Separate groups of food-and water-deprived rats pressed a lever for food or water, respectively, on continuous reinforcement and various fixed-ratio and fixed-interval reinforcement schedules. Food-reinforced rats on continuous, FR 2-, or FI to-eec schedules showed consistently longer mean lever contact durations per leverpress than did water-reinforced rats on the same schedules. Mean lever-eontact-duration differences between food-and water-reinforced rats were greatly attenuated or disappeared under FR 4-, FR 8-, FI 20-sec, and FI 30-sec schedules of reinforcement. These results are interpreted as supporting earlier hypotheses that there are respondent components of operantly conditioned and autoshaped leverpresses, but that these respondent components weaken with partial reinforcement and the leverpress topography comes under the control of operant contingencies.Using pigeons as subjects, Brown and Jenkins (1968) described a procedure labeled "auto-shaping," which conforms to a respondent conditioning paradigm. Their discrete trials procedure involved periodic key illuminations followed immediately by grain access; no operant response was necessary for grain presentation. Within SO trials, most pigeons began to peck an illuminated key that signaled grain presentation.Several subsequent studies (Jenkins, 1973; Jenkins & Moore, 1973;Moore, 1973) employing autos haping procedures have compared response topographies of pigeons when food or water was the reinforcer. These studies have shown that when grain access is paired with key illumination, the pigeon "eats" the key, its pecks resembling subsequent grain pecking. When water presentation is paired with key illumination, the pigeon "drinks" the key, its pecks resembling subsequent drinking. One study (peterson, Ackil, Frommer, & Hearst, 1972), using rats as subjects, showed that when food reinforcers were presented contingent upon presentation of a retractable lever, rats often bit or chewed on the lever; rats receiving brain stimulation contingent on lever presentation seldom bit or chewed the lever. Hearst and Jenkins (1974) stated that if acquisition and maintenance of autoshaped responses directed at the key (or lever) involve respondent conditioning, those response topographies ought to be similar to topographies of responses elicited by the specific reinforcer. The studies More recently, operant response topographies of pigeons (Spetch, Wilkie, & Skelton, 1981) and rats (Cook & Hull, 1979;Hull, 1977)have been studied as a function of reinforcer received. Spetch et al. deprived pigeons of both food and water, then alternately provided grain or water reinforcers for keypecking on fixed-interval schedules. They found that, when different stimuli signaled impending food or water reinforcers, pigeons reliably showed different operant response topographies, as measured mechanically and as judged by human observers. The operant response topography differences were attenuated, however, when simple unsignaled alternation of food and water reinforcers was used, r...