266 • CHAPTER 9 Knowledge
The spreading activation feature of the model is sup-
ported by priming experiments.- The Collins and Quillian model has been criticized for
several reasons: It can’t explain the typicality effect,
the idea of cognitive economy doesn’t always hold,
and it can’t explain all results of sentence verification
experiments. - Collins and Loftus proposed another semantic network
model, designed to deal with criticisms of the Collins
and Quillian model. This model was, in turn, criticized
because it was so flexible that it could explain any result. - The connectionist approach proposes that concepts are
represented in networks that consist of input units, hid-
den units, and output units. Information about concepts
is represented in these networks by a distributed activa-
tion of these units. This approach is therefore also called
the parallel distributed processing (PDP) approach. - Connectionist networks learn the correct distributed pat-
tern for a particular concept through a gradual learn-
ing process that involves adjusting the weights that
determine how activation is transferred from one unit to
another.- Connectionist networks have a number of features that
enable them to reproduce many aspects of human con-
cept formation. - The idea that concepts are represented by specialized
brain areas has been supported by single neuron record-
ing (Freedman’s monkey experiments), neuropsychologi-
cal evidence (category-specific knowledge impairments),
and by the results of brain scanning experiments in
humans (animals versus tools). The conclusion from this
evidence is that knowledge about concepts is distributed
over many areas of the brain. - Newborn infants are capable of crude categorization.
The familiarity/novelty preference procedure has been
used to determine the development of categorization
from global to basic to specific between 2 and 7 months
of age. Further learning during childhood adds more spe-
cific knowledge to categories.
Think ABOUT IT
- In this chapter we have seen how networks can be con-
structed that link different levels of concepts. In Chapter
7 we saw how organizational trees can be constructed
that organize knowledge about a particular topic (see
Figures 7.5 and 7.6). Create a tree that represents the
material in this chapter by linking together things that
are related. How is this tree similar to or different from
the semantic network in Figure 9.12? Is your tree hier-
archical? What information does it contain about each
concept? - Do a survey to determine people’s conception of “typi-
cal” members of various categories. For example, ask sev-
eral people to name, as quickly as possible, three typical
“birds” or “vehicles” or “beverages.” What do the results
of this survey tell you about what level is “basic” for
different people? What do the results tell you about the
variability of different people’s conception of categories?- Try asking a number of people to name the objects
pictured in Figure 9.10. Rosch, who ran her experi-
ment in the early 1970s, found that the most common
responses were guitar, fish, and pants. Notice whether
the responses you receive are the same as or different
from the responses reported by Rosch. If they are differ-
ent, explain why you think this might have occurred.
If You WANT TO KNOW MORE
- More on concepts. If you want to read more about con-
cepts, see The Big Book of Concepts, which starts by
asserting that “concepts are the glue that holds our men-
tal world together.”
Murphy, G. (2004). The big book of concepts. Cambridge, MA:
MIT Press.- Culture and categorization. Cross-cultural research on
members of the Itza culture indicates that culture can
affect which level of categories is considered basic. Thus,
a basic category for members of one culture may differ
from what is basic for members of another culture.
Medin, D. L., & Atran, S. (2004). The native mind: Biological
categorization and reasoning in development and across cul-
tures. Psychological Review, 111, 960–983.- Personal and institutional categories. People and major
institutions create their own categories, some of which
apply only to them individually. This type of categoriza-
tion is related to the increased use of the Internet.
Gleshko, R. J., Maglio, P. P., Matlock, T., & Barsalou, L. W.
(2008). Categorization in the wild. Trends in Cognitive Sci-
ences, 12, 129–135.Copyright 2011 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part.