“…The evidence for the sufficiency of S:IS contingencies for producing or influencing instrumental responses comes most clearly from studies of autoshaping (e.g., Brown & Jenkins, 1968;Gamzu & Schwartz, 1973;Gamzu & Williams, 1971;Locurto, Terrace, & Gibbon, 1976;Stiers & Silberberg, 1974;Wasserman, 1973Wasserman, , 1975 Williams & Williams, 1969;Woodruff, Conner, Gamzu, & Williams, 1977). The theoretical schemes that have been proposed to accommodate the fmdings of these studies range from those that attribute important roles to both S:IS and R:IS contingencies (e.g., Bolles, 1972;Estes, 1969Estes, , 1972Hearst & Jenkins, 1974;Schwartz & Gamzu, 1977;Walker, 1969) to those that try to make the S:IS contingency the most important, if not the sole, basis of the observed learned modifications (e.g., Bindra, 1972Bindra, , 1974Bindra, , 1976Lajoie & Bindra, 1976;Moore, 1973). However, little progress has been made so far toward experimentally understanding the relative roles of the two contingencies in the acquisition or performance of different types of responses.…”