A quote famously attributed to Skinner about controlling a population in terms of "masturbating the lions" – his reference to some other author's work – in its full context is seen to be criticism by Skinner rather than approval. The quote is in the 5th paragraph below, highlighted. For the bottom line just skip to there.
The pdf of Beyond Freedom and Dignity is here: /psychology/
(Paragraphs are numbered for reference.)1
The importance of the literature of freedom can scarcely be questioned. Without help or guidance people submit to aversive conditions in the most surprising way. This is true even when the aversive conditions are part of the natural environment. Darwin observed, for example, that the Fuegians seemed to make no effort to protect' themselves from the cold; they wore only scant clothing and made little use of it against the weather. And one of the most striking things about the struggle for freedom from intentional control is how often it has been lacking. Many people have submitted to the most obvious religious, governmental, and economic controls for centuries, striking for freedom only sporadically, if at all. The literature of freedom has made an essential contribution to the elimination of many aversive practices in government, religion, education, family life, and the production of goods.2
The contributions of the literature of freedom, however, are not usually described in these terms. Some traditional theories could conceivably be said to define freedom as the absence of aversive control, but the emphasis has been on how that condition feels. Other traditional theories could conceivably be said to define freedom as a person's condition when he is behaving under non-aversive control, but the emphasis has been upon a state of mind associated with doing what one wants. According to John Stuart Mill, 'Liberty consists in doing what one desires.' The literature of freedom has been important in changing practice (it has changed practices whenever it has had any effect whatsoever), but it has nevertheless defined its task as the changing of states of mind and feelings. Freedom is a 'possession'. A person escapes from or destroys the power of a controller in order to feel free, and once he feels free and can do what he desires, no further action is recommended and none is prescribed by the literature of freedom, except perhaps eternal vigilance lest control be resumed.3
The feeling of freedom becomes an unreliable guide to action as soon as would-be controllers turn to non-aversive measures, as they are likely to do to avoid the problems raised when the controllee escapes or attacks. Non-aversive measures are not as conspicuous as aversive and are likely to be acquired more slowly, but they have obvious advantages which promote their use. Productive labour, for example, was once the result of punishment: the slave worked to avoid the consequences of not working. Wages exemplify a different principle; a person is paid when he behaves in a given way so that he will continue to behave in that way. Although it has long been recognized that rewards have useful effects, wage systems have evolved slowly. In the nineteenth century it was believed that an industrial society required a hungry labour force; wages would be effective only if the hungry worker could exchange them for food. By making labour less aversive - for instance, by shortening hours and improving conditions - it has been possible to get men to work for lesser rewards. Until recently teaching was almost entirely aversive: the student studied to escape the consequences of not studying, but non-aversive techniques are gradually being discovered and used. The skilful parent learns to reward a child for good behaviour rather than punish him for bad. Religious agencies move from the threat of hellfire to an emphasis on God's love, and governments turn from aversive sanctions to various kinds of inducements, as we shall note again shortly. What the layman calls a reward is a 'positive reinforcer', the effects of which have been exhaustively studied in the experimental analysis of operant behaviour. The effects are not as easily recognized as those of aversive contingencies because they tend to be deferred, and applications have therefore been delayed, but techniques as powerful as the older .aversive techniques are now available.4
A problem arises for the defender of freedom when the behaviour generated by positive reinforcement has deferred aversive consequences. This is particularly likely to be the case when the process is used in intentional control, where the gain to the controller usually means a loss to the controllee. What are called conditioned positive reinforcers can often be used with deferred aversive results. Money is an example. It is reinforcing only after it has been exchanged for reinforcing things, but it can be used as a reinforcer when exchange is impossible. A counterfeit bill, a bad cheque, a stopped cheque, or an unkept promise are conditioned reinforcers, although aversive consequences are usually quickly discovered. The archetypal pattern is the gold brick. Countercontrol quickly follows: we escape from or attack those who misuse conditioned reinforcers in this way. But the misuse of many social reinforcers often goes unnoticed. Personal attention, approval, and affection are usually reinforcing only if there has been some connection with already effective reinforcers, but they can be used when a connection is lacking. The simulated approval and affection with which parents and teachers are often urged to solve behaviour problems are counterleit. So are flattery, backslapping, and many other ways of 'winning friends'.5
Genuine reinforcers can be used in ways which have aversive consequences. A government may prevent defection by making life more interesting - by providing bread and circuses and by encouraging sports, gambling, the use of alcohol and other drugs, and various kinds of sexual behaviour, where the effect is to keep people within reach of aversive sanctions. The Goncourt brothers noted the rise of pornography in the France of their day: 'Pornographic literature', they wrote, 'serves a Bas-Empire * ... one tames a people as one tames lions, by masturbation.'
* The final period of the Roman Empire
Genuine positive reinforcement can also be misused because the sheer quantity of reinforcers is not proportional to the effect on behaviour. Reinforcement is usually only intermittent, and the schedule of reinforcement is more important than the amount received. Certain schedules generate a great deal of behaviour in return for very little reinforcement, and the possibility has naturally not been overlooked by would-be controllers. Two examples of schedules which are easily used to the disadvantage of those reinforced may be noted.7
In the incentive system known as piece-work pay, the worker is paid a given amount for each unit of work performed. The system seems to guarantee a balance between the goods produced and the money received. The schedule is attractive to management, which can calculate labour costs in advance, and also to the worker, who can control the amount he earns. This so-called 'fixed-ratio' schedule of reinforcement can, however, be used to generate a great deal of behaviour for very little return. It induces the worker to work fast, and the ratio can then be 'stretched' - that is, more work can be demanded for each unit of pay without running the risk that the worker will stop working. His ultimate condition - hard work with very little pay - may be acutely aversive.8
A related schedule, called variable-ratio, is at the heart of all gambling systems. A gambling enterprise pays people for giving it money - that is, it pays them when they make bets. But it pays on a kind of schedule which sustains betting even though, in the long run, the amount paid is less than the amount wagered. At first the mean ratio may be favourable to the bettor; he ·wins'. But the ratio can be stretched in such a way that he continues to play even when he begins to lose. The stretching may be accidental (an early run of good luck which grows steadily worse may create a dedicated gambler), or the ratio may be deliberately stretched by someone who controls the odds. In the long run the 'utility' is negative : the gambler loses all.
Etc., etc ...
Beyond Freedom and Dignity