Lateral thinking is approaching a problem from an unexpected angle instead of working straight down the obvious line of reasoning. Edward de Bono coined the term in 1967. What almost nobody writing about it mentions is that its most famous demonstration — the puzzle you solve by drawing outside the square — was put to an experimental test, and the test did not go the way the story goes.
Where the term comes from
Edward de Bono, a Maltese physician and psychologist, introduced lateral thinking in 1967 in The Use of Lateral Thinking. He set it against what he called vertical thinking: the sequential, logical progression that takes the obvious starting point and works down from it. His image for the difference has outlived most of the rest of the theory — vertical thinking digs the same hole deeper, lateral thinking starts a hole somewhere else.
The methods he attached to it are the ones still taught in workshops today, and they are genuinely specific rather than vague encouragement:
| Technique | What you actually do |
|---|---|
| Challenging assumptions | List what you are treating as fixed, then ask what follows if each one is not |
| Random input | Take an unrelated word or object and force a connection to the problem |
| Provocation | State something deliberately absurd, then look for a usable idea on the way back from it |
| Reversal | Ask how you would make the problem worse, and invert the answers |
De Bono built a substantial training business around this and around later programmes such as Six Thinking Hats. That commercial success is part of why the idea is so widely known — and part of why the evidence question is so rarely asked in the places that explain it.
The nine-dot puzzle: what happened when researchers told people to think outside the square
The nine-dot problem — join nine dots in a three-by-three grid with four straight lines without lifting the pen — is the standard illustration of lateral thinking, and de Bono used it as one of his own examples. The standard explanation of why it is hard is that people are fixated on an assumption nobody imposed on them: that the lines must stay inside the square the dots form. Remove the assumption, the story goes, and the solution follows.
In 1981, Robert Weisberg and Joseph Alba published a series of experiments in the Journal of Experimental Psychology: General that tested exactly that. Rather than leaving people to discover the constraint was imaginary, they told them it was — subjects working on the nine-dot problem were informed that it could be solved only by extending their lines outside the boundaries of the square, and subjects working on a matchstick triangle problem were told to work in three dimensions.
The result, in the authors' own summary: “removing the alleged fixation did not result in sudden and direct solution of the problems, indicating that fixation was not a very important factor in making these problems difficult.” People who had been handed the insight still largely failed. What did help was something much less romantic — relatively detailed information about the solution itself. Weisberg and Alba concluded that the words fixation and insight were not useful descriptions of what was going on, and that problem-specific knowledge was doing the work.
This is worth sitting with, because it cuts against the whole shape of the popular advice. The puzzle everyone uses to prove that the barrier is a mental box turns out, when tested, to be a puzzle where telling people about the box does not get them out of it. Nothing on the first page of Google's results for lateral thinking mentions this experiment — the page is Wikipedia, a Reddit thread, de Bono's own organisation, a design company's resource library, a coaching vendor, a personal blog and three YouTube videos, all of which explain the technique and none of which report a test of it.
Lateral thinking, divergent thinking and convergent thinking
If you go looking for research on lateral thinking you will find very little, and it is easy to conclude the question has never been studied. It has — under a different name. Academic psychology measures divergent thinking: how many different responses a person generates to an open-ended prompt, scored for fluency (how many), originality (how unusual), flexibility (how many distinct categories) and elaboration (how developed). Its counterpart, convergent thinking, is what a conventional test of reasoning asks for — narrowing to the single defensible answer.
| Lateral thinking | Divergent thinking | Convergent thinking | |
|---|---|---|---|
| Where it lives | Books, workshops, business training | Psychological research and testing | Psychological research and testing |
| What it asks of you | Break the obvious frame deliberately | Produce many different responses | Find the one right answer |
| How it is scored | Not scored — it is a method, not a measure | Fluency, originality, flexibility, elaboration | Correct or incorrect |
| Relationship to IQ tests | Never established | Weak — see below | This is largely what they measure |
The practical consequence: when you read that “research shows creative thinking is separate from intelligence”, the research being referred to almost certainly measured divergent thinking. Claims about lateral thinking specifically are inherited from that literature rather than drawn from studies of de Bono's method. Our guide to critical thinking covers the convergent half of the same picture — the discipline of evaluating an answer rather than generating one.
Are creative thinkers more intelligent? The threshold question
The old idea here has a name: the threshold hypothesis. It holds that above-average intelligence is a necessary condition for high creativity but not a sufficient one — so the two should correlate below some cut-off, conventionally an IQ of 120, and come apart above it. It is repeated constantly. The evidence for it is unsettled, and has become less settled rather than more.
