We accept that right and wrong (in the ethical sense) exist as valid descriptions of events or states of affairs. Furthermore, we accept that right and wrong are normative, which is to say that we should do what is right, and avoid what is wrong. (The exact source of this normative power is, of course, a topic for another time.) What about praise and blame? Even if we accept the previous claims we aren’t forced to conclude that we should blame people for wrongdoing and praise them for following the morally upright course of action. Some incorrectly conclude that because ethics is normative praise and blame is justified, but this simply isn’t true. For example, consider that if you want to enter your house you should open the door. In this case opening the door is normative. However, if you fail to open the door, and are thus unable to get into your own house, people don’t blame you or punish you (although they may laugh at or pity you).
This doesn’t mean that praise and blame are indefensible, simply that if they are to be defended it must be because they are good (ethics implies that we should praise and blame people), not as part of the nature of ethics. Assuming that we have a consequentialist view of ethics (because ethical analysis is almost always simpler when the ethical system in question is consequentialist, and usually the conclusions of different ethical theories agree in the majority of cases) it is rather easy to defend praise and blame. We can defend them for the obvious reason that praise for good behavior and blame (and punishment) for bad behavior encourages people to act more ethically, and thus in the long run things are better. It is true that many people act ethically for reasons other than the possibility of praise and blame, but what got many people started thinking about ethical behavior, and what they should do, were the consequences, to them, of their acts, in the form of praise and blame.
And although this view encourages us to behave in basically the same way we do now (for example, throwing criminals in jail) the reasons are not quite the same. One notable difference is that under this view imprisoning an innocent man is just as bad as imprisoning a guilty man, since both cases result in the same amount of harm being done, and suffering isn’t made less wrong because it happens to a person who has acted immorally. Now this doesn’t mean that we should go around punishing innocent people, since if our system is really trying to encourage good behavior then we must be accurate with our punishments. If we punish indiscriminately people will feel free to behave badly, since they are just as likely to be punished if they do nothing wrong. It does mean, though, that it is acceptable to punish some innocent people, as long as we do it sufficiently infrequently, and this shouldn’t be a surprising conclusion, since many times we convict criminals without indisputable evidence, and thus by the laws of probability we must be convicting some innocent people.
A final case to consider is if it makes sense to praise or blame a person when they couldn’t have changed the outcome even if they wanted to, or were unaware that it needed changing. In these cases we have to consider whether blame or praise can really do any good. Certainly it can’t help the person who commits a wrong unintentionally, or when under a compulsion, since they would have done the right thing if they could have, and no amount of pervious encouragement could have changed the results.* What we do need to consider is what effect blaming or praising these people will have on the actions of others who can choose. We can’t simply exempt from praise and blame all people who do wrong as a result of a compulsion or an accident, since then if someone wishes to do wrong they may feel free to ingest some drug that will make them lose control, in the hope that they will commit the crime they wanted to, and at the same time won’t be blamed for it. What we need to do then is to exempt people from praise and blame only if their actions were out of control in both a first order and second order sense. This means that not only must they have been in a situation where they couldn’t have known to choose, or had the power to choose, the right outcome over the wrong outcome, but also that they couldn’t have avoided being in that situation either. And in fact this is how most modern legal systems treat such situations. Maybe we aren’t doing such a bad job after all.
* Special case: some people may be under a permanent compulsion to commit some wrong. In these cases we usually imprison them to prevent them from acting on this compulsion, but since there are many cases where the compulsion might be a one-time event I am passing over this special case in our discussion here.