This discussion arose from the USA Today article by Jerry A. Coyne that you can link to it here

One of the issues raised had to do with this quote from the article:

But we should continue to mete out punishments because those are environmental factors that can influence the brains of not only the criminal himself, but of other people as well. Seeing someone put in jail, or being put in jail yourself, can change you in a way that makes it less likely you'll behave badly in the future. Even without free will then, we can still use punishment to deter bad behavior, protect society from criminals, and figure out better ways to rehabilitate them. What is not justified is revenge or retribution — the idea of punishing criminals for making the "wrong choice." [bold highlight added]

The question was asked by a member of the group: “When the editorialist says we “should” take a particular course of action, does he mean to imply that we may be able to make a different decision? If so, isn’t that the exercise of free will?”

And another group member said that he had an argument against the claim that the purported fact that human behavior can be controlled by some mechanisms implies that humans do not have free will. This argument was based one objections to some passages by B. F. Skinner, from Walden II.

Some background about Walden II:

The character Frazier is the behaviorist (the mouthpiece for Skinner). Mr. Castle is the free will advocate.

Frazier claims that the behavior of an individual is already shaped by society toward “good conduct” (whatever conduct contributes to the success of society). Society has moral and legal rules and norms, parents and educators teach right and wrong, we punish transgressors, and reward conformists. The problem is only that society is inefficient at this—what is done is not methodical; it is not based on science. So, he says, why not experiment to discover first, what type of behavior (eg, paying taxes, helping others) is advantageous to society, and secondly, discover how society can make the individual adopt that type of behavior.

Rather than society trying to impose good conduct, society should set up reward schedules so that the individual is self-correcting toward good conduct. This is called self control—“behavioral processes which will lead the individual to design his own good conduct.” At this point in the discussion in Walden II, Frazier asks Mr. Castle what he would do if he possessed a science of human behavior. We will return to this in a moment, but first some explanations.

Skinner probably believes there is such a science, namely behaviorism. Skinner believed that the behavior of human individuals could be shaped by giving rewards for some behavior, or giving punishment for other behavior. The primary behaviorist mechanism by which human behavior is changed is shaping.

Shaping is just changing an individual’s actual behavior to have some behavior occur less often, or to have some behavior occur more often. A reward or “rewarding event” is just something that increases behavior, and a punishment is just something to decrease behavior. Skinner believed that all human behavior could be shaped by these methods. But what is more important to the argument is that Skinner is claiming that all human behavior is completely determined by these mechanisms, or causes. And this means that he is claiming that all human action is determined and that humans do not have free will.

The preferred method of shaping is through rewards: the rewarding of closer and closer approximations to the desired behavior, until success is achieved. When teaching this section on Skinner in a college classroom, one of us used to ask students what behavior they would like to see the instructor be performing by the end of the quarter. The students usually couldn't think of anything, so the instructor would recommend “putting this whiteboard marker in my ear.” We would agree that this would be a good thing to see. So, then the instructor proposed some kind of rewarding behavior: say, the students smiling at the instructor. Then the instructor said that every time his hand with the marker got close to his ear, the students should smile. After that, they should only smile when his hand was just as close as the first time, or closer. Thereby they would be rewarding closer and closer approximations to the behavior they wanted to see. Okay, now back to the article and argument.

Mr. Castle says that humans would loose their freedom, if all human behavior could be controlled by the mechanisms of behaviorism. He thinks that humans have free will already, but this science, if really successful, would effectively take away free will.

It's at this point that Frazier say that a science of behavior is impossible, if there is really free will (because a true science of behavior would be able to predict future behavior, but if there is free will future behavior cannot be predicted). However, this is a claim tangential to the main argument, which is also presented in Skinner's work. The main claim we are going to focus on is the claim made by Frazier that the increasing success of a science of human behavior makes it less and less plausible to believe that free will exists. Followin this, there are some discussions and arguments about free will, and discussions of “Skinner's” view of behaviorism that follow from these claims, but we can pass over them to take up the objection to the main argument.

Let's set aside the issue about whether or not behaviorism is the only science of human behavior. Humans probably learn behavior in other ways than simple reward and punishment, and shaping, but probably the objections can be applied to other arguments implying that if we can control human behavior in some way, then there is no free will. There also may be other causes of human behavior, genetic or otherwise, but we are concerned with those mechanisms that can change behavior. Let us just assume for simplicity that behaviorism is true and that there are no other mechanisms by which humans learn.

Here is the basic argument: supposing that human beings were successful in implementing a science of behavior, then it would seem that humans can act in ways otherwise than predicted by the science.

First, what would it mean to have a successful science of behavior, such that in general and for the most part the behavior of human beings in a society could be shaped toward behavior that that society wanted? Probably what we have to imagine is that most or all members of a society understand the principles and mechanisms of behavior control and shaping. That is, we can’t imagine that just a few people “in control” or “at the top” know these principles and mechanisms, because as individuals, our behavior is most affected by those nearest us. Those most affecting our behavior, and thus those most effective at shaping our behavior are those with whom we come in contact in our daily lives. It shouldn’t be imagined that some few “controllers” are in charge of our behavior. The authorities have very little control of our everyday behavior.

So, in this imagined case, we all would be using the principles of behaviorism to shape the behavior of those around us. This is for the good of society. We want others to act legally and morally, and we want them to want to act legally and morally.

We must also imagining that the mechanisms for shaping are pretty specific—specific enough so that each member of society could understand the mechanisms—and these mechanisms would be the type of rewards needed to be used to shape the types of behaviors desired.

Now, it seems, as a member of this society, that if one understands these mechanisms, then one can act otherwise than what would be predicted to be determined by these mechanisms. In general, knowing the mechanisms of shaping and the behaviors they are meant to shape, one can now do otherwise than the behavior meant to be shaped. For example, if a person knows that one way that others get one to increase a behavior is to smile at her or him, then one knows that when people smile at her or him, they may be trying to shape one's behavior. Suppose that one doed not even know what the behavior in question is. Nevertheless, one begins to note this smiling behavior on the part of others, and some behavior of one's might increase (unbeknownst to one, so far). But one is on one's guard, and after a while one recognize that smiling follows one's saying, “I helped an X, today” where X is some kind of person. Behaviorism would predict that this behavior would increase. But one knows the mechanisms of behaviorism as well. And one knows that the other is attempting to shape one's behavior. So, it seems one can act to ignore or act in ways counter to the shaping—contrary to what is predicted by the theory.

The claim was that humans could act in ways otherwise than those predicted by a science of human behavior. This isn't a proof that free will exists, unless we further prove that humans can always act in ways otherwise than any prediction. This argument also doesn't address the original question, which should be broken down as follows: if we believe that human behavior can be controlled or predicted, and we also believe that society should mete out punishment (or reward) for behavior, then isn't the second inconsistent with the first, because we are claiming that we could freely chose to punish, or not. I think another way to see an inconsistency in the quote is in the first line and the last line. The criminal is not wrong to behave as they do, but we, as a society should do, or are right to do, what we do.