Agency and Anthropology in Kant’s Groundwork

  • Onora S. O’Neill
Part of the International Archives of the History of Ideas / Archives Internationales d’Histoire des Idées book series (ARCH, volume 128)


Recently Bernard Williams described Kant’s Groundwork of the Metaphysic of Morals as “the most significant work of moral philosophy after Aristotle, and one of the most puzzling.”1 This seems to me right on both counts. The puzzles would be fewer if we had a strategy for reading the entire Groundwork and making sense of its multiple transitions between distinct frameworks of argument. One central strategic task is to make sense of Kant’s account of action and agency. He discusses agency under various headings — including willing, rational willing, and freedom — throughout the work. The discussion is complex not only because of this variety of headings, but still more because it is conducted in three distinct frameworks, and because the transitions between these coincide only approximately with the transitions between chapters. The first chapter analyses our common account of morality and claims that it centres on the idea of good willing; the second chapter undertakes a metaphysics of morals by analysing good willing in terms of an abstract account of rational willing as such; the last chapter (in the sections from 450 onwards) argues from the perspective of a critique of reason that rational willing is free and yet compatible with the causality of nature.


Human Agency Categorical Imperative Free Agent Causal Judgement Abstract Account 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Ethics and the Limits of Philosophy (Fontana, 1985), p. 55.Google Scholar
  2. 2.
    This suggests that we must take the remark that moral philosophy “must not borrow from acquaintance with man” (G 4: 412) to mean that it must not draw on anthropology to determine its own principles, but that reference to specific circumstances and forms of agency is needed in actual deliberation. Cf. Maximilian Forschner, “Reine Morallehre and Anthropologie,” Neue Hefte für Philosophie 22 (1983): 25–44.Google Scholar
  3. 3.
    See, for example, his discussion of our lack of insight into the fundamental maxims, Religion within the Limits of Reason Alone,Book I.Google Scholar
  4. 4.
    See the Introduction to the Metaphysic of Morals for a systematic account of the spectrum of agency.Google Scholar
  5. 5.
    Susan Mendus, “The Practical and the Pathological,” Journal of Value Inquiry 19 (1985): 235–43, shows why this is an unconvincing reading.Google Scholar
  6. 6.
    It is less clear to me that he would deny a Humean view of desires (motives) as causes of action; he might regard this as innocuous (if beside the point) because a Humean causal claim asserts no necessary connection between desire and the resulting action.Google Scholar
  7. 7.
    From this alone we can tell that the Categorical Imperative underdetermines action, for if it provided a principle for generating a required action in each situation the thought experiment of a being in whom desire never trumps reason would be empty, since there would be no content to the thought that such a being had desires.Google Scholar
  8. 8.
    Cf. discussions in J. Bennett, Kant’s Dialectic (Cambridge University Press, 1974), p. 193; R.C.S. Walker, Kant (London: Routledge & Kegan Paul, 1982); Bernard Williams, “Morality and the Emotions,” Problems of the Self (Cambridge University Press, 1973 ), p. 228; Allen W. Wood, “Kant’s Compatibilism,” in Self and Nature in Kant’s Philosophy, ed. Allen W. Wood (Cornell University Press, 1984 ).Google Scholar
  9. 9.
    This suggests a serious defect in the thought experiment of the Holy Will. It may be stated as a dilemma. If Holy Wills are embodied beings, they too are part of nature and subject to alien causes, so (contrary to hypothesis) experience desire and temptation and must view moral principles as imperatives. If Holy Wills are not embodied beings, so not part of nature or subject to alien causes (as the text suggests), then it is not clear how they can distinguish those sequences that come about from those that they bring about. Mere conformity to their (disembodied) will cannot show that a change is their doing. Hence it does not seem that Holy Wills can view themselves as agents of any sort.Google Scholar

Copyright information

© Springer Science+Business Media Dordrecht 1989

Authors and Affiliations

  • Onora S. O’Neill
    • 1
  1. 1.University of EssexUK

Personalised recommendations