“…shall consider a situation which involves one which reinforces Aj ( j # i), then wit,h probability c the pattern becomes connected For our first illustrative interpretation, we to Aj, while with probability 1 -c its connection to Ai remains unchanged. uncertainty with respect to trial outcomes, but no differences in utility for different outcomes, vie., the familiar experiment on simple predictive behavior or verbal conditioning (Estes & Straughan, 1954;Grant et al, 1951;Neimark, 1956). The subject's task is simply to choose between two response alternatives, e.g., to operate one of two response keys or to say one of two designated words a t the beginning of each trial, attempting to be "correct" as often as possible.…”