rich and complex set of research issues (Bailenson, Shum,
Atran, Medin, & Coley, 2002; Medin, Ross, Atran, Bur-
nett, & Blok, 2002). In the present article, we examine
some of these in the context of judgments about global
geography. Specifically, we address how some biases in
global location judgments can be attributed to the cat-
egorical nature of geographical representations and the
processes that use them, whereas others are attributable to
cultural asymmetries in geographical knowledge.
Our previous research with Canadian participants
(Friedman & Brown, 2000a, 2000b; Friedman, Brown, &
McGaffey, 2002) indicated that their location estimates for
cities in the old and new worlds were based on a category-
driven system of plausible reasoning (Collins & Michal-
ski, 1989). Geographic categories were psychologically
distinct regions that could be independently influenced
by new information. Some countries had more than one
region, and some regions comprised either one or several
countries. The plausible reasoning framework assumes
that biases in judgments about global locations are mul-
tiply determined because they are influenced by accurate
and inaccurate beliefs about geographic regions acquired
over the lifespan from a variety of sources.
The key features of the data (Friedman & Brown, 2000a,
2000b) were that (1) Canadian participants divided North
America into four distinct regions (Canada, the northern
U.S., the southern U.S., and Mexico), (2) there were usu-
ally large boundary zones (gaps) between regions, (3) there
was little north–south discrimination among the estimates
within most regions, and (4) the estimates became more
biased as the cities’ actual locations were farther south.
Indeed, the average location estimate for most Mexican
cities was near the equator, which was an error of approxi-
mately 1,500 miles.
These four observations are consistent with the influ-
ence of categorical information on location estimates
(Brown, 2002; Friedman
