Can responsibility survive causal explanation?

A name removed

Imagine that Tomas deletes his colleague Kavita’s name from a report so that he will receive all the credit. He understands her contribution, knows the deletion is unfair and has time to reconsider. Nobody threatens him. Asked to explain, he describes his competitive upbringing and his fear of being overlooked. If those causes fully explain his decision, does the explanation remove his responsibility?

The question is one of moral responsibility: whether someone deserves praise or blame for what they do. Causing an outcome is not enough by itself; a fallen branch causes damage without deserving reproach. Assuming that Tomas understands the relevant facts and the wrongness of his action, we are asking what control a person must have.

Determined does not mean coerced

Causal determinism says that the past, together with the laws of nature, fixes what happens next: with that entire past and those laws unchanged, no different outcome is possible. Merely showing that an action has causes does not establish this stronger claim. [2]

Coercion concerns a threat or pressure through which someone compels another’s conduct. Suppose instead that a manager threatens Tomas with serious harm unless he removes Kavita’s name. That would change the conditions under which he acts. Determinism, by contrast, is a claim about how events follow from earlier conditions, and it does not establish that any such threat is present.

Spinoza made understanding itself part of nature’s causal order. Deliberation need not be useless because it has causes: considering Kavita’s contribution could help cause a different response. Beauvoir asked how someone’s situation shapes their possibilities. Neither connection establishes that every caused action deserves blame, or that every constrained person has lost all agency.

Hume: causes can connect an action to us

In his Enquiry concerning Human Understanding (1748), David Hume locates liberty in being able to act or refrain according to one’s will. If Tomas chose to restore the name, could he do it? This conditional question asks what would happen if his choice differed; it does not say that he could choose differently with every prior condition unchanged. [1]

Hume also argues that blame depends on a connection between conduct and the person’s character. If an action revealed nothing of the person’s motives or dispositions, why blame that person for it? Causal explanation can therefore help show why an action is attributable to someone. Hume’s account of necessity emphasizes regular connections and the inferences we draw from them, so it is not simply the modern definition given above. [1]

His view supports a compatibilist approach, on which determinism need not exclude freedom or responsibility. It does not show that wanting something is always enough control: a desire may itself be compulsive, or someone’s deliberation may have been manipulated. [3]

Van Inwagen: could the choice itself differ?

Incompatibilism holds that free will cannot exist if determinism is true. Peter van Inwagen’s 1975 argument asks whether someone could act otherwise when a complete earlier state of the world and the physical laws leave only one outcome. The judge in his example deliberates without threats or restraints, so the difficulty remains even once coercion has been ruled out. [2]

Applied to Tomas, the challenge runs like this. He cannot change the past before his birth or the laws that govern it. If these fix his deletion, acting otherwise would require a difference in something beyond his control, and so, van Inwagen argues, Tomas cannot act otherwise. This reconstructs the argument’s main idea, not its full formal proof.

The disputed step carries our lack of control over the past and the laws through to their consequences. Saying “he would have restored the name if he had chosen to” does not yet justify rejecting that step. To infer that Tomas cannot deserve blame, we also need a control condition: deserved blame requires that the person could have refrained. That requirement is itself disputed. Nor does the argument establish that determinism is actually true. [2] [3]

Another account of control

A contemporary compatibilist response asks whether the process producing the action is appropriately responsive to reasons, that is, capable of recognizing relevant considerations and changing conduct in response. On John Martin Fischer’s approach, the control that responsibility needs can concern how the actual action is guided, without requiring an alternative with the whole past and the laws held fixed. This is a later response, not Hume’s own theory. [3]

Was Tomas able to understand Kavita’s claim? Could his way of deciding respond to reasons against taking the credit, even though he discounted them this time? Failing to follow a good reason is different from being unable to respond to reasons at all, and the failure can be precisely what we blame.

An incompatibilist can still ask whether such control is enough when the character guiding his response has origins beyond his control. Manipulation cases press that question: someone’s reasoning can seem intact even when another person has engineered its direction. Current debate therefore examines both a person’s capacities and their history. [3]

What a brain finding would need to show

The consciousness lesson distinguished explaining a capacity from explaining experience. The question here is which capacity responsibility requires, and a headline announcing brain activity before a reported decision cannot settle it. What was measured, in what task, and why would it undermine the proposed kind of control?

Specialists dispute how to interpret the readiness potential, a brain signal recorded before some movements. This lesson does not assess those experiments, nor does it treat conscious certainty as proof of freedom. Neuroscience can inform an account of control, but to conclude that nobody is responsible, we would still need an argument connecting the finding to the control that responsibility requires. [4]

Try stating the missing condition

Tomas says, “My upbringing caused the deletion, so I was coerced.” Kavita replies, “You wanted the credit, so we need ask nothing else.” What has each left out? State the control condition that would make a compatibilist and an incompatibilist judgment diverge. This exercise is optional.

