Handbook of Psychology, Volume 4: Experimental Psychology

Haptic Perception of Two- and Three-Dimensional Patterns 163

3B

3A

1A

2A

6B 6A 8B

8A

7A

7B

1B

5B

5A

4A

4B

2B

EEEEEEEE

BBBBBBBBB

HHHHHHHHHHH

IIIIIIIIIIIIIIIIIIII

IIIIIIIIIIIIIIIIII

HHHHHHHHHHH

EEEEEEEE

BBBBBBBB

AAAAAAAAA

IIIIIIIIIII

AAAAAAAAA

IIIIIIIIIII

Figure 6.6 Objects used by Lakatos and Marks (1999, Figure 2) and their similarity as determined by a clustering algorithm.

The algorithm distributes the objects in a tree, such that objects at the end of a common branch are similar. Letters on the

branches indicate features that are found on objects falling on different branches (e.g., the letter Adenotes sharp protrusions,

as found on objects 1B and 2B).

haptically than when vision was available. Longer exposure
time (increasing from1sto16s)produced greater similarity
ratings for objects that were locally different but globally sim-
ilar, indicating the increasing salience for global shape over
time.
When people do extract local features of three-
dimensional objects, they appear to have a bias toward en-
coding the back of the object—the reverse of vision. Newell,
Ernst, Tian, and Bülthoff (2001) documented this phenome-
non using objects made of Lego blocks. The participants
viewed or haptically explored the objects, and then tried to
recognize the ones to which they had been exposed. On some
trials, the objects were rotated 180 (back-to-front) between
exposure and the recognition test. When exposure and test
were in the same modality (vision or touch), performance

suffered if the objects were rotated. When the modality

changed between exposure and test, however, performance

was better when the objects were rotated as well: In this case,

the surface that was felt at the back of the object was viewed

at the front. Moreover, when exploration and testing were

exclusively by touch, performance was better for objects

explored from the back than for those explored from the

front.

Although the previously described studies emphasized the

role of shape, no doubt a critical factor in recognizing real,

common objects by touch is material. Material is locally

available, whereas extraction of the shape of an object

requires following its contours or enclosing it in the hand

(Lederman & Klatzky, 1987). A number of studies by

Klatzky, Lederman, and associates point to the importance of