Topic: abstraction by common attributes

topics > Group: philosophy

inheritance of properties
names as abbreviations for descriptions
objects as a set of attributes
problem of classifying information
set definition by extension or intension


Aristotle formed concepts from common attributes. This simplifies a description without losing essential details. The meaning then determines reference. If this were so, then a formalist approach could be successful. Two problems with this is that few concepts have a set of defining properties, and that attributes tend to miss the rich interconnections between our concepts. (cbb 4/93)
Subtopic: attributes as relevant, recurring details up

Quote: thought processes abstract information to a skeleton of relevant features [»smitDC6_1975]
Quote: abstraction selects certain details as important and ignores the rest [»borgA1_1985]
Quote: an abstraction should conceal undesirable properties without hiding desirable ones
Quote: a whole is one concept made of subparts with something in common; can be enumerated [»leibGW_1666]
Quote: each attribute corresponds to an entity, F-ness, which characterizes an individual; e.g., wisdom and "Socrates is wise"; realist approach [»louxMJ_1970]
Quote: 'something is B' is nothing other than B-ness itself; abstractions are attributes [»leibGW_1686]

Subtopic: universal as shared attributes up

Quote: things are one when they share the same definition; e.g., despite size differences [»aris_322a]
Quote: abstraction avoids stating something more than once; recurring patterns are factored out [»maclBJ_1987]
Quote: a general thing abstracts from an actual distribution; so the meaning of general things determines reference [»stawPF_1954]
Quote: a universal is anything that is shared by many particulars [»russB_1912, OK]
Quote: properties or universals are things that particulars have and can share with others; predicates similarly shared [»martAP_1990]

Subtopic: prototype as shared attribute up

Quote: prototypical examples share attributes with other members of a category; strong correlation [»roscE10_1975]
Quote: prototype as family resemblance of attributes within category and least overlap with other categories [»roscE10_1975]
Quote: real-word attributes are interrelated; things are intrinsically separate [»roscE7_1976]

Subtopic: interconnections and universals better than attributes and instances up

Quote: a poor way to form concepts is through a list of characteristics; instead the elements of a definition are richly interconnected [»fregG_1884]
Quote: consider universal concepts instead of instances of concepts; e.g., 'gold' as a concept is richer than 'metal', while metal covers more instances [»leibGW4_1679]
Quote: the principles of eternal things are necessarily true; they explain the being of other things; e.g., fire is why things are hot [»aris_322a]

Subtopic: Plato's forms up

Quote: Plato thought it impossible to find a common definition for sensible things; they change too much
Quote: Plato's forms explain visible things by inventing an equal number of other things [»aris_322a]

Related Topics up

Topic: classification (65 items)
Topic: elements (22 items)
Topic: inheritance of properties (24 items)
Topic: names as abbreviations for descriptions (35 items)
Topic: objects as a set of attributes (39 items)
Topic: problem of classifying information (42 items)
Topic: set definition by extension or intension (18 items)
Topic: taxonomy
(16 items)

Updated barberCB 3/04
Copyright © 2002-2008 by C. Bradford Barber. All rights reserved.
Thesa is a trademark of C. Bradford Barber.