Perceptual and Motor Skills, 2011, 113, 3, 881-893. © Perceptual and Motor Skills 2011
DOI 10.2466/22.24.27.PMS.113.6.881-893 ISSN 0031-5125
INTERACTIVE PROCESSING AMONG PARTS, CONFIGURATION,
AND ORIENTATION OF COMPLEX OBJECTS
1, 2
SHIQI LI AND GUOMEI ZHOU
Department of Psychology
Sun Yat-sen University
Summary.—The aim of the present study was to investigate whether the rep-
resentation of parts and the confguration and orientation of complex objects were
separate or interdependent. Participants were asked to match parts and confgura-
tion of novel 3-D objects with constant or varied features in irrelevant dimensions.
The results of both Exp. 1 (sequential matching) and Exp. 2 (simultaneous match-
ing) showed that both sensitivity and response time (RT) of part matching were
infuenced by whether the irrelevant confguration remained constant or varied.
Both sensitivity and RT for confguration matching were afected by the variation of
irrelevant parts in Exp. 2, but only sensitivity was afected in Exp. 1. These results
demonstrated an interdependent patern of representations of parts and relations
but diferent interactive paterns in memory and perception. The results of Exp. 3
showed orientation dependency in a confguration-matching task but orientation in-
variance in a part-matching task. Consistent with a study done in 2000 by Edelman
and Intrator, these results suggest an interactive processing between confguration
and orientation but an independent processing between the parts and orientation.
When one sees a cup, what information must be detected for immedi-
ate or later discrimination? Some researchers (e.g., Biederman, 1987) have
suggested that an object is mentally represented as a structure of parts and
their spatial relations. Recognizing a cup requires independent coding of
the shapes of parts (“curved” handle and “cylinder” body) and confgural
information (“on the side of”). Thus, the cup may be represented as “on
the side of (‘curved’ handle, ‘cylinder’ body).” In contrast, other research-
ers (e.g., Tarr & Pinker, 1989; Jolicoeur, 1990; Bülthof & Edelman, 1992;
Tarr, 1995) have suggested that objects are represented as the projections
onto the retina in a “template” representation. For them, an object is pro-
cessed holistically, without being parsed into parts and encoded into spa-
tial relations. Edelman and Intrator (2000) combined these two theories
and hold that objects are represented as local features in particular posi-
tions, in a kind of dependent coding. For example, a cup is represented as
“a ‘curved’ handle on the right/lef side of a ‘cylinder’ body.” According
to Edelman and Intrator, if the “curved” shape of the handle is changed,
1
Address correspondence to Guomei Zhou, Department of Psychology, Sun Yat-sen Univer-
sity, 135 Xingang Xi Road, Guangzhou, 510275 China or e-mail (zhougm@mail.sysu.edu.cn).
2
This article was supported by the National Natural Science Foundation of China (Grant
No. 30700230) and the 11th Five-year Plan in Philosophy and Social Sciences of Guangdong
Province (08SXQ001) to G. Z.