A Quarter of the Dial

The Email and the Clock

Eli: Alfred Mele, a philosopher, kept reading news reports claiming scientists, especially neuroscientists, had shown there was no free will. As he later told a panel, he wondered what effect that had on people. Then a woman emailed him. She wrote that she could understand not knowing what thought would come next, "but that that has already happened is beyond comprehension." She was, in her words, "in a lot of despair." Mele wrote back. He was writing a book, he said, and he'd send it when it was done. Its last sentence would be "Don't worry. Be happy."

Mara: A listener asked us the same thing, plainly: do we really have free will, or does the brain decide before we know it? Today's angle is a philosopher's hands-on trial against the clock that claimed to kill free will.

Eli: Mele didn't stop at writing back. He went into the science himself, and at one point sat in a lab chair with electrodes on his own scalp. The puzzle he kept running into was a small electrical rise in the brain, and what it really means.

Mara: And it's more than one woman's peace of mind. Wiebke Melcher points out that free will is crucial for criminal law, "because one must be responsible for an act in order to be found guilty and punished for an act." If the brain settles a choice before awareness can step in, that reaches into a courtroom.

Eli: Mele had wondered what those headlines do to people, and a 2008 study looked at one effect. Some participants read material arguing free will doesn't exist. As it was put on that same panel, "the group that got the no free will news cheated significantly more often than the other groups." That shows the claim has consequences. Whether it's true is a separate question.

Mara: The other two groups behaved alike. That was read as free will being people's default.

Eli: So what was the evidence behind those headlines?

Mara: An experiment Benjamin Libet published in 1983. He tracked the readiness potential, a slow electrical buildup recorded over the brain before a movement. In the classic average, it started about half a second before the movement. The urge people reported came much later, about a fifth of a second before. Libet read that gap as a decision already under way. In Mele's retelling, Libet says, "the brain is unconsciously deciding to flex now and the mind becomes aware of it later." That reading is the contested heart of this whole story.

Eli: So the felt urge comes after?

Mara: Right, a voltage versus a report. Volunteers watched a spot sweep around a clock face, one lap in about two and a half seconds. That's the Libet clock, a way to put a time stamp on something private. They were told to "let the urge to act appear on its own at any time without any preplanning or concentration on when to act." After each flex, they said where the spot had been when the urge came. EEG, electrodes on the scalp reading the brain's electrical activity, ran the whole time, and so did a recording from the muscles.

Eli: So the mind side is a person's say-so?

Mara: Which means the whole conclusion leans on one thing: how well a person can pin a fleeting urge to a spinning spot.

Mele Tries It

Eli: Mele didn't accept that gap at face value. In 2009 he pulled apart order and cause. Every Libet experiment ended with a wrist flick, he wrote, but that didn't justify assuming the early rise caused the flick. Coming first isn't the same as causing. And he offered another possibility. The early signal "may simply mark the beginning of forming an intention to act." Two lab findings sharpen that. The urge time may be a retrospective report, a moment pieced together after the movement rather than read off live. When researchers delayed the feedback people got after acting, the reported intention slid later.

Mara: Then in 2010, Jeff Miller and Judy Trevena found the same readiness-potential signal whether volunteers tapped or didn't. So, they argued, it needn't mean a decision had been made, and it might reflect attention instead.

Eli: Same signal with a tap or without one, and a felt moment that can drift.

Mara: Which puts the weight back on that spinning clock, and on one person's sense of a moment.

Eli: Mele tried one himself. After a lecture on the neuroscience of free will at the National Institute of Health, they made him a subject in a Libet experiment. His method: silently say "now" to himself, treat that as the urge, flex, then report where the spot was at "now." About 40 times. The best he could do was place the spot within a stretch covering a fifth to a quarter of the clock face. And he said volunteers told to watch for an urge might wonder, "Was I just thinking about doing it, was it an urge?"

Mara: Volunteers timed their actual presses within a few milliseconds. Their intentions, not nearly as well.

Eli: So the felt moment blurs across a quarter of the dial, and the early signal still has to earn the word decision. Mara picks it up from there.

Buildup Toward a Threshold

Mara: Here's the brain-first reading at its strongest, the one associated with Libet and Daniel Wegner. Later work moved to fMRI, imaging that tracks changes in blood oxygen. Patterns there predicted simple choices before people reported deciding. One team in 2011 replicated an earlier finding: intentions encoded up to seven seconds before awareness. The accuracy ran around 60 percent. A coin toss gets half.

Eli: Seven seconds is a long head start. Mara, the other reading belongs to Aaron Schurger and Mele.

Mara: Remember the signal that showed up with no tap at all. Here's one way that could happen. In 2012 Schurger's team proposed a stochastic accumulator, brain activity drifting randomly and adding up toward a cutoff. That cutoff is the action threshold, where the system commits to moving. Now, the readiness potential is found by movement-aligned averaging: lining up many recordings at the moment of movement and looking backward. Do that with drifting noise, and the average necessarily shows a gradual climb to the threshold, even if no single trial has a ramp.

