Beyond Freedom and Dignity (B. F. Skinner)

| reading

This is a review of B. F. Skinner's famous book Beyond Freedom and Dignity first published in 1971. The book spells out the practical consequences of a behavioral science that assumes free will does not exist and suggests how this science could be used to improve society. I just finished reading the book. Yes, I'm 50 years late! I try to extract the most important points and order my own thoughts. If you haven't read the book, this should give you a good impression of its content and maybe make you want to read it.

I'll start with a bit of theoretical background that is useful to understand the arguments. Then I give a summary. For the summary I try to put myself in Skinner's position and withhold any criticism. I think it's best to swallow all assumptions first and get a good understanding of the ideas before I allow myself some critical remarks, which will be the last part.


The most fundamental concept necessary to understand Skinner's ideas is that of contingencies. Since it's not defined explicitly in the book itself, I give a quick introduction here. Contingencies of punishment and reinforcement consist of three steps:

  1. antecedent: discriminative stimulus,
  2. behavior: operant response, and
  3. consequence: reinforcement or punishment.

For example, whenever you tell your dog to sit (discriminative stimulus) and he actually sits down (behavior), you give him a treat (reinforcement). After one or several occurrences of these three steps, the discriminative stimulus becomes a signal for the individual that a certain behavior leads to either reinforcement or punishment with some probability. Accordingly, the probability of the behavior given the discriminative stimulus either increases or decreases. Reinforcement increases the probability, punishment decreases the probability. Reinforcement and punishment can be qualified as either positive, if a stimulus is added, or negative, if a stimulus is removed as a consequence of behavior. This leaves us with four types of contingencies by their consequences (examples in parentheses):

Reinforcement and punishment are not necessarily intentional, originating from other people. If you see a stove with steaming pots, touch a hot plate and get burned, this is an instance of positive punishment by your non-social environment.


Skinner applies a strong form of determinism to behavior: humans' and other animals' behavior is entirely determined by immediate stimuli of the environment (reflexes), the individual's past experiences (contingencies of reinforcement and punishment) and genetics. Explanations of behavior based on free will, feelings, or states of mind are dismissed as prescientific references to autonomous man, the illusion of a little man sitting in our head that controls our behavior.

The direction of the controlling relation is reversed: a person does not act upon the world, the world acts upon him. (p. 206)

Defining freedom as being able to do what you want thus becomes meaningless. Instead, freedom is to be free of aversive stimuli. This explicitly includes deferred aversive consequences of behavior, which are difficult to prevent if we follow the illusion of free will or our momentary feelings (I like candy and I'm free to do what I want, so I eat lots of it and a few years later I'm a fat diabetic with rotten teeth). Freedom does not require absence of control, as long as it isn't aversive.

Similarly, without free will, giving someone credit for his behavior is problematic.

The amount we give is inversely propertional to the conspiciousness of the causes of his behavior. If we do not know why a person acts as he does, we attribute his behavior to him. We try to gain additional credit for ourselves by concealing the reasons why we behave in given ways or by claiming to have acted for less powerful reasons. (p. 61)

The dignity or worth of a person is derived from the amount of credit he receives for his behavior. Since behavior is deterministic, crediting somebody for his behavior just means that we don't understand the underlying causes, like people in the past, who credited natural phenomena to god(s) because they did not understand physics. Skinner notes that this notion of dignity stands in the way of progress because advances in technology may reduce the opportunities for receiving such credit (the dishwasher robs me of my wife's credit for washing the dishes by hand).

If you believe in free will, freedom and dignity, but are not fundamentally opposed to control, you probably prefer punishment over reinforcement. Everybody is free to do the right or the wrong thing and will receive his due credit. Punishment is unpleasant, but overt, whereas reinforcement can go unnoticed by the individual and may seem more manipulative. When we punish a person, we leave it up to him to discover how to behave well (p. 69). He might just pick up other bad, but unpunishable behavior or learn to avoid the punisher. If you believe in free will and dignity, this is desirable; if you don't, it's just an ineffective and unnecessarily cruel way of controlling behavior. A person considered good is just a person under less visible control. Consequently, Skinner would rather design a world in which behavior likely to be punished seldom or never occurs, a world that builds automatic goodness via non-punitive contingencies. Since all control is exerted by the environment anyway, we should change the environment instead of the individual. The result should be a world of reinforcing activities without dignity and moral struggle.

If we talk about automatic goodness, we need to talk about values. What is good and what is bad? According to Skinner, good things are positive reinforcers (p. 103) and bad things are negative reinforcers. Things are considered good or bad as a result of evolutionary selection: there is survival value in eating sweet fruits and avoiding bitter (possibly) poisonous fruits. Feelings, as mentioned before, are only by-products of reinforcing effects of things or events. People behave for the good of others because they are under social control, that is, other people arrange contingencies that make us behave for their or our common benefit. Social control and conditional reinforcement help to make deferred aversive consequences effective. For example, telling the truth is beneficial for a group of people in the long run, but the reinforcing effect for the individual is too remote to be effective. A ‘normative statement’ like the following can bridge the gap.

‘You should (you ought to) tell the truth’ is a value judgement to the extent that it refers to reinforcing contingencies. We might translate it as follows: ‘If you are reinforced by the approval of your fellow men, you will be reinforced when you tell the truth.’ (p. 112)

The control of others can be excessive or conflict with our personal reinforcers and we may turn to immediate gratification to avoid it, which is then described by others as a lack of values.

