(e.g. desires could be consistent but deontology prevents the desires from being acted upon)
The question is why the deontological concerns are motivating. If they are motivating though a desire to fulfill deontological concern, then they belong in the utility function. And if not through desire, then how? An endorsed deontological principle might say ‘X!’ or ‘Don’t X’, but why obey it? Deontological principles aren’t obviously intrinsically motivating (in the way anything desired is).
The question is why the deontological concerns are motivating. If they are motivating though a desire to fulfill deontological concern, then they belong in the utility function. And if not through desire, then how? An endorsed deontological principle might say ‘X!’ or ‘Don’t X’, but why obey it? Deontological principles aren’t obviously intrinsically motivating (in the way anything desired is).
Following deontological concerns can be instrumentally useful for biased finite agents: http://wiki.lesswrong.com/wiki/Ethical_injunction