The largest pooled estimate comes from a 2005 meta-analysis that retrieved 447 correlation coefficients from 21 studies and 45,880 participants. The mean correlation between creativity test scores and IQ scores was r = .174 (95% CI .165–.183) — small, heterogeneous, and described by the author as negligible. Age contributed most to variation in the relationship, and which creativity test was used contributed second most. That analysis did not support threshold theory.
A 2013 study took a different approach: rather than splitting the sample at 120 because that is the number everyone uses, it used segmented regression to let the data locate any breakpoint itself, in 297 participants. It found thresholds only for measures of creative potential, not for creative achievement, and the threshold moved with how demanding the criterion was — around 100 IQ points for a liberal originality criterion, around 120 for a demanding one, and around 85 for ideational fluency.
The most recent large test points the other way again. A 2026 study in the Journal of Intelligence applied the same segmented-regression approach to 9,358 Chinese students aged six to twenty across twelve cohorts, controlling for cohort, gender, grade and age. The hypothesised breakpoint at IQ 120 produced non-significant results and failed to improve on a simple positive linear relationship.
Three studies, three answers — which is the honest summary. What none of them supports is the confident version you will hear at a conference: that there is a known IQ level above which intelligence stops mattering for creativity. There is a weak positive association, and a genuine dispute about whether it bends anywhere.
Does creativity training work?
This one has a clearer answer than the threshold question, and it is moderately encouraging. A 2004 meta-analysis reviewed 70 prior evaluation studies of creativity training programmes and found that well-designed ones typically produce gains in performance, with the effects generalising across criteria, settings and target populations, and holding when internal validity considerations were taken into account.
The interesting part is which programmes worked. The more successful ones focused on developing cognitive skills and the heuristics involved in applying them, using realistic exercises appropriate to the domain at hand. That is close to the opposite of a generic mindset seminar. It is also consistent with what Weisberg and Alba found twenty-three years earlier: what moves the needle is knowledge about the actual problem, not an instruction to think differently about problems in general.
Can it be measured? The Torrance tests and a fifty-year follow-up
The best-known instruments for divergent thinking are the Torrance Tests of Creative Thinking, administered from the late 1950s onwards. Because they were given to schoolchildren that early, they support something almost no measure in this field has: a fifty-year follow-up.
Published in 2010, it reported that test scores were moderately correlated with personal achievement but not with public achievement. What did relate to public achievement was an interaction of intelligence and creativity — neither one alone. Men scored significantly higher on public achievement than women, with no significant gender difference in personal achievement, a pattern that says at least as much about which achievements were publicly visible to that cohort as about the people tested.
So the honest position on measurement is: divergent thinking can be scored reliably, the scores predict something real about a life, and what they predict is modest and does not stand apart from ability.
Lateral thinking puzzles are a different thing
A large share of the traffic around this topic is looking for the puzzles — the situation riddles you solve by asking yes-or-no questions until an odd scenario resolves. They are enjoyable and they do exercise assumption-checking. They are not a test.
A test, in the technical sense, needs three things a riddle collection does not have: norms (how a representative sample performs, so your result means something relative to other people), reliability (the same person scoring roughly the same on a second occasion) and validity evidence (a demonstrated relationship to whatever it claims to measure). Lateral thinking puzzles have none of these published. Enjoy them; do not read a score off them. Our guide to how IQ tests work walks through what standardisation actually involves.
What to take from all this
Lateral thinking is a useful vocabulary and a set of specific, practisable prompts, and there is nothing wrong with using them. What it is not is a demonstrated distinct mental faculty that some people have and others lack. The nearest measured construct, divergent thinking, correlates only weakly with measured intelligence; whether that relationship has a threshold is genuinely disputed; training works best when it teaches something concrete about a real domain; and the flagship demonstration of the whole idea did not survive the one experiment that took it literally.
Trust and scope notes
This page is an explanation of published research, not a diagnostic tool. Our IQ test measures reasoning — the convergent side of the picture described above. It does not measure creativity, divergent thinking or lateral thinking, it produces no creativity score, and it cannot tell you whether you are a creative person.
IQ Revealed is not affiliated with, endorsed by or connected to Edward de Bono, the de Bono organisations or any programme trading under the names Lateral Thinking or Six Thinking Hats, nor with the publishers of the Torrance Tests of Creative Thinking. Those names are used here only to describe what they refer to.