You may have noticed how easy it is to slide from “there is such a thing as human nature” to “humans naturally behave in a certain way”. These statements seem semantically equivalent, but for many people, to say that something is “natural” is almost the same as saying that it is virtuous (see the Appeal to Nature.) In this worldview, saying that “selfishness is natural” is a logical contradiction. It cannot be true.
Philosophers have long noted the difference between saying that this is the way things are (these are the facts) and this is what we should do (see the Is and Ought problem). But is it entirely unsound to move from is to ought? If we can't use what we know about the world, about ourselves, and about how we got this way, then we have no rational basis at all for moral argument.
Consider the observation that the human hand happens to be peculiarly well-formed to be used as a weapon (by making a fist, see Fighting Shaped Human Hands). This has the obvious evolutionary explanation that at some point in our history there was a significant advantage to having a powerful punch. People who could form a closed fist where the thumb buttressed the fist were more successful at surviving and reproducing than those who (like non-human primates) couldn't punch properly.
You might question the validity of this sort of reasoning in general (see Just-So Stories) or based on the details of this specific claim, but for the moment suppose that you accept this as true. Where does this leave the moral status of punching people? “Your honor, people of the jury, this man's hand was made for punching.”
This is ridiculous as a legal defense of aggressive behavior, and is equally suspect as a moral argument. Yet to deny the importance of violence in the formative environments for human nature is equally unreasonable. A capacity for violence and an inclination to be violent when our interests are threatened is part of our behavioral legacy. The appeal of taking what you want by force is also sufficiently obvious that it can be rediscovered by each generation of toddlers.
See Evolutionary Ethics.
The is/ought distinction should be seen more as a statement of the limits of philosophical ethics than about what it means to be good. Narrowly, the is/ought claim is Skepticism applied to ethics. The same problem also exists in many situations that are not moral dilemmas, and that may have little or no moral overtones at all. We say things like: “I should eat healthier foods” or “You ought to ask for that promotion.” These are about self control and appropriate motivation, but don't have to do with the sorts of social conflicts that we have evolved moral judgement to regulate.
So Hume's criticism of the jump from “is” to “ought” applies to any claim about “how things should be”, what is “the good life”, or any judgment about a desire or intention.
Do we need to take this challenge to “should” stories seriously? In philosophical ethics, one major response had been is to come up with a fundamental “ought” that is broad enough to be the foundation for all the “ought” claims we would like to make (see Consequentialism). In the real world, moral argument is of course little changed from Hume's time, though there is somewhat less appeal to divine authority. Instead we support our position by direct appeal to values (which we expect the listener has in common) together with factual and hypothetical claims which we feel underlie our judgment. We can put a rational gloss on this by relying on shared normative understandings of the human condition, and some pro-social mythology. See Norms and Values.
We think the real meaning of the “is/ought” divide becomes much clearer when you realize it applies to any statement about desire, intention or motivation that goes beyond simple description. We can say that: “Joe wants to go to Hawaii.” We may have reasons to believe this is true. Perhaps he said he wanted to go, or he's always talking about the last time he went. But we can't so easily make a solid argument that he should (or shouldn't) want to go.
What makes the “is” and “ought” modes of thought subjectively distinct is Representational Opacity. We don't have access to all the considerations we unconsciously weigh, so we have to make up stories that ring true against our intuition. This dual process view does say we have judgments for reasons, but that we don't reliably know the reasons.
What most clearly distinguishes “should” statements is that they provide explicit guidance about behavior. With a statement of fact: “the cup broke”, or a causal generalization: “cups break when they fall on the floor” we will never get to “you should be careful with your cup” unless we fill in the gaps with our understanding that broken cups are bad.
What the “is/ought” argument says is that we can never get to a statement of the correct way to behave by observing what happens in the world, or even by learning how the world works well enough to predict what will happen (the goal of science). Is this actually true?
We believe that if you include humanity in “the world”, then the argument starts to break down. People do tend to behave in ways that promote their own individual flourishing, and that of their society. Societies do have ways of making sure that individuals don't excessively favor their own interest, harming the group. This covers a lot of the territory of common should statements. We survive in the face of competition for limited resources, with other individuals and groups always ready to step in if we don't show up one day.
Well, is merely surviving, or even flourishing, the correct goal? That gets us back to the traditional gridlock of philosophical ethics. In the ordinary daily world of moral argument, we take survival as a given, and mostly debate what it would mean to flourish and how to move in that direction.
