2.4 Spatial Cognition
Human societies vary in their linguistic tools for, and cultural practices associated with,
representing and communicating (1) directions in physical space, (2) the color spectrum, and (3)
integer amounts. There is some evidence that each of these differences in cultural content may
influence some aspects of non‐linguistic cognitive processes (D’Andrade 1995, Gordon 2005,
Kay 2005, Levinson 2003). Here we focus on spatial cognition, for which the evidence is most
provocative. As above, it appears that industrialized societies are at the extreme end of the
continuum in spatial cognition. Human populations show differences in how they think about
spatial orientation and deal with directions, and these differences may be influenced by
linguistically‐based spatial reference systems.
Speakers of English and other Indo‐European languages favor the use of an egocentric (relative)
system to represent the location of objects relative to the self (e.g., “the man is on the right
side of the flagpole”). In contrast, many if not most, languages, favor an allocentric frame which
comes in two flavors. Some allocentric languages such as Guugu Yimithirr (an Australian
language) and Tzeltal (a Mayan language) favor a geocentric system in which absolute
reference is based on cardinal directions (“the man is west of the house”). The other allocentric
frame is an object‐centered (intrinsic) approach that locates objects in space, relative to some
coordinate system anchored to the object (“the man is behind the house”). When languages
possess systems for encoding all of these spatial reference frames, they often privilege one at
the expense of the others. However, the fact that some languages lack one or more of the
reference systems suggests that the accretion of all three systems into most contemporary
languages may be a product of long‐term cumulative cultural evolution.
Weird People 5-Mar-09
In data on spatial reference systems from 20 languages drawn from diverse societies—including
foragers, horticulturalists, agriculturalists, and industrialized populations—only three languages
relied on egocentric frames as their single preferred system of reference. All three were from
industrialized populations: Japanese, English and Dutch (Majid et al. 2004).
The presence of, or emphasis on, different reference systems may influence non‐linguistic
spatial reasoning (Levinson 2003). In one study, Dutch and Tzeltal speakers were seated at a
table and shown an arrow pointing either to the right (north) or the left (south). They were
then rotated 180 degrees to a second table where they saw two arrows: one pointing to the left
(north) and the other one pointing to the right (south). Participants were asked which arrow on
the second table was like the one they saw before. Consistent with the spatial‐marking system
of their languages, Dutch speakers chose the relative solution, whereas the Tzeltal speakers
chose the absolute solution. Several other comparative experiments testing spatial memory
and reasoning are consistent with this pattern, although lively debates about interpretation
persist (Levinson et al. 2002, Li & Gleitman 2002).
Extending the above exploration, Haun and colleagues (2006, 2006) examined performance on
a spatial reasoning task similar to the one described above using children and adults from
different societies and great apes. In the first step, Dutch‐speaking adults and eight‐year olds
(speakers of an egocentric language) showed the typical egocentric bias, whereas Hai//om‐
speaking adults and eight‐year olds (a Namibian foraging population who speak an allocentric
language) showed a typical allocentric bias. In the second step, four‐year old German‐speaking
children, gorillas, orangutans, chimpanzees, and bonobos were tested on a simplified version of
the same task. All showed a marked preference for allocentric reasoning. These results suggest
that children share with other great apes an innate preference for allocentric spatial reasoning,
but that this bias can be overridden by input from language and cultural routines.
If one were to work on spatial cognition exclusively with WEIRD subjects (say, using subjects
from Japan, the U.S. and Europe) one might conclude that children start off with an allocentric
bias but naturally shift to an egocentric bias with maturation. The problem with this conclusion
is that it would not apply to many human populations, and may be the product of particular
cultural environments. The next contrast highlights some additional evidence suggesting that
WEIRD people may be unusual in their egocentric bias.
We have discussed several lines of data suggesting, not only population‐level variation, but that
WEIRD people are unusual. There are also numerous studies that have found differences
between much smaller numbers of samples. In these studies it is impossible to discern whether
the results of the small‐scale societies or those of the industrialized societies are more unusual.
For example, one study found that both samples from two different industrialized populations
were risk‐averse decision makers when facing monetary gambles involving gains (Henrich &
Weird People 5-Mar-09
McElreath 2002) while both samples of small‐scale societies were risk‐prone. It might be that
such risk‐aversion is a local phenomenon. Similarly, extensive inter‐temporal choice
experiments using a panel method of data collection indicates that the Tsimane, an Amazonian
population of forager‐horticulturalists, discount the future 10 times more steeply than WEIRD
people (Godoy et al. 2004).
2.4 Spatial Cognition