Dahaene (Concepts across Cultures)

Study Population

An isolated Amazonian indigenous group - Most participants had little to no formal schooling and no experience with rulers, compasses, or maps + their language has few words dedicated to arithmetical, geometrical, or spatial concepts, although they use metaphors spontaneously

While they create symbolic miniatures and face paintings, they do not typically use maps or draw metric-accurate pictures of their environment

Research Question

Do the conceptual primitives of geometry constitute a "core" set of intuitions present in all humans regardless of culture

Experiment Hypothesis

If geometry is a universal "core knowledge," uneducated people from isolated cultures should perform similarly to those in Western cultures on basic geometric tasks

The Multiple-Choice "Intruder" Test Def (Test 1 in Experiment)

Participants were shown an array of six images; five instantiated a concept (e.g., parallelism) while one violated it (the "weird" or "ugly" one)

Tested → Topology (connectedness), Euclidean geometry (lines, points, right angles), and basic figures (squares, circles)

The Multiple-Choice "Intruder" Test Results (Test 1 in Experiment)

All participants (even 6-year-olds) performed well above chance → Performance was indistinguishable between Mundurukú adults and children, suggesting these intuitions do not require years of cultural development

Mundurukú (and young American children) struggled with geometrical transformations (mirror symmetry, rotation) that require mental transformations and "second-order" judgments

The Abstract Map Test Def (Test 2 in Experiment)

A 2D map used stars to represent a hidden object among three containers arranged in a triangle (right-angled or isosceles); To succeed, participants had to relate 2D map information to a 3D environment across changes in scale and orientation

The Abstract Map Test Results (Test 2 in Experiment)

To succeed, participants had to relate 2D map information to a 3D environment across changes in scale and orientation; Success averaged 71% (chance was 33.3%)

Performance remained significantly above chance regardless of whether the map was aligned with the environment (allocentric), the participant (egocentric), or rotated

While adults used landmarks (like a red box) when available, they still succeeded using purely geometrical information (distance, angle, sense) when landmarks were absent

Mundurukú vs. Americans

The performance of Mundurukú children and adults was identical to American children

Difficulty Profile for Experiments

There was a high correlation in the ordering of error rates across cultures → This means that the tasks Mundurukú found hard, Americans also found hard, suggesting a shared cognitive architecture for geometry

Impact of Culture

American adults performed at a higher absolute level, indicating that while the "core" is universal, knowledge is substantially enriched by cultural inventions like formal schooling and mathematical language

Rejection of Piaget’s Hypothesis

The data did not support a developmental progression from topology to Euclidean geometry; rather, these intuitions appeared simultaneously across cultures

What kind of knowledge is geometric knowledge?

Geometrical knowledge is a universal constituent of the human mind, similar to basic arithmetic

Can geometrical knowledge exist with or without a specialized lang to describe them?

Results provide evidence that geometrical intuitions can exist without a specialized language to describe them