Kohn, Alfie.  Punished by Rewards
New York: Houghton Mifflin, 1993.
ISBN 0395650283
Page 14

It is not by accident that pop behaviorism has come to suffuse our lives. There are identifiable reasons to account for its popularity, beginning with the belief systems already in place which it complements. One of these I mentioned earlier: our pragmatism, and specifically our tendency to favor practical techniques for getting the job done as opposed to getting bogged down with theories and reasons. A nation of busy pioneers and entrepreneurs has no time for figuring out the source of a problem; much more compatible with the American spirit is a simple declaration that would seem to assure results: "Do this and you'll get that."

Ironically, rewards and punishments not only lie at the core of faith but are central to our idea of rationality as well, particularly as it makes its presence felt in economic choices. Rational decision-makers, by definition, are said to seek what is pleasurable and to avoid what is aversive or costly. Rationality, in turn, is central to what it means to be human, at least to many Western thinkers. A number of writers have recently challenged both steps of this argument, but pop behaviorism makes intuitive sense to us as a result of the assumptions built into our economic system.

In fact, behavioral psychology and orthodox economic theory have established a sort of mutual admiration society that flatters both fields, but only by creating a truncated picture of the human being whose actions they seek to analyze. On the first pages of their textbooks, economists often nod in the direction of behaviorism to justify their fundamental assumptions about what motivates consumers or workers. Psychologists in turn assume that the process of weighing costs and benefits that describe how we go about purchasing an appliance is also what we are doing when talking with a lover.

Page 27

Just as threats are simply a more blatant version of control than bribes are, so token economies merely exaggerate the manipulation that describes other, less systematic, applications of rewards. The point to be emphasized is that all rewards, by virtue of being rewards, are not attempts to influence or persuade or solve problems together, but simply to control.

Page 52

That rewards punish is not due only to the fact that they are controlling. They also have that effect for a second, even more straightforward, reason: some people do not get the rewards they were hoping to get, and the effect of this is, in practice, indistinguishable from punishment. Many managers and teachers make a point of withholding or withdrawing a reward if their charges do not perform as instructed. The goody is dangled and then snatched away. In fact, this is precisely what many behaviorists recommend doing. While taking care to urge that children not be punished (by which is meant making something bad happen to them), they freely prescribe the use of "response costs" (by which is meant making something good not happen to them).*

Unfortunately, those who haven't been trained to make such distinctions might fail to understand that when something desirable has been taken away they are not supposed to feel punished.

A parent tells a child that continued good behavior will be rewarded with a visit to the circus on Sunday. On Saturday, the child does something that annoys the parent, which prompts a familiar warning: "Keep this up and you can forget the circus tomorrow." Can there be any doubt that this threat to remove a reward is functionally identical to a threat to employ a punishment?

Page 53

A reward, by definition, is a desired object or event made conditional on having fulfilled some criterion: only if you do this will you get that. If I promise to give you a banana tomorrow, that is not a reward. If I promise to give you a banana tomorrow for helping me out today, that is a reward -- and if I don't give it to you, you will probably feel as if you are being punished. To avoid having this happen, I must avoid giving you things on a contingent basis.