Contribution
The lectures recast the foundations of ethics as a first-person question, the normative question, and give a historical sequence of answers: voluntarism (Hobbes, Pufendorf), realism (Clarke, Price, Prichard, Moore, Nagel), reflective endorsement (Hutcheson, Hume, Williams) and autonomy (Kant). Their own contribution is the concept of practical identity, which turns Kant’s autonomy argument into an account of where all reasons and obligations, moral or not, come from. Their conclusion is that all four answers are “true”, once each is placed as an aspect of the relation of the reflective self to itself.
Key insight
A being that can step back from its impulses cannot act on them as such. It must decide whether they are reasons, and it can do that only by reference to some law it takes as its own. Taking a law as one’s own is having a conception of who one is. So identity, in a practical sense, is not a metaphysical fact about a person that ethics must respect. It is a precondition of acting at all, and obligation is what that identity forbids. On this view the threat of losing oneself (in “I couldn’t live with myself if I did that”) is the form obligation takes.
Assumptions
- Reflective structure, not luminosity. “The human mind is self-conscious in the sense that it is essentially reflective”. It is not internally transparent: “our knowledge of our own mental states … is no more certain than anything else.”
- Acting is acting for a reason, and reasons come from principles. Kant’s premise: a free will “must have its own law or principle”, and “since reasons are derived from principles, the free will must have a principle”.
- The categorical imperative is the law of a free will, as Kant argues in Groundwork III. This argument is summarised from her paper “Morality as Freedom”, not given in full.
- A distinction Kant does not make. The categorical imperative (act on a maxim you can will as law) is not yet the moral law (the law of the Kingdom of Ends). The step between them is made by practical identity.
- Wittgenstein’s private-language argument, read normatively: meaning is a relation between a legislator and an obeyer.
- Aristotle’s account of living things as self-maintaining forms, used for the argument about pain and animals.
Key results
Lecture by lecture:
- Lecture I. Voluntarism makes obligation depend on a legislator’s authority, and authority cannot be reduced to power or to prior normativity without a regress or circle. Realism stops the regress “by fiat”, like the cosmological argument. Procedural realism (there are correct procedures for moral questions) is separated from substantive realism (there are moral facts the procedures track). Only the second is rejected.
- Lecture II. Hume’s “normativity as reflexivity”: a faculty’s verdicts are normative if the faculty, reflecting on itself, approves itself. The understanding fails the test and the moral sense passes. Williams justifies values by their contribution to flourishing. But reflection applied to particular motives, as in the case of the knavish lawyer, drives the reflective endorsement method to Kant: “The test of reflective endorsement … is morality itself.”
- Lecture III. Autonomy is the source of obligation. Practical identity is introduced, and there is “a jumble of such conceptions”. Obligation is unconditional when violating an identity would be worse than death. Good maxims are “intrinsically normative entities” in virtue of their form, so “realism is true after all”, in a procedural sense. Relativism has limits: valuing any identity requires valuing humanity, so “Enlightenment morality is true”. Reasons are public, so others can obligate us. Pain is “the perception of a reason”, so “we have obligations to animals”.
Claims
| id | claim | strength | support |
|---|---|---|---|
| C1 | Substantive realism cannot answer the normative question, because its metaphysics rests on the confidence the question doubts | argued, Lecture I | pp. 40–48 (printed numbering) |
| C2 | The reflective structure of consciousness requires an agent to act under some conception of practical identity, which is the source of her reasons and obligations | argued, the core of Lecture III; the move from “must have a law” to “must have an identity” is the least defended step | pp. 82–88 |
| C3 | Every agent must value her humanity, so moral identity is not optional | argued by a transcendental step from the necessity of some practical identity; conflicts between identities are explicitly left open | pp. 90–93, note 16 |
| C4 | Practical reasons are public, as meanings are, so others can obligate us | argued by analogy with the private-language argument, and illustrated (calling someone’s name) | pp. 94–101 |
| C5 | Even a “wanton” has a second-order volition, since acting on the desire of the moment treats it as a reason | asserted in a note, with Frankfurt’s definition | Lecture III, note 4 |
| C6 | Pain is the unreflective rejection of a threat to identity, and animals can obligate us | sketched; “I won’t spell out the details here” | pp. 101–106 |
Concepts
- Normative question: “what justifies the claims that morality makes on us”, asked from the first person when confidence is shaken.
