Contribution
The pages read set out the book’s position and its plan. The position is named hard incompatibilism. Both determinism and the indeterminacy of standard quantum mechanics are incompatible with the free will that moral responsibility requires. Agent-causal libertarianism would secure that free will, but we have no evidence for it. So we have no evidence that we are morally responsible, and “morality, meaning, and value remain intact”. Chapter 1’s opening defends one plank: the incompatibilist’s core condition concerns an action’s causal history, not alternative possibilities.
Key insight
Frankfurt-style cases show that responsibility does not need the ability to do otherwise. They do not show that responsibility survives an action’s being produced by a deterministic process that traces back to factors beyond the agent’s control. So incompatibilism can concede Frankfurt and relocate. Its core is origination: the agent must be the source of her decision, not a channel for factors she had no control over.
Assumptions
- Moral responsibility means basic desert. For an agent to be morally responsible for an action is for it “to belong to the agent in such a way that she would deserve blame if the action were morally wrong”. The desert is “basic”: deserved “just by virtue of having performed the action, and not … by way of consequentialist considerations” (p. xx).
- Judgements of responsibility are factual. They are not, at bottom, attitudes toward the person (pp. xx–xxi).
- Responsibility attaches primarily to decisions. Omissions and consequences are set aside, on the ground that responsibility is “a matter of control” exercised mainly over decisions (p. xxi).
- Holding accountable is not the topic. Determinism does not threaten holding people accountable for the sake of moral improvement. A compatibilism about that “avoids the issue” (pp. xxi–xxii).
Key results
Skim: Introduction pp. xiii–xxiii and Chapter 1 pp. 1–10, read in full. Section titles for the rest come from the table of contents.
- The position (pp. xviii–xix).
- “Hard determinism” is rejected as a label, “since I do not claim that determinism is true”, and whether quantum mechanics is deterministic “is currently an open question”.
- “Not only is determinism incompatible with moral responsibility, but so is the sort of indeterminacy specified by the standard interpretation of quantum mechanics, if that is the only sort of indeterminacy there is.”
- “We have no evidence for indeterminacy of the kind that would be required for agent-causal libertarianism, and … therefore, we have no evidence that we are morally responsible.”
- Against “no-free-will-either-way” views: “there is a coherent indeterminist scenario in which we have this sort of free will … and it may well be that agent-causal libertarianism is metaphysically possible.”
- In Strawson’s table of nine positions, hard incompatibilism takes positions 1, 4 and 7.
- The stronger claim, stated once (p. xiii). “I shall argue that our best scientific theories indeed have the consequence that we are not morally responsible for our actions.”
- Plan of the book (pp. xv–xx).
- Chapter 1: Frankfurt-style arguments are “largely successful”, and a new one is offered. They do not threaten “the core incompatibilist claim … to have causal histories that make agents ultimate sources of their actions”.
- Chapter 2: event-causal libertarianism fails to make agents sources. Agent-causal libertarianism “might well satisfy this requirement”.
- Chapter 3: agent causation is “seriously challenged by empirical considerations”.
- Chapter 4: against P. F. Strawson, “determinism can indeed be relevant” to the reactive attitudes. Causal-integrationist compatibilisms are met with “counterexamples” and “a general argument that no relevant distinctions can be made between cases in which agents are determined in ways compatibilists think are consistent with moral responsibility and those in which agents are determined in ways that clearly undermine moral responsibility”. The table of contents names it “the four-case argument” (§ “Haji’s response to an earlier version of the four-case argument”, p. 117).
- Chapters 5–7: morality, criminal justice (no retributivism; preventive detention and rehabilitation can be justified) and meaning in life survive.
- What it does not deny (pp. xxii–xxiii). The hard incompatibilist does not deny “the ability to choose in accord with one’s values, or the capacity to be responsive to the reasons there are for acting”, or freedom from coercion. The position is “essentially a negative position”.
- Chapter 1, the two intuitions (pp. 1–6).
- Principles (1) and (2): free action requires that the agent could have done, or chosen, otherwise. Pereboom will argue that “any such requirement that is relevant to explaining why an agent is morally responsible for an action falls to counterexamples”.
- Fischer’s Black/Jones case is quoted. With Fischer, Pereboom holds that it does not refute “the claim that moral responsibility requires that the actual causal history of the action not be deterministic”.
- Condition (3): “An action is free in the sense required for moral responsibility only if it is not produced by a deterministic process that traces back to causal factors beyond the agent’s control.”
- Condition (O): “If an agent is morally responsible for her deciding to perform an action, then the production of this decision must be something over which the agent has control, and an agent is not morally responsible for the decision if it is produced by a source over which she has no control.” This is “the most fundamental and plausible incompatibilist intuition”, and it also explains why an uncaused decision would not be one the agent is responsible for.
- Leeway against causal-history incompatibilism: Pereboom is the second.
- “The more powerful, and indeed the best, type of challenge to compatibilism develops the claim that causal determination presents in principle no less of a threat to moral responsibility than does covert manipulation. We shall turn to that challenge in Chapter 4” (p. 6).
