The Divergent Thinking Test
In The You On AI Field Guide
The test's design was deceptively simple. Land and Jarman had been commissioned by NASA to identify engineers and scientists with exceptional capacity for generating multiple solutions to novel problems. The instrument they developed measured four dimensions of divergent cognition: fluency (number of responses), originality (statistical rarity), elaboration (depth of development), and flexibility (willingness to cross categories). A genius-level score required not just many answers but answers that violated the implicit premise of the question—answers that imagined the paperclip as two hundred feet tall, made of foam, alive.
The longitudinal data produced a curve so consistent that Land called it 'non-creative behavior learned.' The same children, the same brains, showed a monotonic decline in divergent scoring across the years they spent in school. The decline correlated not with biological maturation but with institutional exposure. Children who had spent more years in standardized educational settings showed steeper declines than children who had not. The test was measuring something the system was actively producing.
Robinson's reading of the data was structural rather than psychological. The capacity for divergent thinking did not disappear; the confidence to use it did. Ten years of daily practice at producing single correct answers trained children that being wrong carried costs—the red mark, the lowered grade, the narrowed future. The architecture of assessment taught them that deviation was failure, and deviation is precisely what divergent thinking requires.
The data acquires new urgency in the age of large language models. The convergent competence that schooling developed at such cost is now performed by machines. The divergent capacities that schooling suppressed are the only capacities the economy cannot automate. The test results, which Robinson deployed for decades as philosophical argument, have become an economic diagnosis of civilizational scale.
Origin
Land and Jarman conducted the original NASA-commissioned study in 1968, publishing results that would later appear in Land's 1993 book Breakpoint and Beyond. The longitudinal protocol—testing the same cohort at ages five, ten, and fifteen—made the findings unusually robust. Subsequent replications by independent researchers across multiple countries produced similar decay curves, suggesting the phenomenon was institutional rather than cultural.
Robinson first cited the data in his 2001 book Out of Our Minds and featured it prominently in his 2006 TED talk, which became the most-watched TED talk in history. The figures traveled from academic literature into popular consciousness through his advocacy, acquiring the status of a diagnostic baseline that every educational reform argument now must address.
Key Ideas
Capacity is not lost, confidence is. The children retain the neural architecture for divergent thinking; what they lose is permission to use it in contexts that matter to them.
The curve is institutionally produced. The decline correlates with years of schooling, not with biological maturation—making the result a measurement of the system rather than of human development.
Four dimensions of divergence. Fluency, originality, elaboration, and flexibility together constitute what the test measures; standardized education develops none of them and actively erodes the fourth.
The economic inversion. The suppressed capacity is now economically scarce in ways the developed capacity is not, making the longitudinal decline a civilizational liability rather than an educational curiosity.
Debates & Critiques
The test has been criticized for cultural specificity—whether divergent fluency means the same thing across populations—and for the question of whether declining scores reflect increasing sophistication rather than decreasing creativity. Some critics argue that adults filter responses through judgments about social appropriateness, producing apparently lower scores that reflect maturation rather than suppression. Robinson's response was that the filtering itself is the suppression: the capacity to generate without judging is precisely what creative work requires and what adults have lost.
In The You On AI Book
Further Reading
- George Land and Beth Jarman, Breakpoint and Beyond: Mastering the Future Today (HarperBusiness, 1993)
- Ken Robinson, Out of Our Minds: Learning to Be Creative (Capstone, 2001; revised 2011)
- Ken Robinson, "Do Schools Kill Creativity?" TED talk (February 2006)
- E. Paul Torrance, Torrance Tests of Creative Thinking—the lineage of instruments from which Land and Jarman's protocol descended
- Kyung Hee Kim, "The Creativity Crisis: The Decrease in Creative Thinking Scores on the Torrance Tests of Creative Thinking" (Creativity Research Journal, 2011)