Transcript Taylor 4

Taylor 4
Prototype Categories II
Two main issues:
• What exactly are prototypes?
• Do ALL categories have a
prototype structure?
4.1 Prototypes can be understood
in two ways:
1) As a central member, an actual artifact
2) As a schematic representation of the
conceptual core of a category
-for this model, a given entity instantiates
the prototype
-for many categories (e.g. TALLNESS),
only 2) is possible
What is Similarity?
• Membership is assigned by virtue of
similarity to the prototype
• BUT: similarity is a difficult concept
because
– 1. it is a graded concept
– 2. it is a subjective notion
– 3. similarity is based on attributes, which
themselves show prototype structure
• (cf. Gentner’s “similarity space”)
Nominal Kinds
• These are categories with essential
conditions for membership.
• The existence of such categories is not
inconsistent with a prototype approach.
• Facts of this type do not lead to all-ornothing category membership.
Natural Kinds
• These are categories with a clear
boundary.
• The existence of a clear boundary does
not preclude prototype organization.
• There can still be better & worse
examples, gradience within the category.
4.2 Prototypes according to
Langacker
• Prototype – a typical instance of a
category
• Schema – an abstract characterization
that is fully compatible with all members of
the category it defines
Either a schema or prototype + extensions
may suffice. Taylor will focus on prototype
+ extensions, because not all categories
yield schemas.
4.3 Folk Categories &
Expert Categories
• Even ODD NUMBER and EVEN NUMBER
show prototype effects – small numbers (3
vs. 2, 4) are better examples
• “Prototype effects…arise from an
interaction of core meaning with nonlinguistic factors like perception and world
knowledge”
Folk vs. Expert Categories
• Expert categories – defined by the
imposition of a set of criteria for
membership
• Folk (natural) categories – structured
around prototypical instances and
grounded in how people normally perceive
and interact with things in their
environment
Folk vs. Expert Categories, cont’d.
• Some words, like gold and water are
subject to both expert and folk definitions.
• The folk definition is prior to the expert one
and is often used even when a person
knows the expert one.
4.4 Hedges
• Our everyday folk theory of what a
category is contains the belief that
categories are definable in terms of what
their members have in common.
• Language requires us to use one word
(form) or another, to choose among
categories – this reinforces the folk belief
in discrete categories.
4.4 Hedges, cont’d.
• Every language has hedges, which
enable a speaker to express degree of
category membership.
• Some hedges: loosely speaking, strictly
speaking, par excellence, technically
• These words manipulate categories and
boundaries
4.4 Hedges, cont’d.
• Hedges provide evidence
– That we distinguish between central and peripheral
members (par excellence, strictly speaking)
– That we distinguish between different degrees of nonmembership (strictly speaking)
– That category boundaries are flexible (loosely
speaking)
– That categories can be redefined by ad hoc selection
and re-weighting of attributes (in that)
– That in some cases categories are defined by
classical principles, but these are felt to be exceptions
(technically)