- Practical identity: “a description under which you value yourself, a description under which you find your life to be worth living and your actions to be worth undertaking”. It is not “a theoretical one, a view about what as a matter of inescapable scientific fact you are”.
- Reflective endorsement / reflective rejection: reasons arise from the first, and obligation from the second.
- Procedural vs substantive realism: whether there are correct answers because of correct procedures, or correct procedures because of independent facts.
- The thinking self and the acting self: the two sides of reflective consciousness between which there is “a relation not of mere power but rather of authority”.
Connections
The lectures engage Frankfurt (the wanton, “Freedom of the Will and the Concept of a Person”), Nagel (The Possibility of Altruism and The View from Nowhere), Williams (Ethics and the Limits of Philosophy), Hume and Kant. None of those works is read in the record, except Frankfurt’s later Tanner Lectures (LIT-293), which answer the normative question through caring and love rather than autonomy. The later “Normative Constitution of Agency” (LIT-587) turns practical identity into self-constitution: we “constitute our own practical identities” by choosing under the categorical and hypothetical imperatives.
Bearing on the record
- THEORY-029 (the regress of higher-order attitudes). The lectures offer a stopping point the theory has not yet weighed. The agent’s standpoint is not a higher-order attitude but the practical identity she must have in order to act at all. The wanton note (C5) makes every reflective agent’s action express some law. Against THEORY-029, this is specified by the structure of reflection, independently of any particular attitude. For THEORY-029, which identity counts as one’s own is still “how she thinks of herself”. Identities form “a jumble”, and the lectures give no criterion that discriminates a self-governed choice from a wanton one, except that the latter treats the desire of the moment as law. So the lectures may supply what THEORY-029’s refutation condition calls “a stopping point that is specified independently of agency”, but only if practical identity can be specified without presupposing the agent’s participation. Korsgaard says practical identity is chosen or constituted. That looks like the second horn of the dilemma. Worth a sentence in THEORY-029’s connections when it is next revised. The decision is the owner’s.
- THEORY-047 and SDT. Korsgaard’s relativism about identity (“human identity has been differently constituted in different social worlds”) is the philosophical counterpart of SDT’s claim that how a value was taken in fixes how it is held. The lectures do not discuss how identities are acquired.
- A THEORY, filed as THEORY-075 (Proposed). Obligation is constituted by reflective rejection of what would violate an agent’s practical identity, so that every reason is identity-relative and moral reasons follow only from valuing humanity as an identity. Sources: these lectures and LIT-587. Rivals within the record: Frankfurt’s caring (LIT-293). It is Proposed until the book’s reply to its commentators (LIT-599), still unread, is read.
It carries no instruction for machine-learning practice.
Limitations
- By Korsgaard’s own account the lectures are “necessarily … sketchy, and sketchily argued”, and they leave the content of obligations, and how to resolve conflicts between identities, “for another day”.
- The step from “the will must have a law” to “the agent must have a conception of her identity” is made by appeal to how deliberation seems from inside (“it is as if there were something over and above all of your desires”), not proved.
- The publicity argument depends on one reading of Wittgenstein.
- The commentators’ objections in the book version (Cohen on obligation to bad identities, Williams, Nagel, Geuss) are not here, and nor is the reply.
Open questions
- Can practical identity be specified without the agent’s participation? If not, does the account answer the regress, or relocate it?
- What happens when one’s identities conflict? The lectures leave it open (note 16), and the book’s reply is the next place to look.