The principle proposed by K. Cheng and C. R. Gallistel (see record <RelatedUID>2005-03796-012</RelatedUID>) has the merit of explaining empirical data on reorientation of chicks and rats, invoking the simple computation of a global shape parameter: the extraction of the principal axis. However, despite its elegance and economy, it is inconsistent with previous observations on the geometric determinants encoded and used by both chicks and rats in place-learning tasks based on the shape of the environment. Further investigations are necessary to clarify whether different geometric features are used in different spatial tasks (i.e., reorientation vs. place learning). (PsycINFO Database Record (c) 2010 APA, all rights reserved)