Read a possible response

Tomas has described influences, not a coercive threat, and he has not shown that the entire past and the laws fixed his decision either. A causal explanation could reveal ordinary deliberation, coercion or a serious impairment, so the details matter. Kavita is right to examine his purpose, but wanting the result cannot by itself establish adequate control.

A compatibilist could hold him responsible if the decision came from his own appropriately reason-responsive capacities, with understanding and without coercion that excuses it. An incompatibilist who requires the ability to refrain, with the same past and laws, would deny that this condition can be met under determinism. They disagree about the control required, even if they accept the same account of what happened. Had the manager threatened him, that would raise a further question about excuse, and the threat’s seriousness and Tomas’s options would matter.

Module checkpoint: what does the explanation leave open?

This module has asked how we test explanations and judgments when different methods reveal different parts of a problem. Scientific testing and fair treatment of testimony examined how inquiry earns trust. Institutional justice and dependency asked what measures of advantage and cost can overlook. The lessons on art, conscious experience and responsibility then separated what an account explains from the further conclusions drawn from it. The questions are connected, but they are not stages in a proof that all these thinkers agree.

Two concepts make the connection concrete. From Miranda Fricker comes testimonial injustice: prejudice that unfairly reduces a speaker’s credibility. From the compatibilist accounts in this lesson comes reasons-responsive control: a way of deciding that can recognize reasons and respond to them. The first helps us assess the evidence on which someone is blamed; the second helps us state what control blame requires. A fair hearing, however, cannot by itself settle how much control responsibility demands.

Now extend the case. The chair dismisses Kavita’s complaint because he assumes that junior colleagues invent grievances. He then cites an assessment predicting Tomas’s competitive behavior: “His action was predictable, so nobody is responsible and we can leave the report alone.” How do the two concepts bear on the module’s question here? Then say what question they leave open, and what the sources used here cannot show. This checkpoint is optional.

Read a checkpoint response

The chair’s prejudice undermines the hearing Kavita receives. That is testimonial injustice, and it bears directly on the module’s question about what supports a judgment: her drafts and contributions should be examined without treating her rank as evidence against her. Her account remains open to checking, since fair credibility is no guarantee of truth.

Prediction is a different matter. An accurate forecast establishes neither determinism nor coercion, and it does not show that Tomas lacked reasons-responsive control. The question is how he understood and responded to the reasons available to him. Even if the assessment describes him accurately, the chair has offered no argument that it defeats this control condition.

That still leaves a real question open: is such control enough for deserved blame when its own history lies beyond the agent’s ultimate control? And remember what kind of evidence we have been reading: selected philosophical passages and specialist overviews, which do not investigate this invented workplace or validate its assessment. They help us frame the questions, but they cannot supply the missing evidence. Meanwhile, restoring Kavita’s credit can be justified even while the grounds or degree of Tomas’s blame remain disputed.

Carry this forward

Before deciding whether an explanation excuses someone, identify what it explains and what control the judgment requires. The next module turns to shared responsibilities, beginning with what we owe beyond the present human world. It uses these distinctions without requiring agreement on a single theory of freedom.

Sources and limits

[1] David Hume, An Enquiry concerning Human Understanding, section VIII, first published 1748, electronic text at Hume Texts Online, for liberty, necessity and the connection between blame and character. Read paragraphs 21–32, with the account of necessity in paragraph 5; the central passages on liberty and responsibility are paragraphs 23 and 29–31. The site’s labels “E 8.23” and “SBN 95” identify the section and paragraph and the Selby-Bigge–Nidditch pagination. Linked notes were not read, and no printed edition was compared. This was a selected reading, not the complete work. Hume’s broad historical generalizations are not adopted here.

[2] Peter van Inwagen, “The Incompatibility of Free Will and Determinism”, Philosophical Studies 27 (1975), pp. 185–199, for determinism and the argument that free will is incompatible with it. Read the main text, pp. 185–197, especially the judge example and the argument in part III, pp. 190–194, and the discussion of conditional analysis, pp. 196–197. Printed pp. 190–191 were also inspected as images. The endnotes and later versions of the argument were not read. This is an argument about compatibility, not evidence that determinism is true.

[3] Michael McKenna and D. Justin Coates, “Compatibilism”, Stanford Encyclopedia of Philosophy, revised 16 April 2024, for reasons-responsive accounts and their difficulties. Read §4.4, including §§4.4.1–4.4.4, and consulted the opening framing; the symbol § marks a section. The underlying works by Fischer and other contemporary philosophers were not independently read.

[4] Timothy O’Connor and Christopher Franklin, “Free Will”, Stanford Encyclopedia of Philosophy, revised 3 November 2022, for the difference between disputed interpretations of experiments and arguments about agency. Read the closing discussion of empirical challenges in §3.1, beginning “Pereboom’s empirical basis,” and §3.2. The experiments and scientific reviews cited there were not independently assessed.

Sources accessed 20 September 2026. Tomas, Kavita, the manager, the chair and the assessment are original teaching material. Earlier lessons support the named comparisons and the recap; their underlying sources were not newly researched. No whole-book reading, independent scholarly review or learner testing is claimed.