In the world surrounding the ivory tower we do indeed see people and societies behaving in ways that are roughly consistent with survival, and this has to be true because otherwise these examples wouldn't be here. Our scope for individual and political freedom lies within this constraint. Not only does a society have to survive in isolation, it has to survive in the presence of other societies with different values. Even if we rule out war, cultures still compete for mind share and societies still compete for citizens. If a society fails to provide what its people want, then they may either leave or adopt ideas from another culture.
Evolutionary psychology adds the prediction that our judgments, our motivations, and so on, will be adaptive. This is why we believe you should get out of bed in the morning: not because reason demands it, but because those who never got out of bed didn't leave any descendants. This wisdom of evolution is embedded in our minds, but hidden from us by Intentional Opacity. We can see that an organism that wishes to survive and prosper must behave in certain ways, but that adaptive intention is mediated by other psychological processes, such as motivation and desire toward goals which might turn out to be important intermediate steps.
What if humans happened to be non-social animals? The stories that we would recruit our reason to support would be quite different. Consider the biography of an alien ambush predator philosopher. They are egg layers, which gets rid of the unpleasantly social/cooperative aspect of internal fertilization. Then their entire life could be alone, never meeting another member of their species except in a territorial dispute. Reason would demand it, of course. All the complexities of social cooperation are gone. No free rider problem. Unfortunately such a beast would have no use for language, so would scarcely be recognizable as a conscious being.
Perhaps the is/ought problem is a contingent fact of the evolved construction of the human mind. If our thought processes were entirely accessible to us, then would we have no need for “ought”? Certainly many philosophers have imagined that a being could be purely rational, and moreover that we have (or should aspire to) such rationality. Homo philosophicus might engage reason to decide whether to get out of bed. Is this even credible? Of course Homo philosophicus still can't escape the adaptive logic. Either some philosophers reliably come up with reasons to get out of bed, or find a way to live entirely in bed, or the species dies out, and is not available as an example.
The thing is, it works just as well to get out of bed “because the early bird gets the worm” as because “blah blah whatever rational blah blah”. So unless some divine force prevents H. Philosophicus from abandoning perfect reason it is likely the species will evolve various irrational shortcuts to the morning debate. They will usually get out of bed without thinking, and will find that they care about life.
Caring (in the emotional-motivational sense) is an intentionally opaque motivational shortcut. We could come up with reasons to care, but we could come up with reasons for apathy too. Yet beings who don't act as though they care about life won't stay around long enough to for us to study them. This is the motivational anthropic principle. While motivations are in a sense arbitrary, in practice we don't observe the full range of conceivable motivations. The evolutionary argument is that it doesn't matter what causes our behavior, whether it is innate or learned. Cultures with maladaptive norms will not survive. Behavior must needs be reasonably adaptive.
Of course many philosophers see our subjective values as being at the core of the human condition: our sense that we care, and that something matters. We do care, but it's hard to escape the evolutionary logic that the ultimate cause for our subjective world is no more (and no less) than survival. Evolutionary psychology argues that our entire behavior (unconscious included) is “rational” (understandable) from the God's eye view. We just don't have access to the reasons, and are also inhibited by moral standards from acknowledging the pervasiveness of opportunistic behavior.
You can come up with some stories about why reason demands we get out of bed: that we be productive, support our families, aid our fellow man, but even ignoring the technical is/ought problem, to someone who is depressed, these arguments are impotent. If you are entirely lacking in motivational foundation, then no story can make you care, since there is no subjective basis for the story to connect with.
Is/ought is not just a philosophical problem, it's a problem with any argument about what to do when there is no standard of truth or causality, only rhetoric. Political argument faces the same theoretical difficulty. The thing is, this isn't where actual political argument falls down. With normal non-depressed philosophically-innocent people, it is very difficult to convince them that there is anything arbitrary or non-obvious about wanting the good things. The real problem with politics is the lack of a connection to a quantitative casual theory. When an engineer says “we should make this thicker”, not only do his peers understand that bridges not falling down is good (the is/ought problem), they also agree on ways of deciding what casual mechanisms are important and how to make cost/benefit tradeoffs. Though they are on much shaker ground, policy wonks and economists think they are developing a similar discipline of political engineering. See Evolutionary Politics.