Spatial cognition is fundamental to human behavior, but people differ in how they remember spatial relations, variably using body-based (egocentric) and environment-based (allocentric) spatial reference frames. Despite decades of study, the causes of this variation and flexibility in spatial memory remain unclear. Here we show that people spontaneously use different reference frames on different spatial axes at the same time. When remembering the placement of a target object in a 2-dimensional array, Indigenous Tsimane' adults preferentially used allocentric space to determine lateral placement and egocentric space to determine sagittal placement in the same action. This effect of axis was also significant among US university students, whose overall preference for egocentric space was stronger on the sagittal than lateral axis. These findings support a novel account of spatial cognitive diversity and suggest that people across cultures habitually integrate egocentric and allocentric spatial reference frames into the same action.

People implicitly map numbers onto space, but what aspect of numbers do people spatialize? When cardinality (i.e. mag-nitude; 5 objects) is pitted against ordinality (i.e., sequential position; the 5th object), people show an implicit ordinalitymapping, at least in lateral space. We hypothesized that if people spatialize numerical magnitude at all, they should do soon the vertical axis, according to the way they talk about numbers (i.e. low, high). Participants memorized sequences ofrandomized numbers (e.g. 85913) and then classified them (as small or large) using two response keys, oriented either lat-erally or vertically. Participants showed reliable ordinality mappings on both axes; they were faster to press the left/upperkey for numbers earlier in the memorized sequence and the right/bottom key for later numbers, regardless of numbersmagnitudes. People map exact numbers onto both lateral and vertical space according to their ordinality.

People conceptualize both time and numbers as unfoldingalong a horizontal line, either from left to right or from rightto left. The direction of both the mental timeline (MTL) andthe mental number line (MNL) are widely assumed to dependon the direction of reading and writing within a culture.Although experimental evidence supports this assumptionregarding the MTL, there is no clear evidence that readingdirection determines the direction of the MNL. Here wetested effects of reading experience on the direction of boththe MTL and MNL. Participants read English text eithernormally (from left to right) or mirror-reversed (from right toleft). After normal reading, participants showed the space-time associations and space-number associations typical ofWesterners. After mirror reading, participants’ space-timeassociations were significantly reduced but their space-number associations were unchanged. These results suggestthat the MTL and MNL have different experiential bases.Whereas the MTL can be shaped by reading experience, theMNL is shaped by other culture-specific practices throughwhich people experience numbers arrayed in left-right space.

People implicitly associate different emotions with differentlocations in left-right space. Which dimensions of emotion dothey spatialize? Across many studies people spatializeemotional valence, mapping positive emotions onto theirdominant side of space and negative emotions onto their non-dominant side. Yet, other results suggest a contradictorymapping of emotional intensity (a.k.a., emotional magnitude),according to which people associate more intense emotionswith the right and less intense emotions with the left, regardlessof valence. To resolve this apparent contradiction, we firsttested whether people implicitly spatialize whicheverdimension of emotion they attend to. Results showed thepredicted valence mapping, but no intensity mapping. We thentested an alternative explanation of findings previouslyinterpreted as showing an intensity mapping; these data mayreflect a left-right mapping of spatial magnitude, not emotion.People implicitly spatialize emotional valence, but there is noclear evidence for an implicit lateral mapping of emotionalintensity.

Spatial cognition is central to human behavior, but the way we conceptualize space varies over development and across cultures. When remembering the locations or movements of nearby objects, educated adults predominantly rely on a body-based spatial reference frame (e.g. to the left), whereas other groups prefer environment-based frames (e.g. toward the road), at least in some contexts. We propose that this variation in spatial thinking partly reflects differences in the ability to reliably discriminate left-right space, an ability that is common only among educated adults. To evaluate this proposal, here we tested US children’s spontaneous use of spatial reference frames on two axes. On the front-back axis, where spatial discrimination was relatively high, participants remembered object locations and movement directions using a body-based reference frame. On the left-right axis, where their spatial discrimination was significantly worse, the same participants preferred environment-based reference frames. This reversal reveals remarkable flexibility in children’s spontaneous use of spatial reference frames, extends findings in indigenous adults, and clarifies the likely mechanisms underlying spatial cognitive diversity.

People use space to conceptualize abstract domains like timeand number. This tendency may be a cognitive universal, butthe specifics of people’s implicit space-time and space-numberassociations vary across cultures. How does culture shape ourabstract concepts? In Western cultures, both time and numbersare arranged in people’s minds along an imaginary horizontalline, from left to right, but in other cultures the directions of themental timeline (MTL) and mental number line (MNL) arereversed. The directions of both the MTL and MNL have longbeen assumed to depend on the direction in which people readand write text. Here we argue that this assumption is false, andshow that the MTL and MNL are shaped by different aspectsof cultural experience. In a training experiment, participantsspatialized time and numbers in opposite directions across theirfingers. Training changed the MTL and MNL in oppositedirections, as predicted by a general principle called theCORrelations in Experience (CORE) principle: peoplespatialize abstract conceptual domains in their minds accordingto the ways these domains are spatialized in their experience.

The ability to represent large exact numbers is unique to humans. On some proposals, this capacity depends crucially onlanguage; learning the count list (”one”, ”two”, ”three”, etc.) allows children to represent the exact cardinality of numberslarger than four. On alternative proposals, this ability depends not on language but on innate pre-verbal counting processes.Here, we conducted a non-verbal test of large exact number concepts in the Tsimane’, an indigenous Amazonian culturein which adults vary widely in their knowledge of the verbal count list. Participants correctly matched the number ofobjects in a response set to the number in a sample set but only for cardinalities that were within their verbal count range.For larger cardinalities, they reproduced sets that were only approximately matched in number. The findings challengeaccounts that posit pre-verbal number concepts and support the Whorfian view that language can enable new conceptualabilities.

The physical properties of space may be universal, but the way people conceptualize space is not. In some groups, people tend to use egocentric space (e.g. left, right) to encode the locations of objects, while in other groups, people encode the same spatial scene using allocentric space (e.g. upriver, downriver). These different spatial Frames of Reference (FoRs) characterize the way people talk about spatial relations and the way they think about them, even when they are not using language. Although spatial language and spatial thinking tend to covary, the root causes of this variation are unclear. Here we propose that this variation in FoR use reflects the spatial discriminability of the relevant spatial continua. In an initial test of this proposal in a group of indigenous Bolivians, we compared FoR use across spatial axes that are known to differ in discriminability. In two non-verbal tests, participants spontaneously used different FoRs on different spatial axes: On the lateral axis, where egocentric (left-right) discrimination is difficult, their behavior was predominantly allocentric; on the sagittal axis, where egocentric (front-back) discrimination is relatively easy, their behavior was predominantly egocentric. These findings support the spatial discriminability hypothesis, which may explain variation in spatial concepts not only across axes, but also across groups, between individuals, and over development.