ct> Chunking and Perceptual Unitization.
ct> Fisher's model of visual search illustrates that a target shape is
ct> harder to identify when it is placed among other shapes with which it
ct> shares similar features. However training can significantly
ct> increase the speed taken to identify the target shape.
This is a well known finding and seems reasonable to accept on the
basis of Czerwinski et al's notion of perceptual unitization.(ie
perceptual features are formed from a set of elementary components ).
ct> This supports Schyns et al's feature creation theory as it suggests
ct> similar features will be chunked together forming a catagory for
ct> items that share these common features which can then be
ct> distinguished from distractor items.
Yes. Schyns et.al believe that chunking can not be understood
completely unless category contrasts and similarities are taken into
consideration. There are both similarities and differences between
Schyns et al's feature creation theory and the utilization and chunking
ct> in some situations we
ct> may recognise something as a whole, and therefore we need not
ct> discretize everything we see.
Yes, large features are sometimes registered without having to be
broken down into smaller features, and smaller features can be created
by breaking down larger features.
ct> Constructive Induction.
ct> New names may be given to catagories by identifying common features
ct> of members of the catagory.
That is the problem - how do we identify common features of a
category? Constructive induction is a form of machine learning and it
may be useful to look at machine learning in order to describe and
understand how new featural descriptions are created in humans.
ct> Or, as Wisniewski and Medin illustrated,
ct> individuals may alter their verbal description of an object in order
ct> for it to fit a catagory label, based on links between abstract
ct> background knowledge and concrete object information.
Yes. Schyns argues that the immediate appearance of objects can be
altered by experience.
ct> Some argue there is an innate conceptual core that
ct> could bias the features infants will notice in objects. The fact
ct> that newborn infants will be more responsive to a picture consisting
ct> of organised features of the human face than a picture of random
ct> lines, supports this idea by suggesting that the infant has some
ct> innate tendancy to notice the features of the human face.
This suggests that human infants have a predisposition to attend to
particular stimuli (features)due to the fact that they are social
beings which need to interact with humans for emotional support, food
etc. For example, Spelke (1994) suggests that children have an innate
bias to group parts that move together into single objects. Children
also seem to have a bias towards local salient properties. However we
also have the ability to detect certain features spontaneously as a
predisposition could not possibly prepare us for all the features we
may come across in our lifetime.
Further research is needed to explain how children and adults use
categorization to discover new object features. Schyns argues that we
create object features from raw, unprocessed representations.
This archive was generated by hypermail 2b30 : Tue Feb 13 2001 - 16:23:20 GMT