I can't resist posting this. The following principle seems plausible:
- You cannot attempt to kill someone whom you know to be dead.
Suppose you know that your great grandfather died between 1900 and 1910, but you do not know when. (You know your grandfather was conceived before 1900.) Moreover, a very complicated inheritance case turns on the question. It turns out that if he died before 1905, you can claim a large inheritance. Now, you happen to run into a Nobel-prize winning physicist, Dr. Mort, who is known for brilliance and honesty, and shows you an odd machine. Dr. Mort tells you that the machine is a Temporally Reversed Death Ray. You specify a past date and enter identifying information, and the Temporally Reversed Death Ray kills the person on that date. Oddly (you suspect this may have something to do with the Grandfather Paradox), she tells you that the Death Ray only works on victims past reproductive age. All this sounds really dubious, but, hey, Dr. Mort is reputed for brilliance and honesty, and who are you to doubt her word. So, you enter your great, great grandfather's identifying information, reflect that in 1902 your great, great grandfather would be past reproductive age, set the date to January 1, 1902, and press the big red button.
What have you done? Well, you have attempted to kill your great great grandfather, that's what. You have attempted to do this whether or not the machine works. You have attempted to do this even if it turns out that it is logically impossible for the machine to work. As long as you believed that the machine would work, you tried to kill your great great grandfather.
I wonder what the law would say. Normally, the law does not convict people of attempted murder if the method used was one that no reasonable person would expect to work—for instance, one is not going to be convicted of attempted murder if one attempts to kill by magic. That limitation is reasonable in law, but morally speaking sure the person who attempts to kill by magic is an attempted murderer. Moreover, given Dr. Mort's reputation, it seems that a reasonable person might believe her. And if the word of one physicist is not enough, we could imagine that she gets ten others of equally good reputation to speak, too. (Maybe one could find eleven Nobel-prize winning physicists of high reputation to do this as a joke. It is at least possible.)