In "Justice as a Larger Loyalty" Richard Rorty argues that Justice is simply a sub-category of loyalty. Skipping his main argument, I started thinking about his examples and how he went about discussing the issue.
The problem with hypothetical examples for moral dilemmas is that they invariably oversimplify the situation. It trying to highlight the dilemma, they posit a few facts, and then ask the reader to consider what they would do. But what any of us would do is ultimately dependent on a far greater range of data than the hypothetical example can provide. And, of course, it also tends to assume that our actions are largely the result of our moral reasoning, something that should not be taken for granted.
Two examples in Rorty’s essay include that dilemma of a family after a nuclear holocaust who now shoot their neighbors to preserver their own dwindling food supplies, and the classic lifeboat dilemma. As to the example of the family after a nuclear holocaust, the problem is can be considered from several angles. For one, everyone is going to die anyway, so why worry about whether you can feed your family for two extra days. This is the real problem with hypotheticals: as models they are always incomplete. There are always factors that the model excludes. For example, someone once asked me (knowing I am married), “If you could sleep with another woman and no one would ever find out, would you do it?” The expected answer is, “… well, if no one would ever know…” However, there is another dimension to the issue. I would know, even if no one else ever did. And this is not insignificant. So back to the nuclear holocaust example: food is limited and family and neighbors all want to eat. The choices are: fight off the neighbors in the name of family, or share with all and run out sooner. Given the bleak situation, I question whether the moral course is to put a black spot on your soul by killing your neighbors for the sake of your kids, or to go out with a clean conscience, knowing you did best for everyone. The radiation will probably get you before the hunger.
Move the scenario to a life boat on the open ocean, and the problem of hypoteticals again shows up. There are simply too many unknowns. You don’t know how long until you will be rescued, if you will be rescued, whether you will be able to catch fish, and where ocean currents will take you. And anyway, dehydration is a bigger problem than malnutrition. To kill off the strangers so as not to have to share food and thus provide better for your family sets a terrible moral example for them, and they will have to witness the deed. Then if you all survive, you will never know if you could have done things differently and brought everyone through.
Finally, as a principle, looking out for those closes to you does not create a society I or anyone else would want to live in. So if it doesn’t serve well under normal circumstances, how is it any better in extraordinary ones? Once you accept the expediency excuse (desperate circumstances call for desperate measures) then you have the slippery slope problem: when are circumstances desperate? Nuclear holocaust is remote, but what about unemployment? Is that desperate enough to justify stealing food (my family needs it more than the store proprietor)? Where is the line?
Rorty, Richard. "Justice as a Larger Loyalty." Justice and Democracy: Cross-Cultural Perspectives. Ed. Ron Bontekoe. Honolulu, HI: U of Hawaii P, 1997. 9-22.