- Chapter 1, robustness and Widerker (pp. 6–10). A flicker of freedom must be “robust” to ground responsibility: securing it must be what would have got the agent off the hook. Blushing is not robust. The Kane/Widerker dilemma is stated. If the prior sign determines the action, the libertarian lacks the intuition of responsibility. If it does not, the agent could have done otherwise. Pereboom calls it “a test for the effectiveness of any Frankfurt-style argument”. His answer, a scenario “without prior signs” and “a new Frankfurt-style scenario” (pp. 14–23 by the contents), falls outside the pages read.
Claims
| id | claim | strength | support |
|---|---|---|---|
| C1 | We have no evidence that we are morally responsible in the basic-desert sense | not assessable here | stated in the Introduction; the arguments are in Chapters 2–5, not read |
| C2 | Frankfurt-style cases leave standing a causal-history condition on responsibility | moderate | the argument from Fischer’s concession and condition (3), pp. 2–6 |
| C3 | Origination (O) is the core incompatibilist intuition | weak | asserted with Kane, Honderich and Watson cited, pp. 4–5 |
| C4 | Determination is no less a threat to responsibility than covert manipulation | not assessable here | announced, p. 6; argued in Chapter 4, not read |
| C5 | Hard incompatibilism leaves morality, meaning and value intact | not assessable here | announced; Chapters 5–7, not read |
Concepts
- Hard incompatibilism. Moral responsibility is incompatible with determinism and with event-causal indeterminacy of the quantum kind, and we have no evidence for the agent-causal indeterminacy that would allow it. Not “hard determinism”, since determinism is not asserted.
- Basic desert. Desert of blame or credit just for having performed the action, not for consequentialist reasons.
- Leeway incompatibilism. Responsibility needs an indeterministic causal history only because that secures alternative possibilities.
- Causal-history incompatibilism. The causal history matters to responsibility independently of alternative possibilities.
- Origination (O). The production of a decision must be something over which the agent has control.
- Robust alternative. An alternative whose securing would by itself have made the agent not responsible for what she did.
Connections
The pages read engage Frankfurt and Fischer (Black/Jones), Kane’s ultimate-responsibility condition (UR), Honderich on origination, Widerker’s and Ginet’s defences of alternative possibilities, McIntyre’s princess case (footnote 9), Galen Strawson’s table and “no-free-will-either-way” view (Double’s term), and P. F. Strawson’s reactive attitudes. The acknowledgments say the book’s general conception comes from “Determinism Al Dente” (Noûs, 1995), and that part of Chapter 1 is “Alternative Possibilities and Causal Histories” (Philosophical Perspectives, 2000).
Bearing on the record
- THEORY-040’s promote_when asks for “a first-hand reading of Pereboom’s four-case argument … that shows a manipulation meeting a named historical condition”. This reading does not supply it. The four-case argument is in Chapter 4, which is not read. All that is read is the Introduction’s one-sentence description of it and p. 6’s announcement that “causal determination presents in principle no less of a threat to moral responsibility than does covert manipulation”.
- THEORY-119 (PAP is false by Frankfurt-style cases) gains an ally with a limit. Pereboom agrees that such arguments are “largely successful” and treats Widerker’s dilemma as the test any of them must pass. His buffer case, which THEORY-119’s promote_when names, is on pp. 18–23 and was not read. He also argues that THEORY-119’s conclusion leaves incompatibilism standing (THEORY-316).
- THEORY-121 names Pereboom’s four cases, rebuilt to manipulate the taking of responsibility, as its candidate refuter. That rebuilding is in Chapter 4, p. 117 by the contents (“Haji’s response to an earlier version of the four-case argument”), not read.
- QUESTION-071 says that in Pereboom’s four cases “an agent’s present structure is produced by someone else and matches an unmanipulated agent’s in every respect the account names”. The pages read do not describe the cases, so this cannot be confirmed. On the book’s own description the argument runs from cases where agents are “determined in ways that clearly undermine moral responsibility” to cases compatibilists accept. The case at the compatibilist end is ordinary determination, where nobody else produces the structure. The question’s phrase fits the manipulation cases, not all four.
- The LIT’s summary said the book holds that “whether determinism is true or not, we lack the free will that basic-desert moral responsibility requires”. The Introduction is more careful. It says agent-causal libertarianism “may well” be metaphysically possible and would give us that freedom. The claim is that we have no evidence for it, so “no evidence that we are morally responsible”. Elsewhere it says our best theories have the consequence that we are not. The summary is corrected. Its description of the four cases is unverified here and is marked so.
- THEORY-316 is filed from Chapter 1’s pp. 1–6.
- No instruction for machine-learning practice; nothing for the anthology.
Limitations
- Only the Introduction and ten pages of Chapter 1 were read. The arguments for every distinctive thesis (the new Frankfurt-style case, the case against libertarianism, the four-case argument, the practical chapters) are outside them.
- The Introduction’s summaries are the author’s own and are not arguments.
Open questions
- What are the four cases, exactly? In particular, does the agent in each meet every compatibilist condition Chapter 4 names, and is the agent in case 4 described in present-structure terms only? QUESTION-071 needs that answer.
- Does the new Frankfurt-style case (pp. 18–23) escape Widerker’s dilemma? THEORY-119 depends on it.
- How does the hard incompatibilist keep “the capacity to be responsive to the reasons there are for acting” while denying responsibility, and is that capacity the will line’s notion of organization?