The Deontic Transfer Principle


The Deontic Transfer Principle states that if it is permissible for a person A to cause another person B harm H then, other things being equal, it is permissible for A to impose a risk of harm H on B. In this article we show that the Deontic Transfer Principle is vulnerable to counterexamples, and that the same is true of a range of closely related principles. We conclude that the deontic status of a risk imposition is not directly inherited from the deontic properties of deterministic acts.

This is a preview of subscription content, log in to check access.


  1. 1.

    To harm someone is to impose a risk of harm that occurs with probability 1. The first set of acts is therefore a strict subset of the second, meaning that there are (many) more acts of the latter type.

  2. 2.

    What we call the Deontic Transfer Principle should not be mixed up with the Deontic Inheritance Principle in deontic logic. The latter holds that if (p → q), then O(p) → O(q), where O(p) means that it is obligatory that p. We will return to this principle in note 17.

  3. 3.

    They seem to have done so independently of each other.

  4. 4.

    See, for instance, McCarthy (1997: 205), who seems to endorse the Deontic Transfer Principle: “It would be very surprising if facts about the morality of imposing risks of harms did not connect importantly with the morality of harming”.

  5. 5.

    We leave it open what type of harm it would be permissible to inflict on the dictator.

  6. 6.

    In Standard Deontic Logic necessary propositions are obligatory. The exhaustive disjunction of all available alternatives is therefore obligatory, which is the reason why we exclude this case in the formulation of the principle.

  7. 7.

    We would like to thank Karsten Klint Jensen for raising this objection.

  8. 8.

    We are grateful to an anonymous reviewer for suggesting this line of response.

  9. 9.

    Perhaps the rifle has a special “discharge responsibility mode” which randomly fires at one of two previously selected targets once the trigger is pulled.

  10. 10.

    This proposal can be regarded as a specification of the ceteris-paribus-clause in previous formulations of the deontic transfer principles. We are grateful to Klaus Steigleder for suggesting a similar version of the principle to us.

  11. 11.

    This principle has been suggested by Ibo van de Poel.

  12. 12.

    E.g. within a maximizing act-consequentialist framework, this is obvious: Suppose that A is one of the actions with the best consequences within the set of options {A,…, Z}; then A is permissible. But if A* has strictly better consequences than A, then A will be impermissible for an agent choosing among {A,…, Z, A*}.

  13. 13.

    One might think that this is rather unsurprising since deontic properties fail to propagate in other domains as well. For instance, deontic properties do not easily transfer from ends to necessary means, as illustrated by the Professor Procrastinate case. Professor Procrastinate ought to accept the invitation to review a paper on which he is a leading expert and write the review. Accepting the invitation is a necessary means for doing what he ought to do. However, as the example goes, Professor Procrastinate will not write the review once he accepts the invitation, so it seems wrong to conclude that he ought to accept the invitation in the first place. It would have been better to decline the invitation and let someone else write the review. This can be understood as a counterexample to the Deontic Inheritance Principle mentioned in footnote 2: \({\text{A[ccept]}} {\land} {\text{W[rite]}}\rightarrow{\text{A[ccept]}}\), but it is not true that \({\text{O}}({\text{A}}{\land}{\text{W}})\rightarrow{\text{O}}{\text{A}}\). We of course agree that our discussion can to some extent be understood as reinforcing a difficulty that is already well known in other domains, but it is worth keeping in mind that the Deontic Transfer Principle concerns a different domain. Influential scholars do actually claim that the principle to be true in this domain, so our arguments should be surprising to at least some thinkers. It is also worth noting that the failure of deontic properties to propagate from determinate to risky contexts is not due to their failure to propagate from ends to necessary means, or from some proposition to what is entailed by it. There is no such means-end-relation or logical entailment in the cases we consider. In particular, it is conceptually possible (but perhaps not very plausible) to reject the Deontic Transfer Principle but accept the Deontic Inheritance Principle in deontic logic. In that sense, the truth or falsity of the Deontic Transfer Principle does not depend on the Deontic Inheritance Principle.

  14. 14.

    Cf. Hansson (2010, 587).


  1. Hansson, S. O. (2010). The harmful influence of decision theory on ethics. Ethical Theory and Moral Practice, 13(5), 585–593.

    Article  Google Scholar 

  2. Lackey, D. P. (1986). Taking risk seriously. Journal of Philosophy, 83(11), 633–640.

    Article  Google Scholar 

  3. McCarthy, D. (1997). Rights, explanation, and risks. Ethics, 107(2), 205–225.

    Article  Google Scholar 

  4. Thomson, J. J. (Eds.). (1986). Imposing risks. In Rights, Restitution, and Risk. Essays in Moral Theory (pp. 174–191). Cambridge, MA: Harvard University Press.

Download references

Author information



Corresponding author

Correspondence to Christian Seidel.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Peterson, M., Seidel, C. The Deontic Transfer Principle. Erkenn (2019).

Download citation