From: Grant Callaghan (grantc4@hotmail.com)
Date: Thu 06 Mar 2003 - 16:49:54 GMT
Do you see how, even among people that supposedly share a
culture (for
instance, we could concievably find ourselves talking about the
subject at a
departmental cocktail party) the meaning of the Illiad is
different
depending on what other memes are in our *individual* heads?
Best,
Reed
The act of "seeing" and thus "reading" begins with the act of taking what we
perceive and placing it in a category based on experience within a culture.
Everything else we do with what we see or hear begins here. The following
excerpt from the archives of the Cogsci department at UCSD explains what
they have learned about the process.
Perceiving a New Category:
The Neurobiological Basis of Perceptual Categorization
Kara D. Federmeier
Department of Cognitive Science
University of California, San Diego
Categorization is a ubiquitous cognitive phenomenon -- ubiquitous in more
than one way. The influences of categorization can be seen in nearly every
aspect of higher cognition. We categorize to determine what to pay attention
to, to talk and understand, to remember, and to reason. However,
categorization can also be observed at every level of cognitive processing.
From the earliest stages of sensory processing, neurons and groups of
neurons respond to some kinds of inputs and not to others, creating
perceptual categories that are the roots of all human experience. While
current approaches to the
study of categorization generally acknowledge the important relationship
between categorization and
other higher cognitive processes, these approaches have largely failed to
acknowledge that the roots of
categorization begin in early sensory processing and are thus deeply
entrenched in the functioning of the
human brain. To understand categorization -- and, thereby, significant
portions of all of higher cognition
-- it is useful to try to understand those roots.
Current theories and models of categorization have tended to define
categorization as a grouping process that operates on perceptual inputs.
Perception, defined as the extraction of the perceptual features of a
stimulus, is believed to precede categorization and to provide the input for
it. Grouping done on the basis of perceptual features is often termed
“perceptual categorization” while grouping done on the basis of more
abstract features has been termed “semantic categorization.” Note that the
term “concept” is sometimes conflated with the term “category”. While the
line between the two can be fuzzy, in general, a category is defined as a
group of items that are treated similarly, while a concept can be defined as
a flexible abstraction of a category (often linguistic). If an organism
treats a set of perceptually different items similarly by eating all of
them, for example, these items could be said to define the category “food”
for that organism. If that organism also has a representation in which
“food” is understood to be anything potentially edible, the organism might
then be said to also have a concept of food. Categories, not
concepts, will be dealt with in this paper.
These traditional definitions will be maintained throughout the paper.
However, the goal is to demonstrate that categorization happens via rather
than on perception and that one cannot be understood without reference to
the other. In particular, this paper focuses on the relationship between
visual processing and perceptual categorization done on the basis of visual
features. First, it is argued that, far from being a phenomenon unique to
human higher cognitive processing, categorization is a general solution to a
basic conflict that arises for any information processing system. Such
systems must be able to recognize and act on important (relevant)
differences between stimuli in their environment while still ignoring other,
irrelevant differences. A simple, effective solution to this conflict is to
divide stimuli into groups -- to categorize. In fact, human categorization
is structured in two dimensions to both maintain information and allow for
appropriate generalizations.
Having established categorization as a general solution to an information
processing problem, it is then shown that the visual system itself faces
this problem and solves it by categorizing. Models of categorization have
generally ignored perception, assuming that, prior to categorization, the
visual system has acted to extract the features needed for categorization
while suppressing information about irrelevant stimulus differences.
However, this is exactly the problem that, as this paper argues, entails
categorization. Neurobiological evidence suggests that bottom-up visual
processing mechanisms, from the earliest levels, sort stimuli into groups
according to features known to be important for perceptual
categorization. Complex feature maps, with many category-like properties,
can be observed in higher visual areas, such as the inferotemporal cortex.
At the same time, top-down factors that determine which features are
important and used for categorization have been shown to influence visual
physiology as well.
----------
Perceptual categorization is thus an inherent part of visual processing, and
it is difficult to even talk about
object recognition without reference to categorization.
Grant
_________________________________________________________________
Help STOP SPAM with the new MSN 8 and get 2 months FREE*
http://join.msn.com/?page=features/junkmail
===============================================================
This was distributed via the memetics list associated with the
Journal of Memetics - Evolutionary Models of Information Transmission
For information about the journal and the list (e.g. unsubscribing)
see: http://www.cpm.mmu.ac.uk/jom-emit
This archive was generated by hypermail 2.1.5 : Thu 06 Mar 2003 - 16:46:30 GMT