As is well known, Hume's guillotine forbids a derivation of ought from is. But that does not mean that it cannot be derived at all.

One should first distinguish between a morally non-binding ought and a morally binding one.

Then you have to say that you can derive an ought from a will, to put it more precisely, from another will, a will different from one's own.

The foreign will is to be understood as a request, claim, demand or command, all of which only another person can do.

Since I am German, my reasoning comes from the logic and semantics of the German verb "sollen", which translates to ought or should in English.

Almost all grammars of German explain "sollen" or ought by saying that there must be another personal agent who wants something from you. So, if A ought to do X, this implies that someone wants A to do X.

For example: I go to the doctor and he tells me to take pills twice a day. Then I later tell my wife I ought to take pills twice a day. It is important to add that it is not just a recommendation of the doctor, it is already more, so more binding.

The question that now arises is, how can I distinguish morally binding ought from morally non-binding ought? Because not everything that people demand of me is really binding.

Let's take the concrete example of when a person is lying on the street seriously injured and is just unable to help herself. But she calls out to me to help her, as I am just in the visible vicinity.

Here, any normal and reasonable person would say that there is a moral ought in me. To whom can this particular ought be traced?

1. To the injured person, who has nothing left at the moment but her being an end in herself. This would be the position of the German philosopher Gerold Prauss, who says the following according to my rough translation: "Thus, ought is nothing but the synthetic result of the encounter of subjects, each of which is a knowing end in itself in relation to the other, a knowing willing and thus also a knowing demanding." It proves for Prauss to be the case that the order of morality in general, is conditioned by a very peculiar situation: a claim to morality is grounded only when the person handled is precisely not in a position to help herself and as long as she remains in this position. This self-help represents then the decisive criterion. When we do not help someone in need, we do not solely prove to be non-meritorious but we commit an evil, whatever our maxims might be.

2. Or to society or God, who additionally demand something from me or give a binding note to the demand of the injured.

If I bring in God or society, my ought could be reduced to a have to or must.

Here is the logical semantic scheme for the must:

(1) A must do X for Y to happen, and
(2) A wants Y to happen.
______________
(3) So A must do X.

In concrete terms, this means that I do not want to be punished by God or society, so I must help the injured person.

So not only ought, but also must can be normative. Here is a quote from the German action theorist Peter Stemmer, roughly translated again by me:

"To express a normative situation linguistically, we usually use a "must" sentence: "You must, in order to catch the 10 o'clock train, leave now." With these formulations, we might say, strictly speaking, only a must of a necessary condition is brought into language. But it is presupposed in each case that the corresponding wanting/willing is present. It would be pointless to tell someone who does not want to catch the 10 o'clock train that he must leave now in order to catch the train. That he wants to do so is assumed in the situation, and thus it is presupposed that the must is normative."