during IL to suppress the responses from earlier lists.The question of whether unlearning is a as the interval between the end of IL and reversible process has critical implications for the test of first-list retention is lengthened, current analyses of the mechanisms of in-Experimental tests of this hypothesis have terference in forgetting.The reduced focused on temporal changes in retention availability of first-list associations after in-under the A-B, A-C paradigm as measured terpolated learning (IL), which is taken as by the method of modified free recall evidence of unlearning, has been widely (MMFR). The MMFR test has been attributed to the operation of a process hav-viewed as providing a relatively pure meaing the functional properties of experimental sure of unlearning on the assumption that extinction (cf. Keppel, 1968; Postman, performance under this procedure is immune 1961). It follows that unlearned associa-to the effects of response competition and tions, like extinguished conditioned re-cannot be influenced by failures of list difsponses, should show spontaneous recovery ferentiation.

Critical issues in interference theory

“…Thus, while the present data support an extinction hypothesis they do not indicate why items are not, nor are likely not to be, extinguished [po 102]." The differential resistance of first-list items to unlearning, which has also been demonstrated in other studies (e.g., Postman, Stark, & Henschel, 1969), presents an important explanatory problem. At the same time, however, such findings rule out the assumption that the acquisition of A-C depends on the unlearning of A-B.…”

“…& Fraser, 1968). Recent experiments indicate,however, that rises in first-listrecall are likely to be observed after intervals of the order of half an hour (Forrester, 1970;Kamman & Melton, 1967;Martin & Mackay, 1970;Postman, Stark, & Fraser, 1968;Postman, Stark, & Henschel, 1969;Shulman & Martin, 1970). The conditions under which recovery may be expected after intervals of a day or more (e.g., Abra, 1969;Ceraso & Henderson, 1965, 1966Silverstein, 1967) remain to be fully specified.…”

Sources of Proactive Inhibition on Unpaced Tests of Retention

“…This convergence in error scores results from an increase for the second list and a decrease for the first list. This trend is consistent with the assumption that the second list is dominant at the end of interpolated learning but that the degree of this dominance declines as a function of time (Postman, Stark, and Henschel, 1969). The error probabilities and the temporal trends were quite similar for the two paradigms.…”

Auditory and visual serial position functions obey different laws

“…The effects of PI were weak immediately but slowly strengthened over time (see, e.g., Keppel & Underwood, 1962;Postman, Stark, & Fraser, 1968). The effects ofRI were strong immediately, but dissipated comparatively rapidly (see, e.g., Briggs, 1954;Postman et aI., 1968;Postman, Stark, & Henschel, 1969;Underwood, 1948aUnderwood, , 1948b. These time course differences seemed to hold over time scales ranging from minutes to days.…”