📄 social science

What Would an Empirical Counterexample to Motivational Internalism Require?

This paper proposes a rigorous empirical counterexample criterion to distinguish genuine amoralist cases from masked or defeated motivation, thereby establishing the evidential standards required to refute unrestricted psychological internalism while delineating the limits of empirical challenges against semantic or rationalist variants.

Original authors: Voin Milevski

Published 2026-10-19

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Voin Milevski

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine a person standing on the side of a road, watching a dog lie injured and whimpering in the gutter. This person knows, with absolute certainty, that they ought to stop, pick up the animal, and take it to a vet. They have the time, the money, and no other pressing obligations. Yet, they simply walk past. When asked later, they sincerely say they felt no pull to help, no inner urge to act, and no motivation to do anything at all. To an observer, this seems like a clear case of someone making a moral judgment without feeling the drive to follow it. But in the world of moral psychology, this simple story hides a deep puzzle. For decades, philosophers have debated whether making a sincere moral judgment necessarily creates a spark of motivation to act on it. This idea, known as motivational internalism, suggests that you cannot truly believe you should help without feeling some desire to do so. If the person walking past the dog is telling the truth, they might be the "amoralist" that disproves this rule. However, proving that someone truly lacks motivation is incredibly difficult. A person might be motivated but too tired, too distracted, or too overwhelmed to act. Their motivation might be hiding, or "masked," by other parts of their mind. Without a way to tell the difference between a missing spark and a hidden one, the debate has often stalled on stories rather than facts.

A new paper by Voin Milevski from the University of Belgrade tackles this stalemate not by arguing about philosophy, but by asking a practical question: what kind of evidence would we actually need to prove that a person has a moral judgment but zero motivation? The author argues that simply pointing to a person who doesn't act is not enough. To call such a person a genuine counterexample, we need a rigorous method to rule out the possibility that their motivation is just buried or blocked. The paper proposes a strict set of rules for how scientists and philosophers should test these cases. It suggests that we cannot just assume a person is unmotivated because they didn't act; we must first show that no hidden motivation could possibly explain their behavior. This requires comparing two different explanations for the same behavior: one where the motivation is truly gone, and another where it is still there but being stopped by a specific obstacle.

The core of the paper's proposal is a "paired-model test." Imagine you are trying to figure out why a car won't start. One theory is that the engine is dead. Another is that the engine is fine, but the battery is disconnected. You cannot just look at the car and decide; you have to test the battery. Similarly, to prove a person lacks moral motivation, researchers must identify a specific reason why a hidden motivation might be blocked, such as a habit of avoiding difficult tasks or a mental block caused by exhaustion. This "blocker" must be something we can point to independently, not just a made-up excuse invented after the fact. If a researcher claims a person is unmotivated, they must also be able to show that if you removed this specific blocker, the person would suddenly show signs of that motivation. If the motivation was truly absent, removing the blocker would change nothing. If the motivation was just hidden, removing the blocker would make it appear.

The paper also introduces a crucial rule about time, called the "temporal provenance" constraint. This addresses a common mistake in experiments. Suppose a researcher removes a blocker and the person suddenly starts helping the dog. Does this prove they were motivated all along? Not necessarily. The intervention itself might have created a new motivation that wasn't there before. Perhaps the person only helped because the researcher made it easier, or because they felt pressured to do so. To count as a true counterexample, the evidence must show that the motivation existed before the experiment started, not that it was created by the experiment. The researcher must be able to link the person's new actions back to a specific mental state that was present earlier, rather than a new reaction to the situation.

