Under many circumstances, children and adult rats reorient themselves throu
gh a process which operates only on information about the shape of the envi
ronment (e.g., Cheng, 1986; Hermer & Spelke, 1996). In contrast, human adul
ts relocate themselves more flexibly, by conjoining geometric and nongeomet
ric information to specify their position (Hermer & Spelke, 1994). The pres
ent experiments used a dual-task method to investigate the processes that u
nderlie the flexible conjunction of information. In Experiment 1, subjects
reoriented themselves flexibly when they performed no secondary task, but t
hey reoriented themselves like children and adult rats when they engaged in
verbal shadowing of continuous speech. In Experiment 2, subjects who engag
ed in nonverbal shadowing of a continuous rhythm reoriented like nonshadowi
ng subjects, suggesting that the interference effect in Experiment 1 did no
t stem from general limits on working memory or attention but from processe
s more specific to language. In further experiments, verbally shadowing sub
jects detected and remembered both nongeometric information (Experiment 3)
and geometric information (Experiments 1, 2, and 4), but they failed to con
join the two types of information to specify the positions of objects (Expe
riment 4). Together, the experiments suggest that humans' flexible spatial
memory depends on the ability to combine diverse information sources rapidl
y into unitary representations and that this ability, in turn, depends on n
atural language. (C) 1999 Academic Press.