Nine dots, arranged in a square grid, and a single instruction: connect all nine using only four straight lines, without lifting the pen. Try it with the assumption that the lines must stay inside the square the dots seem to form, and it cannot be done. The psychologist J.P. Guilford ran this puzzle as a lab exercise in the early 1970s and found that only around one in five people managed to solve it in the time given, because almost everyone assumed a boundary that the puzzle never actually stated.
Where the phrase actually came from
This is the puzzle behind “thinking outside the box.” The earliest known written use of the phrase appears in a 1971 issue of the journal Data Management, and the nine-dot puzzle became the go-to illustration for consultants and management trainers explaining why people get stuck: they impose limits that were never actually there, and the fix is to notice the imagined wall and step past it.
It is a tidy story. It is also, according to a later and more careful test of the same puzzle, not quite what is going on.
What happened when researchers just told people to go outside the lines
If the imagined boundary really is the whole problem, then telling people plainly that the solution requires lines outside the square should make the puzzle much easier. In 1969, the psychologists Clarke Burnham and Kenneth Davis tested exactly that. Participants who were told outright that they were free to draw beyond the edges of the square solved the puzzle only marginally more often than participants given no such hint, a difference small enough that it could plausibly be explained by chance rather than by the instruction itself.
This is one classic study, not the final word on insight problems generally, but it is a finding that later research on the nine-dot puzzle has kept coming back to. Removing the assumed boundary, in Burnham and Davis’s data, barely moved the solve rate. What did move it, in later work on the same puzzle, was giving people two extra dots positioned outside the original square, which dramatically increased how many people found the solution. The real obstacle looks less like an imaginary wall and more like the sheer length and specific order of moves the solution requires, something simply being told to think more broadly does not hand a person for free.
A stranger, more physical version of the same question
The phrase itself got a second, unrelated test decades later. A 2012 study in Psychological Science, by Angela Leung and colleagues, built an actual five-foot cube out of pipes and cardboard and asked some participants to solve creative word problems while sitting inside it, and others to solve the same problems sitting just outside it. The people sitting outside the box produced more correct, more original answers than the people sitting inside it.
This is a different mechanism from the nine-dot puzzle entirely. Leung’s team was testing embodied cognition, the idea that physically enacting a metaphor can shift the mental state associated with it, not testing whether removing an imagined constraint helps a specific problem. The two studies sit oddly next to each other: verbally telling people the constraint does not exist barely helped with the puzzle that gave the phrase its name, while physically acting out the metaphor of being outside a box measurably helped with a separate set of creative tasks. The phrase captures something real about embodiment, on this evidence, but not the specific mechanism the nine-dot puzzle is usually used to illustrate.
The Danish version of the same argument runs the other way
One of the more sustained bodies of applied research on creativity and constraint has come out of Billund, a small town in Denmark that is home to the LEGO Group. Since the 2000s, the company’s research arm, now the LEGO Foundation, has funded and published work on play and creative thinking, including a 2013 report called Cultures of Creativity, co-authored by the researcher David Gauntlett and the LEGO Foundation’s Bo Stjerne Thomsen.
The interesting thing about LEGO bricks, from a research standpoint, is that they are a fixed system of interlocking constraints. A brick only connects in specific ways, at specific angles, to a specific range of other pieces. The creativity the system produces does not come from removing that structure. It comes from working skilfully inside it, discovering how much variety a small, closed set of possible connections can still generate. The most visible Danish contribution to this whole conversation about creative thinking is not really about stepping outside any box. It is a business built on the opposite claim: that a firmly bounded system, used well, can produce more surprising results than an unbounded one.
What actually seems to help, on the evidence so far
None of this settles what “thinking outside the box” should mean in practice, and the research here comes from a handful of separate studies rather than one unified body of evidence. But a pattern worth taking seriously runs through all of it: simply being told that a limit does not exist, on its own, does not appear to do much. What changed outcomes, across these different experiments, was something more concrete: extra information that made a hidden path visible in the nine-dot case, a physical shift in posture in the embodied-cognition case, or an actual, well-designed set of constraints to work within in the LEGO case.
The instruction to think outside the box is not wrong so much as incomplete. On the evidence collected so far, what tends to move a person past a stuck point is rarely the general permission to ignore limits. It is a more specific change, in the information available, the body, or the structure of the problem itself.