Timothy O'Connor on Libertarian Free Will & Agent Causation

Event causal libertarianism asserts that human beings have free will when their actions are caused by agent-involving mental states and events, combined with an indeterministic event that results in the final decision. Agent causal libertarianism asserts that an agent acts with free will when the decision is caused by herself, when she initiates a new causal chain by herself, totally independent of prior mental states and events and conditions.

There are two main versions of libertarian free will: agent causal libertarianism and event causal libertarianism. Agent causal libertarianism imparts certain causal powers to agents as unique substances that are more than just a collection of events—they are more than just their physical makeup and brain activity. Agents as substances have a sui generis kind of causal power, the ability to be essentially uncaused causes, meaning they can initiate action without themselves being causally determined by antecedent events. Event causal libertarianism introduces indeterminacy at the level of events, saying all there are in the universe are physical events, but they reject determinism to preserve free will. The main objection to event causal libertarianism is that indeterminate events seem no more within the control of the agent than determined events, and the agent disappears at the moment that outcome occurs.

Event causal libertarianism proposes that decisions are caused by mental states (beliefs, desires) but with indeterminism introduced. Robert Kane's model places indeterminism between reasons and decisions, so in equilibrium states where reasons don't strongly favor one option, decisions become genuinely undetermined. This preserves freedom in difficult choices. The model addresses concerns about responsibility by introducing transferred responsibility: even actions in compromised states can be attributed to agents if they made prior conscious decisions to enter those states. The Buridan's ass problem (equally balanced choices) poses a challenge: if reasons are equally balanced, indeterminism must somehow tip the balance. Critics note that deterministic models require external mechanisms that agents don't control, while libertarian models claim only the agent's reasons matter.

Event causal libertarianism holds that freedom consists in decisions being caused indeterministically by other events, rather than deterministically. On this view, events like desires at specific times cause actions, and if these causal chains involve indeterministic causation, agents can be free. Robert Kane presents this view, arguing that indeterministic event causation allows for genuine freedom.

Emergentism proposes a middle path: mind is a property that emerges from complex physical systems. Like water molecules (H2O) alone are neither solid nor liquid, but billions organized in a crystal lattice produce solidity (ice), mind emerges from billions of neurons. A single neuron is not conscious, but the complex network produces consciousness. Mind is caused by neurons and depends on brain structure, but cannot be reduced to individual neurons - it's a property of the complete system.

Emergentism is the philosophical view that certain properties of a system, particularly consciousness, are new outcomes that cannot be reduced to or deduced from the properties of their constituent parts, and while these emergent properties may supervene on lower-level physical processes, they possess their own causal powers and explanatory status that resist complete reductionist explanation.

Agent causation can be understood as an emergent causal power - a power that emerges from the interactions of simpler components (like neurons) but is not itself determined by those interactions. Just as the mind emerges from brain activity but is not determined by it, agent causation might be an emergent property that allows agents to act on the basis of reasons in ways that are not fully determined by prior physical events. This view attempts to preserve both the causal closure of physics and the special status of human agency.

Emergence is not merely a coarse-grained description of complex physical systems but represents a genuinely fundamental aspect of reality, particularly at the level of consciousness, where new kinds of properties appear that cannot be fully explained by underlying physical laws; while chemistry and biology can be understood through organized physical phenomena, consciousness involves distinctive phenomena that are causally dependent on but qualitatively different from physical processes, suggesting a fundamental feature of the universe that may be explained theistically as part of a design argument.

Emergentism is a non-reductionist dualist position in philosophy of mind that posits mental phenomena, particularly consciousness, emerge from the interaction of physical components (such as neurons) without being reducible to them. The video presents three main variants: classical materialist emergentism (which holds that mental states are caused by sensory contact with the world through neuronal stimulation), token emergentism (which explains mental causation through chain reactions similar to dominoes), and strong emergentism (which argues for ontological dependence of mind on brain). Each variant faces significant challenges: classical emergentism struggles with psychophysical closure and non-monotonicity, token emergentism fails to explain higher mental processes like imagination and dreams, and strong emergentism risks collapsing into reductionism.

This lecture introduces the central question of whether results matter to moral blameworthiness. The speaker presents causation as fundamental to both criminal and tort law—causation is an explicit element in liability rules, determines punishment severity (completed crimes vs. attempts), and establishes remedy amounts. The alternative legal systems—punishing attempts as severely as completed crimes, or basing liability solely on risk rather than causation—would dramatically alter legal structures. The traditional framing of 'moral luck' as whether we can be blamed for things beyond our control is conceptually flawed. Using ordinary language, 'luck' refers to freakish, abnormal, unforeseeable events that surprise us. Morality and law actually blame people only for things that are in the normal course of their actions—what they could reasonably expect and control. The control argument against causation's relevance to blameworthiness fails because it equivocates on 'control': using it in two different senses.

This section presents Pereboom's hard incompatibilist position, distinguishing basic desert moral responsibility from forward-looking goals. Basic desert—the idea that people deserve praise/blame regardless of consequences—is challenged by determinism, but moral formation and reconciliation remain intact. The manipulation argument demonstrates that if we accept manipulation doesn't eliminate moral responsibility, ordinary causal determination shouldn't either. Pereboom evolved his presentation from extreme manipulation cases to more psychologically plausible scenarios. He argues indignation is a blunt instrument that resists effective behavior change, proposing refined emotional attitudes focused on reconciliation and protection. If basic desert is rejected, criminal punishment should shift from desert-based to incapacitation-based justification, comparing dangerous individuals to disease carriers requiring quarantine.

Spinoza distinguished between causal responsibility (being the cause of something) and moral desert (deserving reward or punishment). When we say someone is 'responsible' for an action, we mean they were its cause. This causal responsibility exists regardless of whether they could have acted otherwise. Spinoza argued that moral systems based on desert (rewarding the deserving, punishing the undeserving) are problematic because they assume free will where none exists. However, causal responsibility is sufficient for meaningful ethics: we should avoid harmful actions because they cause harm, not because we will be punished for them. This shifts moral motivation from fear of cosmic retribution to understanding of natural consequences.

Moral responsibility exists on a spectrum: intentional killing (most blameworthy), reckless killing (knowingly taking risks), negligent killing (failing reasonable care), and accidental killing (unpredictable outcomes). All except accidental involve some moral responsibility. But for causation refers to distant connections where events wouldn't have occurred without particular factors—it can be nearly anything. Proximate cause matters for responsibility, not but for causation. Using but for causation to assign responsibility is a misuse that can establish distant connections irrelevant to actual blameworthiness. In overdetermined situations where multiple causes could independently produce the same effect, each cause may bear full responsibility.

There are multiple senses of responsibility that do not require free will: (1) Causal responsibility—being causally responsible for actions like a tree causing damage to a roof; (2) Liability responsibility—being liable for damages in a car accident; (3) Attributability responsibility—having characteristics, personality traits, and qualities attributed to you. Forward-looking moral responsibility grounded in practices like moral formation, safety, reconciliation, and restitution can exist without basic desert. The only thing denied is the kind of control required for retributive punishment, moral anger, resentment, and indignation based on desert.