It's called the trolley problem, and it's all about how far you'd be willing to go to save lives in an emergency – even if it meant killing somebody.
Now, scientists have tested this famous thought experiment in real life for the first time: with almost 200 human participants, caged mice, electric shocks – and one heck of a decision to make.
You're probably already aware of the classic trolley problem itself, but here's a quick recap - because it's essential to be familiar with it to understand the moral dilemma posed in the new experiments.
Imagine seeing a runaway trolley (or train carriage) hurtling down the tracks, headed directly for five people who are tied to the rails ahead.
The good news is you have the power to save their lives – by simply pulling a lever that will divert the runaway trolley onto another track so that it avoids these poor, tied-up people.
There's just one problem, but it's a big one.
On the other track, there's also a single person tied to the rails, and if you intervene to save the five people on the original track, you'll end up killing this other person.
There are lots of variants and twists that expand upon the dilemma of this classic scenario, each giving a different spin on the hypothetical rightness and wrongness of pulling the metaphorical lever (or not).
But at its heart, the ethical question posed by the trolley problem is whether you should save five lives by taking one – which means getting your hands dirty – or if you should refrain from actively choosing to kill someone, which perversely results in even more death.
This probing dilemma has pondered moral philosophers since the 1960s, but in a provocative twist on the classic problem, psychologists in Belgium have brought the nightmare scenario into the real world (or at least half-way, you might say).
In an experiment with almost 200 student volunteers, participants were admitted to a lab, one at a time, and presented with a difficult choice.
In the lab, an electroshock machine was connected to two separate cages. One of these cages had five mice within it. The other cage had a single mouse occupant. You can probably tell where this is going.
The participants were told they had 20 seconds to make a decision. If they did nothing, a very painful but non-lethal electric shock would be applied to the cage containing the five mice.
If, however, they simply pressed a single button placed before them, then those five mice would be spared the electric shock, which would instead be administered to the single mouse in the other cage.
Before you start penning hate mail, please note: in actual fact, no animals were ever shocked or otherwise harmed in the test.
But during the experiment this was never explained to the participants, who were given the impression their decision would result in electric shocks being applied to at least one mouse, or at most five, depending on how they chose to react.
Ultimately, 84 percent of the participants who took part in the real-life test elected to press the button, sparing the five mice by consciously choosing to zap the other mouse – which, you might reason, results in fewer animals suffering overall (if they were receiving shocks, which they weren't).
What's interesting is that this real-life experiment didn't match up with another experiment run by the researchers, in which they asked a separate group of participants how they would react in the exact same situation. This time it was purely hypothetical, with no lab setup, mice, or electroshock machine actually present.
In that experiment, only 66 percent of people said they would zap the solitary mouse.
There are a number of limitations with the study, and the extent to which it fully embodies the trolley problem.
For a start, it's hard to ethically equate the prospect of human death with the experience of a mouse receiving an electric shock, and at least some of the participants involved in the experiment later said they saw through the researchers' setup, understanding no animals would be harmed.
But to the extent that it explores the trolley problem, the results suggest that, in the heat of the moment, more of us lean towards consequentialism (based on the overall outcome) than deontological thought (which argues it would be immoral to act to hurt the one mouse, despite the overall outcome), than we might otherwise think.
Hmm. Lots of tricky questions, and no clear answers. What do you think you would do?
The findings are reported in Psychological Science.