The principle "It is always wrong directly and intentionally to take the life of an innocent human being," is often cited; I think it sums up the fifth commandment rather nicely, at least as regards being able to determine why certain things are wrong. For example, both abortion and euthanasia are always wrong because in both the life of an innocent human being is directly and intentionally taken--but if an ectopic pregnancy must be removed, or if an elderly person undergoes surgery which unfortunately fails to provide healing but instead hastens death, in neither case is the death of the innocent person directly intended.
But an additional point is provided in CCC 2269, under the discussion of homicide: "The fifth commandment forbids doing anything with the intention of indirectly bringing about a person's death. The moral law prohibits exposing someone to mortal danger without grave reason, as well as refusing assistance to a person in danger."
So, how is it possible that the Catechism spells out so clearly that it is always wrong directly or indirectly to take the life of an innocent human being--and yet the Catechism allows for the possibility that a person may kill another in various circumstances without incurring any moral penalty?
The key word in the discussion of both direct and intentional killing and and indirect, intentional killing is, I believe, the word "innocent." There is no circumstance where it might be morally permissible to kill, or to allow to die through indirect action willed toward the person's death, an innocent human being.
In the situation where a person is an aggressor, threatening the life of another, the principle of self-defense may permit the use of lethal force to stop the aggressor. The Catechism puts it this way:
2263 The legitimate defense of persons and societies is not an exception to the prohibition against the murder of the innocent that constitutes intentional killing. "The act of self-defense can have a double effect: the preservation of one's own life; and the killing of the aggressor. . . . The one is intended, the other is not." (65)
Then in the discussion about war (in which the "just war" doctrine's principles are restated) and in the discussion about the death penalty, these issues, too, are framed as matters of self-defense, where the "self" in question includes one's nation in the event of war, or of society as a whole in the event that society can't defend the innocent without executing the criminal.
I find this interesting, because not only is the word "innocent" important, but so is the word "intentional." In other words, one may defend oneself (and by extension one's country and one's society) from unjust aggressors using force up to and including lethal force, but only, it would appear, because the principle of double-effect applies, in that the killing of the aggressor is the unintentional, secondary effect and the preservation of innocent life is the primary one.
But because we are talking about a double-effect scenario, it follows that not only may the death of the aggressor not be sought for its own sake, nor desired as the outcome, nor held to be a good, but also that the death of the aggressor is not morally permissible in circumstances where a lesser restraint against him which does not cause his death is possible. What does that mean?
For starters, it would mean that something like this would be completely impossible from a moral perspective. Anwar Al-Awlaki may be plotting against the United States, and there may be a completely legitimate need to stop him from doing this. But plotting acts of aggression and actively carrying out such acts are two different things; it may be necessary to shoot a terrorist who is attempting to walk into a crowded restaurant while wearing a bomb because he is actively attempting to kill innocent people, but it would not be necessary to place snipers on the roof of the terrorist's apartment building because intelligence suggests that he might be considering such an act. There are plenty of ways to stop him from carrying out his planned attack without necessitating his death, and the moral law is clear: when non-lethal means of self-defense remain an option, the defender has an obligation to make use of those means before resorting to lethal ones, which must be considered a last resort.
It is true that in many self-defense scenarios, a person seeking to preserve his life may not have the ability to assess the situation calmly and correctly before using lethal means. This is true in home-invasion situations, on the battlefield, and in other situations where it simply isn't possible to tell whether non-lethal means will work to stop the aggressor before resorting to potentially lethal ones. But no one has the right to seek to kill an aggressor intentionally--and the duty to assess the situation and come up with non-lethal means to stop the aggression grows the further removed from a direct attack the situation becomes.
I bring up all of this because it is sometimes said, "We have the right to kill someone in self-defense. If we can kill him, we can certainly torture him for the same reason." But as the Catechism makes clear, we do not have a right to kill anyone, in self-defense or otherwise. We have the right to protect ourselves from an aggressor, up to and including the use of lethal force. But it is protection, not the death of the aggressor, which is willed; the death of the aggressor is the unintended and undesired consequence in this double-effect case.
And there is no possible way that it can become morally right to torture someone on the grounds of self-defense, because torture is always a direct and intentional act. Attempts to frame the use of torture in double-effect terms will always fail, because the protection of the innocent does not permit the defender to inflict pain and suffering, directly and intentionally, on the human being who is not actively threatening him but is instead wholly in his power and at his mercy.