When the author applies these strict rules to the kinds of cases usually discussed in philosophy, the results are sobering. Simple stories about people who don't act, often called "vignettes," fail the test. These stories describe a person who doesn't act and says they aren't motivated, but they do not provide the necessary evidence to rule out hidden blockers or to prove that the motivation was there all along. Similarly, studies of people with brain injuries, specifically damage to a part of the brain called the ventromedial prefrontal cortex, do not yet provide a definitive answer. While these patients often struggle to act on moral judgments, it is not yet clear if they have lost the motivation entirely or if their ability to process the value of actions has been damaged in a way that mimics a lack of motivation. The paper suggests that until we can show that these patients retain their moral understanding while their specific motivation disappears, and until we can rule out other explanations like general confusion or lack of effort, we cannot claim to have found a true counterexample.

The paper also looks at experiments where ordinary people are asked to judge whether a character in a story is moral. These studies show that people's opinions vary wildly depending on how the story is told. Some people think a character who doesn't act is still moral; others think they are not. The author argues that these experiments tell us more about how people use words and classify behavior than they do about the actual inner workings of the human mind. They reveal our habits of judgment, not the hidden reality of whether a person is truly unmotivated.

Ultimately, the paper concludes that finding a true "amoralist"—someone who sincerely judges an action is right but feels absolutely no motivation to do it—is much harder than it seems. It is not enough to find someone who doesn't act. To earn the title of a counterexample, a case must survive a battery of tests that rule out hidden motivations, prove that the motivation was present before any experiment began, and show that the person's moral judgment remains intact even when their motivation is gone. The author does not say that such a person does not exist, but rather that we have not yet found one with the kind of proof that would settle the debate. The paper shifts the goalposts from storytelling to evidence, demanding that if we want to claim that moral judgment can exist without motivation, we must be able to prove it with the same rigor we use in any other scientific investigation. Until then, the possibility remains that every person who seems unmotivated is simply a person whose motivation is hidden, waiting to be revealed by the right kind of test.

Title: What Would an Empirical Counterexample to Motivational Internalism Require?
Author: Voin Milevski (University of Belgrade)

1. The Problem

The paper addresses the evidential gap in the debate over motivational internalism—specifically, the unrestricted psychological co-instantiation thesis which claims that sincere moral judgement ( $J$ ) necessarily co-occurs with corresponding motivation ( $M$ ) ( $\square(J \rightarrow M)$ ). While the modal project (conceiving of an amoralist) is well-established, the empirical project (identifying an actual agent who instantiates $J$ without $M$ ) faces a critical methodological hurdle.

The core difficulty is distinguishing between two psychological organizations that produce the same observable "negative profile" (inaction, lack of felt pressure, and sincere denial of motivation):

The Absence Model: The agent genuinely lacks the corresponding motive.
The Masking Model: The agent possesses the motive, but it is blocked, defeated, or rendered phenomenally silent by a "blocker" (e.g., depression, exhaustion, competing habits, or executive bottlenecks).

Current literature often treats these as logically possible alternatives without a disciplined method for determining which is the better explanation of actual data. The paper argues that without a rigorous criterion, internalists can always posit undetectable residual motives (making the thesis unfalsifiable), while externalists may incorrectly assume absence based on surface behavior.

2. Methodology

The author proposes a single Empirical Counterexample Criterion designed to adjudicate between the Absence and Masking models. This framework relies on comparative causal inference rather than deductive proof. It integrates three primary components:

The Paired-Model Test: Masking hypotheses are not granted automatic evidential standing. To be considered a serious rival, a masking model must satisfy four conditions:
1. Independent Warrant: There must be antecedent evidence for the proposed "blocker" ( $b$ ) and its capacity to gate the process, independent of the verdict at issue.
2. Motive Warrant: There must be antecedent reason to predict a content-specific motive ( $m$ ) even when bracketing the $J \rightarrow M$ connection.
3. Discriminating Predictions: The model must specify in advance which observable variables (attention, planning, latency, effort) should change when the blocker is varied.
4. Failure Conditions: The model must specify outcomes that would reduce its support (e.g., if the blocker is removed and no content-specific change occurs).
The Temporal Provenance Constraint: This addresses the problem of intervention. If an agent responds to a manipulation (e.g., removal of a blocker) at time $t_2$ , this does not automatically prove the motive existed at $t_1$ . The framework distinguishes between:
- Revelation: The intervention removes a blocker, revealing a persisting motive present at $t_1$ .
- Generation: The intervention creates a new motive or activates a disposition to form one.
- Constraint: Post-intervention motivation supports earlier masked presence only if independent evidence links the response to the motivational organization present at $t_1$ , rather than to the intervention itself.
Scope Conditions: The criterion explicitly defines which internalist theses are vulnerable to empirical counterexamples. It targets unrestricted, psychologically substantive co-instantiation theses. It acknowledges that restricted theses (e.g., limited to rational agents) or constitutive/semantic theses (where motivation is part of the definition of $J$ ) require different evidential routes or may be immune to this specific empirical test.