From social control, we move right to the concept of culture, which the book defines as a set of contingencies arranged by other people and the behaviors they generate. What members of a culture find reinforcing differs, depending on genetics, natural and social contingencies (this is an instance of cultural relativism). Cultures evolve similarly to species: new practices enter a culture like genetic mutations and are selected if they contribute to the survival of the culture in its competition with the physical environment and other cultures. Cultures that select contingencies that make its members work for the survival of their culture are more likely to survive, so it's no surprise that we observe this feature in most cultures. The survival advantage of cultures for their members is that they have the power to bring people under the control of remote consequences of their behavior. If we were to design a culture, we should strengthen this aspect and make the culture self-perpetuating.

At the end of the book, Skinner describes what his theory of behavior means for the self-image of man and addresses some problems in this regard. One such problem is that of thinking, that is, complex ‘cognitive’ activity. He notes that we never observe discrimination, generalization, abstraction, memory, or association. All we can see is behavior and as we know, behavior is determined by genetics (the product of natural selection), the social/cultural environment (also product of a selection process), and the natural environment. Another problem is that of consciousness, introspection, self-knowledge, and self-control. Again, all these things are products of the environment.

Without the help of a verbal community all behavior would be unconscious. Consciousness is a social product. It is not only not the special field of autonomous man, it is not within range of a solitary man. (p. 187)

Skinner builds on Freud's structural model of the psyche, but the ego, the mediating agent between id and super-ego, aka ‘autonomous man’, is gone.

The controlling self (the conscience or superego) is of social origin, but the controlled self is more likely to be the product of genetic susceptibilities to reinforcement (the id, or the Old Adam). The controlling self generally represents the interests of others, the controlled self the interests of the individual. (p. 195)

Autonomous man, just like god, is a piece of explanatory fiction (p. 196) that must be done away with to prevent the abolition of the human species. Still, people are not merely passive victims, because they shape their own environment.

The evolution of a culture is in fact a kind of gigantic exercise in self-control. As the individual controls himself by manipulating the world in which he lives, so the human species has constructed an environment in which its members behave in a highly effective way. (p. 201)

Cultural practices even bring death, the ultimate remote event, to bear on our behavior. A person may find solace in (that is, he may get reinforcement from) his contributions to the goods of others that will outlive him.

Skinner also addresses criticism of the application of his results from studies of lower animals to humans. For obvious ethical reasons, experiments which control for genetic differences and individual environmental histories are difficult to perform with human subjects.

There is, of course, always the danger that methods designed for the study of lower animals will emphasize only those characteristics which they have in common with men, but we cannot discover what is ‘essentially’ human until we have investigated non-human subjects. [...] Some of the complex contingencies of reinforcement now under investigation generate behavior in lower organisms which, if the subjects were human, would traditionally be said to involve higher mental processes. (p. 197)

Critical remarks

The book takes strong determinism as an axiom from which the whole argument is built. For Skinner this determinism seems self-evident or impossible to (dis)proof. Not even some form of randomness is considered, let alone free will. All we get as justification for the rejection of free will are analogies with biology and physics.

We can follow the path taken by physics and biology by turning directly to the relation between behaviour and the environment and neglecting supposed mediating states of mind. Physics did not advance by looking more closely at the jubilance of a falling body, or biology by looking at the nature of vital spirits, and we do not need to try to discover what personalities, states of mind, feelings, traits of, character, plans, purposes, intentions, or the other perquisites of autonomous man really are in order to get on with a scientific analysis of behaviour. (p. 20)

Other options are branded non-scientific or pre-scientific. However, saying that human behavior isn't influenced by states of minds or thoughts because stones falling from a cliff aren't either might not convince every critic. The way I read this book is as a thought experiment: let's assume the world is deterministic and there is no free will, how do we explain human behavior in the light of our experimental findings and what practical consequences can we derive from our explanation?

Skinner has been accused of greedy reductionism or nothing-buttery: all human behavior is nothing but the consequence of contingencies. In science, reducing complexity to simple underlying principles is generally desirable and applauded, but you need to make falsifiable predictions. Skinner makes falsifiable predictions for the behavior of animals in his lab settings, but is not able to make predictions about human behavior. All his explanations of human behavior are just retrofitted to observations after the fact. We can't really blame him because controlling genetics and the complete history of contingencies in an experiment with humans would clearly be unethical. But the fervour with which he brings forward his positions against alternative accounts of human behavior is somewhat unexpected. This brings me to my last point.

In the summary above I omitted all the unspecific rhetoric against the literature of freedom, the literature of dignity, the pre-scientific view etc. Skinner tends to identify his views with science per se. Dissenting views are not scientific, backwards, hindering progress, dangerous, maybe even resulting in the abolition of the human species (p. 196). Given the leaps of faith his ideas require, this appears rather self-righteous. I always admire people who have so much trust in their own abilities that they promote their insights as indisputable truths and would like to reshape the world based on them. At the same time, they make me deeply suspicious because more often than not their behavior is explained by hubris or some ulterior motive.

Despite these critical remarks, I found the book an enriching and rewarding read. Even if not all behavior can be explained with contingencies, much of our behavior is probably less intentional then we like to admit. With climate change and ongoing wars all over the news, the idea of manipulating the social environment to build automatic goodness is as relevant now as it was in the 70s. Appealing to autonomous man to use fewer resources and love thy neighbor has not proved effective.