Respect (reverentia) is likewise something merely subjective; a feeling of a peculiar kind not a judgement about an object which it would be a duty to effect or to advance.  For if considered as duty it could only be conceived as such by means of the respect which we have for it.  To have a duty to this, therefore, would be as much as to say to be bound in duty to have a duty.  When, therefore, it is said:  “Man has a duty of self-esteem,” this is improperly stated, and we ought rather to say:  “The law within him inevitably forces from him respect for his own being, and this feeling (which is of a peculiar kind) is a basis of certain duties, that is, of certain actions which may be consistent with his duty to himself.”  But we cannot say that he has a duty of respect for himself; for he must have respect for the law within himself, in order to be able to conceive duty at all.

XIII.  General Principles of the Metaphysics of Morals in the

{Introduction ^paragraph 160}

treatment of Pure Ethics

First.  A duty can have only a single ground of obligation; and if two or more proof of it are adduced, this is a certain mark that either no valid proof has yet been given, or that there are several distinct duties which have been regarded as one.

For all moral proofs, being philosophical, can only be drawn by means of rational knowledge from concepts, not like mathematics, through the construction of concepts.  The latter science admits a variety of proofs of one and the same theorem; because in intuition a priori there may be several properties of an object, all of which lead back to the very same principle.  If, for instance, to prove the duty of veracity, an argument is drawn first from the harm that a lie causes to other men; another from the worthlessness of a liar and the violation of his own self-respect, what is proved in the former argument is a duty of benevolence, not of veracity, that is to say, not the duty which required to be proved, but a different one.  Now, if, in giving a variety of proof for one and the same theorem, we flatter ourselves that the multitude of reasons will compensate the lack of weight in each taken separately, this is a very unphilosophical resource, since it betrays trickery and dishonesty; for several insufficient proofs placed beside one another do not produce certainty, nor even probability.  They should advance as reason and consequence in a series, up to the sufficient reason, and it is only in this way that they can have the force of proof.  Yet the former is the usual device of the rhetorician.