First, let me clear away some misunderstandings that are no doubt at least part of the reason some people find the view in question counterintuitive. The natural law theorist is NOT saying that you are obliged to tell the murderer where his intended victim is. In fact you are obliged not to tell him. The claim is rather that it is wrong to resort to lying, specifically, as a way of avoiding telling him. You could instead say nothing, or try to distract him, or say something that is vague or ambiguous or subtly off-topic but not untrue. You could threaten him, since he is himself threatening someone under your protection. Indeed, you can do more than threaten him if you are certain that his attempt at murder is imminent. You can punch his lights out, or even kill him if that is the only way to save your own life or that of the person you are hiding. This would be self-defense, and thus not murder. There is no question whatsoever here of your having a duty to sit back and let him do what he wants. The claim is only that it would be wrong to lie. And even if you did lie to him, the claim is not that you would have done something seriously wrong. You would be guilty of at most a venial sin, given the circumstances. So, things are hardly as dire as critics of the view might think.
Of course, many find it counterintuitive to hold that there would be even a slight moral failing in telling such a lie. But the classical natural law theorist has given a reason for thinking there is. As Aquinas says, the basic trouble with lying is that it is a kind of perversity. It takes what has as its inherent, natural end the communication of what is really in one’s mind -- speech and related behavior -- and deliberately turns it to the opposite of that end. The teaching about lying is therefore just a straightforward application of the more general moral system. For the natural law theorist has a worked-out theory of the good, grounded in a sophisticated metaphysics (classical essentialism), that explains why such an action must of necessity be bad (even if not always seriously bad). He argues that unless we accept some such account of the good, no account of goodness as an objective feature of the world, and thus no foundation for morality, would be possible at all. (See chapter 5 of Aquinas, chapter 4 of The Last Superstition, or the first half of “Classical Natural Law Theory, Property Rights, and Taxation” for the rundown.) He also claims that something like the metaphysics that leads to this account of the good is rationally unavoidable in any event. (Again, see the relevant chapters in Aquinas and The Last Superstition.) Then there is the fact that in general, natural law theory conforms to traditional morality, and thus to the common sense of most human beings historically. (Of course it does not conform to the intuitions of contemporary liberal Western academics. But relative to what most human beings think and have thought historically, their intuitions are highly idiosyncratic.) Merely to say that one finds some one small part of natural law theory counterintuitive hardly outweighs all of this.
And we should expect even strong intuitions occasionally to be mistaken. As I argued in the earlier post on intuition linked to above, intuition is only ever a rough and ready guide at best. For example, in general we reason logically, which is why when we begin the study of logic we find that it consists largely in the codification of principles we had already been following implicitly. But even the best common sense reasoning is still rough around the edges, and there are some pitfalls we are prone to if we are not careful. That is why the study of logic is necessary, and why there are such things as fallacies – patterns of reasoning that are bad but which can seem to untutored common sense not to be bad. So, while it is true that from an Aristotelian-Thomistic point of view, philosophical theory should never depart wholesale from common sense, this claim is itself something we know only as a result of philosophical theory, and that same theory tells us that common sense nevertheless needs here and there to be corrected. There is no reason to think that ethics is any different from logic in this respect.
Indeed, there are other cases where the moral intuitions even of religious and traditionally-minded people depart from what we know from natural law theory to be true. Common sense tells us that stealing is wrong, and common sense is right about that. Our intuitions reflect an inchoate understanding that private property is something good for us given our nature. But the concrete details of how a system of private property works out in actual human societies are not natural, but conventional and highly contingent. It requires intellectual effort to determine exactly what the natural law strictly requires of us vis-à-vis private property, what it recommends as best for us when possible but doesn’t strictly require, and what it leaves entirely open to custom and human law. Common sense tracks some of these nuances, but not all of them. It tells us, for example, that the starving man who takes food from a cabin in the woods has done nothing wrong. But it also often nevertheless counts this as “stealing,” which it is not, because (as the natural law theorist argues) the right that the natural law gives us over our property cannot, given its function in the overall moral system, possibly be so strong as to rule out this sort of thing. The owner of the cabin has a property right in his food, but not a right that is so absolute that it would be theft to take it under emergency circumstances of the sort in question. Because of this fuzziness in our intuitions, the starving man might feel guilty, though there is no reason for him to do so. The fuzziness can also lead to extreme positions on the part of those who think they are being “consistent” about treating property rights as natural rights grounded in natural law, such as Rothbardian libertarians (who would regard the cabin scenario as theft, and thus unjust). They are not being consistent, but simplistic and muddleheaded, because the very natural law considerations that ground the institution of private property also entail certain limits on it. (See the article on property rights linked to above for the details.)
