Respect for others can be generalized to a respect for complexity and our limitations in dealing with it. When we think of a situation in fewer variables than it has in real life, we risk unintended consequences which may be harmful to people in that situation (committing evil).
An excellent example of this is the U.S. approach to terrorism. President Bush seems to think of people as “good” or “evil,” identifies groups in these terms, and casts every human activity as a conflict between such groups. There are at least two problems with this perspective. First, there are no good or evil people, just good or evil actions (actions which help or hurt people). Second, behavioral traits are built into the entire human population; they cannot be stamped out by destroying the people who demonstrate them.
Whether or not people will use fear as a tool to achieve political objectives (practice terrorism) depends on, among other things: their personality; the availability of (and their familiarity with, and confidence in) alternative approaches; and their degree of desperation. We can control conditions and awareness, but not inheritance (genetic engineering notwithstanding). Culture (including religion) has been proposed as a way to screen people who might practice terrorism; but it only preconditions them, and not uniformly.
President Bush and others who would objectify people in order to protect themselves would do well to consider the fundamental sources of historical success in preventing violence: screening of behavior, not people; education; and reduction of poverty and persecution. When you screen behavior (through laws and law enforcement), you discourage the actions themselves, forcing people to consider acceptable alternatives. When you educate people, you give them the tools to find those alternatives. And when you reduce poverty and persecution, you reduce the desperation that drives people to act before they adequately consider the consequences of their actions.