When considering the derivation of moral principles from a naturalistic point of view, it would seem to me that moral statements can only be conceived of in one way. That is, if we are trying to make a reasoned argument for why a moral principle ought/ought-not to be performed, you must argue about the utility/end of that action.
One ought not to steal because you run the risk of being arrested.
So if the outcome in this case ("risk of being arrested") is not preferred, then one ought not to steal. In fact, I see this as the potential solution to any is-ought problem: one ought to do something because the outcome is favorable - the outcome leads to some defined end that is preferable. (The nuance is of course deciding what is the most favorable action in a particular circumstance - you can only make the best decision given the information you have).
If someone states:
Killing is wrong, it seems to me that they are really saying "I dislike killing", not that it is morally reprehensible. In this way, I do not see that they are really making a reasoned argument, they are simply stating their personal preference. To turn their scenario into one of utility I would say: "I have a preference against killing because it would bring upon an undesirable circumstance: a) It would lead to psychological disturbance on my part and/or b) it would lead to my potential arrest. Since neither of these options are preferable, I will state that I ought not to do it".
For something less extreme, the statement
one should not lie could be unpacked into a consequence-based scenario: "I interact in a stable community of individuals whose successful cooperation is based on the principle of trust. I have observed in many scenarios that lying, generally, ends up causing further issues and may bring distrust onto me from the community. To maintain this social order for myself and others, I ought not to lie." Is lying permitted under certain circumstances? One could argue that a lie may protect someone (a consequence).
In this way, I could conceive of many moral systems in use today as being defined on this basis even if they differ in the particulars. The a priori assumption is that the moral principles chosen in the particular social context would maximize day-to-day predictability, structure/order, well-being, and facilitation of ideas. And for this reason, they have become the predominant set of standards followed by that group. I suppose we could call this a form of relativism; I'm not particularly sure.
I am having difficulty seeing other ways of conceiving of the justification of moral principles outside of this. The notion of certain things being an end-in-itself does not seem rationally justified and also seems subjective. So I can't see how someone might argue:
You ought to perform good deeds because they are good by definition. This is not a reasoned argument to me and it seems to run into other problems such as how one defines "good" in the first place.
I can conceive of many moral questions being posed in this way. What are some counterexamples against this position? One counterexample I foresee is that one can never know for certain the outcome of one's actions in all scenarios. This might be true, but can't one reasonably be certain that various patterns of actions under probable circumstance can lead to a favorable outcome? Furthermore, even if one were following moral principles for another reason, they cannot be sure that their "good" action could have lead to a good outcome.
Perhaps I am being too simplistic here or I am abusing the definitions of various terminology. I welcome any counterexamples to help clarify my position.