Consider also that the intuition that lying to the murderer would not be even mildly wrong is itself historically and culturally contingent. We live in an age permeated by consequentialist thinking – so much so that even the most conservative and religious of people can convince themselves that deliberately incinerating tens of thousands of innocent people is justifiable “for the greater good,” and indeed a paradigm of moral courage. It is no surprise that in such a cultural context, lying to the murderer at the door should seem devoid of even the slightest moral stain. But other ages had other paradigms. For example:
St. Anthimus, Bishop of Nicomedia, would not allow the soldiers who were sent to arrest him, and who were enjoying his hospitality, to save him by a lie; he preferred to suffer martyrdom.
St. Firmus, Bishop of Tagasta, concealed in his house two young men, whom the emperor had unjustly condemned to death. The officers of justice came to the bishop, and demanded to be told where the young men were hidden. The prelate refused to answer; he was put to torture, but this availed nothing: “I can die,” he said, “but I cannot make others miserable.” The emperor hearing of his heroic conduct, pardoned the young men.
(Quoted from Rev. Francis Spirago, The Catechism Explained: An Exhaustive Explanation of the Catholic Religion, pp. 410-11.)
I have noted before that one finds the view that lying is intrinsically immoral defended by thinkers as different in cultural context and philosophical orientation as Aristotle and Kant. Augustine and (as we've seen) Aquinas were also uncompromising, and while there was some debate about the matter in the early Church, the view that lying is always and intrinsically immoral has been the standard view in Roman Catholic moral theology for centuries. Those to whom it seems intuitively obvious that lying to the murderer at the door would not be even slightly morally problematic should consider the possibility that their intuitions reflect their contingent cultural circumstances more than they do any innate moral understanding.
Finally, anyone who claims that it would not be even mildly immoral to lie to the murderer needs to provide some alternative account which both explains why lying is wrong in other cases but does not forbid it in the case at hand. And there are serious problems with such accounts. For example, it is sometimes suggested that it is wrong to lie only when the person lied to has a right to know the truth, which the murderer at the door does not. One problem with this suggestion is that it fails to capture what is wrong with lying per se. For we can fail to respect someone’s right to know the truth even when we don’t lie – for example, when we simply keep silent when someone who has a right to certain information from us asks for it. The view would also have the absurd implication that we can freely tell falsehoods not only to murderers, but also to innocent people who happen not to have a right to know certain truths. For example, it would entail that when there are certain secrets that a government has a right to keep from its citizens (about sensitive military operations, say), the government may not only refrain from revealing them to the citizens, but even tell outright falsehoods instead. It would entail that parents could tell falsehoods to their children, rather than merely keeping silent, about matters they are too young to understand. It would entail that rather than merely keeping silent, we can tell falsehoods to other adults about private matters we have no obligation to inform them about. It would entail that God might tell us nothing but falsehoods, since we have no rights against Him – contrary to the Thomistic view (defended in the post just linked to) that God can only ever will what is good for us, despite His not being obligated to us in any way. (Which brings to mind a further consideration: If even God cannot lie – as St. Paul famously affirms in Titus 1:2 – then where do we get off thinking that we may sometimes do so? Job 13:7 indicates that it would be wrong to lie even for the purpose of defending God’s honor.)
In short, I would say that the natural law position that lying is intrinsically wrong has powerful arguments in its favor, and when rightly understood is not as counterintuitive as it might seem. (Nor have I by any means exhausted the subject here. Natural law theorists have had much to say about the legitimacy in many contexts of broad mental reservations, evasiveness, stratagems during wartime, and the like.) Meanwhile, the alternative view has no good arguments in its favor, and is at best supported by culturally contingent and fallible intuitions. Thus, it poses no serious challenge to the natural law position. If there is a conflict between that position and our intuitions, it is the intuitions that have to go.