Kant analyzes the "good will." A good will is a well working will. It does not act from some hierarchy of absolute goodness; instead, it is simply a matter of a well-functioning psychological motivation. Its good functioning comes from its acting in accordance with the following three rules that are embedded in our minds innately--perhaps even by way of evolutionary selection. These rules are referred to as "categorical imperatives" because, when we are sensitive to the way our minds work, then we follow those imperatives categorically or in all cases when we act at all. While our minds are equipped with these rules, it is possible either to suppress them or to cultivate them. The ethically responsible person, of course, cultivates them:
Act always as if you are your very own law-giver; make your rules autonomously.
The second rule is also intuitively obvious. We probably all have had a feeling at times that someone has "used" us for his or her personal advantage and to our detriment. We will not see such an action as fair. At the same time, Kant reminds us that being unkind to ourselves is not fair either. If a person sacrifices his or her own life while trying to rescue people from a sinking boat, that person does not have our unequivocal support since s/he has failed to treat himself as an end in him-/herself. We must be evenly concerned about ourselves and about others. Kant observes the same about a person doing military service for his/her country: If, in the process of doing so, one were to lose his or her life, then we should not admire the ethical action without some qualms since the person has failed in his/her obligations to him-/herself.
Finally, we must always take responsibility for our decisions. We cannot blame others for what we have done. Jean Paul Sartre, a French philosopher, suggests that we ultimately always have choices. If someone tries to force our actions, we always have a choice to commit suicide. We also cannot subscribe to some morality by saying that we are, say, Nietzschean or that we are Christian or that we are Buddhists. In all such cases, we are not taking responsibility. We are working with a heteronymous morality when we subscribe to other systems and not with an autonomous morality, the latter being required by the nature of our minds equipped with the categorical imperative. We must reason about our moral choices, infer sound conclusions, and--in short--do our very own thinking.
God, Soul, and Immortality
For Kant, being ethical is a development. To get a good idea where we want to develop to, it helps to have a god concept. God is a perfected being. God also is merely a concept for Kant. We must not confuse "god" with a Christian god and we must also not confuse the ethical progress of development with Christian morality. Kant rejects Christian morality as heteronymous; that is, someone from outside of us is trying to tell us what we should do. That is not acceptable to Kant. Our morality must be generated from our own thinking about problems by way of using our "good will." Again, the "good will" is a psychological make-up that is conditioned by and that abides by the three intuitive rules above. Kant does not give us a list of things to do and things not to do; Kant urges us to use our own analysis and our own thinking under constraint of our in-built good will. Kant also understands that the process of ethical development is endless. For that reason, we may find our progress easier to imagine if we assume an immortal soul. Neither the soul nor god are in any way real in the real, physical world; we have no proof for either one. However, assuming both as metaphors helps our thinking along: one metaphor gives us a goal; the other encourages us to believe that we can get there. One also cannot compel faith of any sort, so it makes little sense to go beyond the metaphors for "god," "soul," and "immortality." As Nietzsche was to say later, "To him who needs a god, there is a god; and to him who needs no god, there is none." Indeed, believing in a god and eternal bliss or eternal punishment would be counter-productive for Kant. Our ethical action should spring from a sense of duty that we are doing a right action; it should not be in response to fear of eternal damnation or anticipation of eternal bliss. Either response would be at too low a level of moral action, being heteronomous and not autonomous. So, Kant's view is quite atheistic; religious terminology is a convenience for thinking, not anything that is in the real world. That's the reason why when Wilhelm I of Prussia followed his open-minded uncle Friedrich the Great of Prussia, Kant did get a nasty letter admonishing him to quit his atheistic teaching or he might lose his job. The letter was signed by a bureaucrat named Wöllner.
Kant also makes a distinction between "legality" and "morality." Following the laws, all the while perhaps even looking for loopholes, is much inferior action to doing what one knows to be right. The first line of obligations should come from one's sense of right action, not from what one thinks the law specifies. "Legality" is a lower form of right action than "morality." We follow what we know to be our duty; if that happens to coincide with what the law says, that's fine; if it does not, the law needs to change, and we need to work toward that end. Frequently, the challenge is made whether Kant considers lying to be ever OK under certain circumstances. Short answer: Yes, he does. In the Critique of Practical Reason, he points out that he will bow to a person of noble rank, an aristocrat; but inside, he never bows to anyone at all because all persons have equal value, no matter at what level they are born into the social hierarchy. As you can see, that's a bow with one's fingers crossed or a lie with some level of expedience. Many of these Kantian values, you can also find among ethical heroes who never have heard of Kant. If you are interested in Korean history, take a look at Admiral Yi Soon-Shin. Many of his views on equality and on considering morality in warfare anticipate in the 16th century already what Kant will express in the 18th century. So, Kant may indeed be onto some universals of human morality here.
Kant argues that one must act from a sense of duty. He contrasts duty ethics with what I'd call sappy ethics. Sappy ethics responds to, say, a hurt dog with tears and immediate intervention. Sappy ethics might respond to images of suffering humans with sorrowful engagement. Typically, such interventions tend not to be all that successful in alleviating problems. If I feel sorry for the homeless and if I then collect canned goods in my church for distribution to the homeless, I have not really understood the full problem involved. Duty ethics requires a thorough analysis of what I should do. For example, I might analyze the social system as an improper agent in leading to a faulty distribution of a society's resources to its members. I would then have a moral obligation to right this unfairness. While I would not do wrong to share some canned goods, the actual duty should push me to alleviating the general wrong by pushing my society to a commitment to fairness and justice. If I am dealing with abortion, I might succumb to a general feeling that the little baby wants to live and so should live. That's sappy ethics. Instead, I need to analyze thoroughly rights and responsibilities of all involved in the situation as well as the latest state of medical and biological knowledge. From the vantage point of such a thorough analysis and the deep understanding, I am then in a position to make a circumspect, fair, and just decision driven by universalizability, kingdom of ends, and autonomy. The outcome might even be in contrast to the sappiness of my nature; however, if the analysis is right, I have a duty to follow the analysis and not the sappiness. Kant also does not nail down the moral community. If rational agents include also non-humans, then my duty would encompass them also. Kant is not prescriptive of an ethics; he simply shows how to think ethically. His is not an ideology; his is a process.
Pleasure as Highest Good
Kant disagrees with the Epicurean hedonist assumption that pleasure is the highest good. Also Mill and the Utilitarians make this assumption. Here is an interesting paradox that results from assuming pleasure as the highest good. If pleasure were the highest good and I acted in a bad manner, then I should be punished for acting bad by not achieving sufficient pleasure. But punishment is ordinarily the withdrawal of comfort or the inflicting of some moderate level of pain. So, ironically, if I failed to achieve the most pleasure (highest good) I could achieve, then more pleasure should be withdrawn from me by way of penalties for not achieving enough pleasure. And that would be a very curious state of affairs, wouldn't it?