Nobody can see a thought, but they can see one end.
For most of its history, research on thinking relied on people describing their own thoughts. A 2020 brain-scan study tried something new: instead of reading what people think, it spotted the moment they moved on to something else.
AI-generated image, not a photograph of a real person or place.
Thoughts turn over all day
Pushing one away can backfire
Inner voices vary a lot
What the studies measured
THE SHORT ANSWER
Thousands a day, and hard to steer.
One study counted when thoughts change. Another showed what happens when we fight them.
THE COUNT
About 6,200 a day
Queen's University researchers estimated that the average person has about 6,200 thoughts per day, based on detecting when brain activity moves to a new pattern.
Pick a research method to see how it works. The bar shows how much it relies on people describing their own thoughts.
Summarised from the published studies listed in the sources.
TWO NUMBERS WORTH KNOWING
A busy mind, measured in scans.
One estimate of how often thoughts change, one on how many brains it came from.
≈ 6,200
Estimated thoughts a day, average person
The estimate comes from counting transitions between brain states. The researchers found the transition rate was stable across watching films and resting, and used it to approximate the rate at which thoughts turn over.
THOUGHTS PER DAY · QUEEN'S UNIVERSITY
184
People whose brain scans were analysed
The study used scans from 184 participants in the Human Connectome Project, recorded while they watched films and while they rested in the scanner.
THE DATA · NATURE COMMUNICATIONS
The researchers also related the resting transition rate to trait neuroticism, which they describe as a first neural confirmation of the 'mental noise' theory.
FOUR THINGS THE STUDIES SUGGEST
What the research does and doesn't say.
These are findings from specific experiments, not rules for living. Each comes with the limits of the method behind it.
Thoughts change often
The 6,200 figure counts changes of thought, detected indirectly, not words or sentences.
Suppression can rebound
A 2001 meta-analysis of controlled studies found a small to moderate rebound after people tried to push a thought away.
A distraction helped
In the second white bear experiment, people given a specific thought to use as a distracter were less preoccupied with the bear later.
Inner speech isn't universal
Experience sampling found inner speaking in many, but not all, moments, with large differences between people.
AI-generated image, not a photograph of a real person or place.AI-generated image, not a photograph of a real person or place.
THE WHITE BEAR
Why pushing a thought away can backfire.
In Wegner and colleagues' 1987 experiment, people asked not to think of a white bear mentioned it and rang the bell anyway. When later told to think about the bear, they did so more than people who had been asked to think about it from the start.
A 2001 meta-analysis found the rebound effect was real but small to moderate, and varied with the kind of thought and how it was measured. People with clinical diagnoses did not show larger rebounds, though few studies included them.
Questions
Questions people ask
How many thoughts do we have a day?
One 2020 study estimated about 6,200 for the average person, by detecting when brain activity shifts to a new pattern. It is an estimate of thought transitions from scans, not a count of words.
What are thought worms?
The researchers' name for a run of adjacent brain states while a person stays on the same idea. When someone moves to a new thought, a new worm begins, and the method detects that change.
Does trying not to think about something work?
In the 1987 white bear study it backfired: people could not suppress the thought and thought about it more afterwards. A 2001 meta-analysis found a small to moderate rebound overall.
Does everyone hear an inner voice?
Not all the time. Descriptive Experience Sampling studies found inner speaking in many, though not all, sampled moments, and large differences between individuals in how often it happens.
Can a scan tell what I am thinking?
The thought worm method deliberately does not try. It detects when a new thought begins, without regard to what the thought is about.