3. Key Contributions

The paper contributes a structured rule for admitting and defeating masking explanations in metaethics:

Admission Rule: Masking hypotheses must "earn" their status through antecedent warrant and predictive commitments; they cannot survive merely as logical possibilities or post-hoc redescriptions.
Provenance Constraint: It isolates the temporal ambiguity in intervention studies, clarifying that later responsiveness does not entail earlier presence unless continuity is independently established.
Graded Counterexample Status: The author introduces a four-grade system to evaluate evidence:
- Grade 1: Negative profile is compatible with absence (masking remains possible).
- Grade 2: Masking is an admitted rival with local comparative support.
- Grade 3: Masking models face prospective, sensitive tests and risk failure.
- Grade 4: Full counterexample status (independent $J$ warrant, failed masking predictions, warranted scope conditions).

4. Results and Applications

The framework is applied to three existing bodies of evidence, yielding specific verdicts:

Amoralist Vignettes: Thin descriptions (e.g., "Patrick" who judges he ought to help but does not) fail to reach counterexample status. They lack the independent warrant for blockers or motives required to distinguish absence from masking. They are useful for modal conceivability but insufficient for empirical identification.
vmPFC Lesion Cases: Current findings regarding ventromedial prefrontal cortex damage do not yet warrant counterexample status. While they offer mechanistic data, the literature remains contested regarding whether moral judgement ( $J$ ) is truly preserved or if motivational deficits are content-specific versus general (e.g., effort or valuation deficits). The framework suggests that future cases could reach counterexample status if $J$ is triangulated independently and masking models fail under heterogeneous revealing conditions.
Folk-Attribution Experiments: Studies asking whether ordinary speakers attribute moral belief to unmotivated agents measure classificatory practices and conceptual policies, not the latent psychological organization ( $J \land \neg M$ ) of the agent. These results do not directly identify the presence or absence of motivation.

5. Significance and Claims

The paper claims to specify when an empirical amoralist has earned counterexample status and which internalisms that status threatens.

Modest Claim: The framework does not claim to prove psychological absence deductively or to settle all semantic disputes about the nature of moral judgement. It acknowledges that psychological attribution is inherently fallible and theory-laden.
Primary Contribution: It prevents the "cheap" move of positing undetectable motives to save a theory while simultaneously preventing externalists from assuming absence based on surface behavior. It establishes that repeated sensitive null results under conditions where masking models predict content-specific differences can make the Absence model predictively superior, even if masking remains logically possible.
Boundary: The paper delineates an "empirical-access boundary." Psychologically substantive internalisms (which posit weak, defeasible, or regulative motives) are testable via this framework. However, semantic, constitutive, or purely rationalist theses (where motivation is definitional to judgement or derived from a priori rational constraints) require different evidential routes and are not directly threatened by this specific empirical criterion.

In summary, the paper argues that an empirical counterexample to motivational internalism is not a single observation of inaction, but a rigorously controlled body of evidence that independently warrants the judgement, rules out admissible masking models through prospective testing, and establishes the temporal continuity of motivation (or lack thereof) across interventions.

1. The Problem

2. Methodology

3. Key Contributions

4. Results and Applications

5. Significance and Claims

More like this