361). Similarly, reinforced training of the separate com ponents followed by tests with the compound stimulus alsc yields a higher level of responding to the compound thar to the components (e.g., Weiss, 1972). Despite the evidence for a summation process, there is also abundant evidence that a compound can act inG estalt-like fashion as a functional unit. The evidence arises primarily from research with patterning schedules, which can produce explicit differential conditioning of a compound and its components. Patterning was first reo ported by Pavlov (1927, p. 144) and, according to Razran (1965, 1971), has been obtained frequently in Soviet laboratories. In the English-language literature, positive and negative patterning have been obtained by Hull (1940), with the human galvanic skin response (GSR), Woodbury (1943), with the dog's rewarded nose press, and Rescorla (1972, 1973), with the rat's rewarded barpress. Patterning, particularly negative patterning, appears inimical to a simple summation rule. Responding to the reinforced compound in positive patterning might be construed as the summation of subthreshold associative strengths acquired to the separate components. However, there is no way that a lack of responding to the nonreinforced compound in negative patterning could represent the summation of the high levels of overt responding to the components (Woodbury, 1943). Additional evidence for the integration of compound stimuli has been

No abstract