0
Some recent research appears to indicate that this really does happen. The research paper was titled Walking straight into circles and was carried out at the Max Planck Institute for Biological Cybernetics. You can view details of the paper at http://www.cell.com/current-biology/abstract/S0960-9822(09)01479-1 . The Abstract of the paper is:
Common belief has it that people who get lost in unfamiliar terrain often end up walking in circles. Although uncorroborated by empirical data, this belief has widely permeated popular culture. Here, we tested the ability of humans to walk on a straight course through unfamiliar terrain in two different environments: a large forest area and the Sahara desert. Walking trajectories of several hours were captured via global positioning system, showing that participants repeatedly walked in circles when they could not see the sun. Conversely, when the sun was visible, participants sometimes veered from a straight course but did not walk in circles. We tested various explanations for this walking behavior by assessing the ability of people to maintain a fixed course while blindfolded. Under these conditions, participants walked in often surprisingly small circles (diameter < 20 m), though rarely in a systematic direction. These results rule out a general explanation in terms of biomechanical asymmetries or other general biases [1,2,3,4,5,6]. Instead, they suggest that veering from a straight course is the result of accumulating noise in the sensorimotor system, which, without an external directional reference to recalibrate the subjective straight ahead, may cause people to walk in circles.
Is it true that "a lost person usually walks in a circle of about one square mile, always turning toward the dominant side"? I found that quote and try to find out more precise references about it. Can someone help? Thanks