Why We Love Moral Dilemmas: The Trolley Problem Explained

A runaway trolley is barreling down a set of tracks. Ahead, five innocent people are tied to the track and completely unable to move. You are standing next to a lever. If you pull the lever, the trolley will seamlessly switch to a different, side track. However, there is one innocent person tied to that side track.

Do you pull the lever, actively killing one person to save five? Or do you do nothing, allowing five to die but keeping your hands clean of direct murder?

This is the classic "Trolley Problem," introduced by moral philosopher Philippa Foot in 1967, and later expanded upon by Judith Jarvis Thomson in the 1970s. It is arguably the most famous thought experiment in modern philosophy, spawning countless memes, internet debates, and even entire television episodes (most notably in The Good Place). But why are human beings so endlessly fascinated by these morbid, completely unrealistic scenarios? And more importantly, what do our answers actually teach us about ourselves?

The Clash of Ethical Titans

Thought experiments like the Trolley Problem are intentionally designed to strip away the messy context of reality and force a direct collision between two fundamental ethical frameworks: Utilitarianism and Deontology.

Utilitarianism, championed by 18th and 19th-century thinkers like Jeremy Bentham and John Stuart Mill, argues that the morality of an action is determined entirely by its consequences. The ultimate goal of any moral action is to maximize overall happiness and minimize suffering. From a strict utilitarian perspective, the Trolley Problem is simple math: five lives are inherently more valuable than one life. You must pull the lever. Doing nothing when you have the power to save a net four lives is a moral failure.

Deontology, famously associated with the philosopher Immanuel Kant, argues that morality is based on absolute rules and duties, regardless of the outcome. A core tenet of Kantian deontology is the "categorical imperative," which states that you must never use a human being merely as a means to an end.

Actively pulling the lever makes you directly responsible for the death of the one person on the side track. You are using that person's death as a tool—a means—to save the five. From a strict deontological perspective, you cannot pull the lever. You are not morally culpable for the trolley's initial malfunction or the people tied to the track, but you become morally culpable the second you intervene and intentionally cause a death.

The Neuroscience of Morality

When people across the globe are presented with the standard lever version of the Trolley Problem, the vast majority (around 80-90%) say they would pull the lever. Utilitarian math wins out over deontological rules.

But what happens when we change the scenario slightly? Imagine you are standing on a footbridge overlooking the tracks. The trolley is barreling toward the five people. There is no lever. The only way to stop the trolley and save the five people is to push a very large, heavy man off the bridge and onto the tracks. His weight will stop the trolley, but he will die.

The mathematical outcome is exactly the same: sacrifice one innocent life to save five. Yet, in this variation, the vast majority of people (around 90%) say they would not push the man. Why does our moral compass completely reverse when the math is identical?

Dr. Joshua Greene, a moral psychologist and neuroscientist at Harvard University, used functional magnetic resonance imaging (fMRI) scans to discover exactly why this flip occurs. Dr. Greene placed subjects in fMRI machines and asked them to solve both versions of the dilemma while he monitored their brain activity.

In the first scenario (pulling a lever), Dr. Greene observed intense activity in the brain's prefrontal cortex—specifically the dorsolateral prefrontal cortex. This is the area associated with cold, rational calculation, working memory, and logical problem-solving. The subjects are doing the math: five is greater than one. Pull the lever.

In the second scenario (pushing a man), Dr. Greene observed something entirely different. The thought of physically laying hands on a stranger and violently shoving them to their death triggers massive activity in the amygdala and the ventromedial prefrontal cortex—the brain's emotional centers. A powerful, visceral, biologically ingrained feeling of revulsion overrides the rational utilitarian math. The brain screams, "Direct, physical murder is wrong!" and the deontological instinct takes over.

The Real-World Stakes of Artificial Intelligence

For decades, the Trolley Problem was dismissed by critics as a useless parlor game for philosophy students. When in real life, critics argued, would anyone ever actually face such a perfectly engineered, binary moral choice?

The answer arrived in the 21st century with the advent of artificial intelligence and autonomous vehicles. The Trolley Problem is no longer abstract; it is an engineering requirement.

Consider the programming of a self-driving car. Imagine the car's brakes suddenly fail while driving down a narrow street. It is heading toward a crowd of five pedestrians. The only way to save the pedestrians is to swerve the car into a concrete wall, which will instantly kill the car's single passenger.

How should the car be programmed? Should the algorithm be utilitarian, sacrificing the passenger to save the five? Or should the algorithm be deontological, prioritizing the safety of the passenger it was hired to protect at all costs? Who decides this? The engineers? The government? The consumer?

Researchers at MIT created the "Moral Machine," an online platform that gathers human perspectives on moral decisions made by machine intelligence, specifically autonomous vehicles. They found massive cultural differences in how people answer these variations, proving that there is no universal human consensus on the "right" answer.

By engaging with moral dilemmas like the Trolley Problem, we are not just entertaining ourselves. We are doing the difficult, necessary work of examining our deepest values, understanding the biological limits of our reasoning, and deciding exactly what kind of ethical framework we want to program into the future.