“…Furthermore, the truly random paradigm was the most stressful of the three conditioning paradigms, since dogs in this group received more shocks and since completely unpredictable shocks (i.e., unsignaled shocks with no signaled safety periods) are known to be more stressful than predictable shocks (e.g., Weiss, 1971;see Weinberg & Levine, 1980, for a review). There is a considerable theoretical (e.g., Bruner, Matter, & Papanek, 1955;Tolman, 1948) and empirical (e.g., Johnson, 1953;Spence & Lippit, 1940)literature that suggests that, in circumstances of extreme stress, an CONTEXTUAL ASSOCIATIONS 319 organism narrows its attentional field to the most salient cues in the environment (for reviews, see Easterbrook, 1959;Melton, 1950;Thistlethwaite, 1951). Other cues, such as some contextual ones, may be functionally ignored and thus may not gain associative strength even when they are correlated with reinforcement.…”