This paper argues that behavior analysis is not technological to a fault, but rather has a faulty technology by being incomplete. The paper examines reinforcers and punishers that result from the outcomes of either (1) striving for better experimental control, or (2) inventing theories to explain why current control is imperfect. (JDD)

This commentary on EC 608 301 points out that devotion to the natural science canon of experimental proof in the area of generality testing requires proof procedures that are difficult, expensive, or impossible, but adopting a different canon of proof will not produce the desired certainty about the accuracy of the answer. (JDD)