In three experiments, the perception of the apparent orientation of block l
etters shown in various orientations above the subject's head in the horizo
ntal plane was examined. A block letter F with its front facing down toward
the observer has two crossbars on its right side; the top is the part with
the long crossbar, and the base has no bar. The experiments involved chang
ing the locations of these parts with respect to the observer. In Experimen
t 1, the subjects using touch most often identified a letter as having its
left and right sides in a normal orientation if the front of the block lett
er faced upwards away from the observer, with the bar on the right and the
top of the letter farther from the subject than the letter's base. In Exper
iment 2, the subjects judging visual uprightness favored positions in which
the bars were on the right, the top of the block letter was near them, and
the letter's front faced downwards toward the observer. In Experiment 3, t
he subjects using touch most often assessed letters as being upright if the
top of the letter was the farthest part and the bar was on the right. The
results suggest that, when assessing orientation, subjects using touch favo
red positions that would be reached by a letter moving vertically upwards f
rom table height, but subjects relying on vision favored positions reached
by a letter moving in an are centered on the subject's head ton the eyes, i
n particular).