Eli: So activity can creep toward a line without being the go signal. In the model, crossing is the go.

Mara: In the model, yes. It generates the curve; it doesn't prove that's what happens on every trial. Prediction has limits too. Marcel Brass and his colleagues could predict which button a person would press ten seconds ahead, but barely better than guessing. Ten seconds out, the brain data sat close to a coin toss. Brass said, "You can't deterministically predict what people are going to do until very close to their decision."

Eli: Schurger's team also tested the idea. During a Libet-style wait, volunteers heard an unexpected tone telling them to move. When their ongoing brain activity happened to be riding high already, in that negative-going swing, they reacted faster. That's what the drift-and-threshold account predicts. It fits the model. It doesn't prove the early signal always means noise.

Mara: So a random swing set the reaction speed, just as the model needs.

Eli: Mele's correspondent couldn't fathom that her action had "already happened." What nobody can yet say is whether one person's rise, on one trial, is drift or a command. And the sharper test is whether someone can stop once that buildup is under way.

Stopping Before the Line

Mara: Libet himself saw room inside his own numbers. Between the reported urge and the movement sat roughly 150 to 200 milliseconds. In that sliver, he argued, consciousness could veto the action, stop it before it runs. That idea has been called "free won't." The end of the window is the EMG, a recording of muscle activity that marks the instant the movement actually begins.

Eli: Before any test, there's a dispute over how far these tasks reach. Robert Sapolsky's reading treats simple motor timing as showing the basic dynamics of all voluntary action. Mele pointed out what the tasks are: arbitrary. No reason to prefer one moment or button over another. In his words, "Nothing hangs on it." The other reading, held by Uri Maoz among others, is that twitches like that aren't choices made for reasons, so they can't disprove free will.

Mara: And Libet's veto is something a lab can put to a direct test.

Eli: In 2016, a team did. Volunteers played a game: press a button, earn points. A brain-computer interface, a system reading brain activity directly, watched for their readiness potentials in real time and fired stop signals. The buildup started around a full second before muscle activity. Even so, people could still cancel after the buildup began, up to a point of no return, the cutoff after which stopping no longer worked. That came about 200 milliseconds before movement. Libet's veto, in this task, now had evidence and a measured deadline.

Mara: Then Uri Maoz and colleagues asked what happens when a choice matters. In 2019, volunteers decided which nonprofit should get a thousand dollars, pressing one key or the other. Before arbitrary presses, Libet's ramp showed up as usual. Before the donation choices, it was substantially reduced, or not statistically significant. A reviewer urged calling it diminished, not absent.

Eli: So the famous ramp weakens where a reason drives the choice, the kind Mele said the old tasks couldn't speak to.

Mara: Whatever freedom turns out to be, Eli, it now hinges on more than who starts a twitch.

Which Freedom Survives

Eli: Mele started wondering. Asked whether people have free will, he answered, "Yes. Yes they do." Then the catch: not everybody, he said, understands the expression free will the same way. For some it means not being coerced. For others, that the person could have done otherwise at the last moment.

Mara: Doing otherwise at the last moment sounds like that 200-millisecond window, but his yes is a contested one. Libet said his findings limited conscious control without ruling it out. The team behind the 2008 imaging study didn't claim to have ruled out free will either. Wiebke Melcher argues the question stays more a matter of belief than data. And as of 2026, by one account, scientists are far more skeptical that free will doesn't exist, and that brain scans will ever settle it.

Eli: Which is where the definitions come in. Compatibilists say freedom can coexist with physical brain processes, even with determinism. What matters to them is reason-responsiveness: whether an action flows from the person's own desires, reasons and character, not whether consciousness beat the neurons by some number of milliseconds. Another account calls free will a capacity: to bias one's own behavior toward reflectively endorsed preferences, the ones a person has thought over and approves.

Mara: So on that view, freedom is a capacity to weigh reasons and steer by them. Nobody has to consciously launch every motor signal for it to count.

Eli: The reports behind that email rested on arbitrary flicks. A buildup before a flick nobody cares about is one thing. Deliberate choices running with no conscious part in them is a far bigger claim, and those flicks never put it to the test.

Mara: Joshua Greene, on a 2014 panel, put it bluntly: "I don't for a second think that neuroscience has shown that our decisions are made before we do them." That's contested. Whether free will survives depends on which freedom is meant. Three things to watch. Whether single-trial recordings show drift or an early command. Whether arbitrary flicks carry over to reasoned decisions. And whether free will means launching a movement or answering to reasons.

Eli: For listeners curious to look into the science, read up on the readiness potential.

Mara: Follow science with us as researchers test what brain activity truly means for conscious agency.

Eli: That's our story for today. I'm Eli.

Mara: And I'm Mara. Thanks for listening to Angle, and until next time.