Concept clarification

From Wikiversity
Jump to navigation Jump to search

The purpose of this page is to collect excellent further reading on the subject of concept clarification (also known as conceptual clarification), to help learners find some of the best material available. It may be expanded later to serve a broader purpose. Best sources are those that are freely available for reading, and that are well done or authoritative.

Concept clarification involves "An analysis of an abstract concept through investigation of examples and the identification of critical and less critical attributes", as per webquest.org.

Concept clarification involves asking the following questions:

  • What is the primary superclass, the "genus"?
  • What are other superclasses? This seems less critical.
  • What are the differentiating characteristic?
  • What are example subclasses?
  • What are example instances?
  • What are counter-example subclasses?
  • What are counter-example instances?
  • What are disputed subclasses?
  • What are disputed instances?
  • What concepts is the concept to be distinguished from?
  • What parts do the individuals belonging to the concept have?
  • What large whole do the individuals belonging to the concept belong to?

The above may seem surprising, from the point of view of genus-differentia definitions usually found in dictionaries, which are in the Aristotelian tradigion. But as counter-intuitive as it may seem, all the above questions can help advance concept clarification.

Important part of concept clarification is watching for ambiguity. Thus, it does not do to say that "cat is an animal" since both "cat" and "animal" are ambiguous. Thus, one could say:

  • cat-domestic-animal is an animal-living-thing-including-horses

Otherwise, "animal" also has a meaning that has humans as the genus, but that is not intended. Furthermore, even the above is ambiguous between an instance of and a class of. Thus, one could say:

  • cat-domestic-animal is-subclass-of animal-living-thing-including-horses

That would be in contrast to this:

  • William-my-dog is-instance-of animal-living-thing-including-horses

The process of finding disambiguating identifiers is recursive and networked, but given sufficient combination of words to name the concept, using the hyphenated convention, can provide good reduction of ambiguity. Whether the ambiguity was truly ultimately reduced by the above method is probably an open philosophical question.

What a dictionary could do, to eliminate ambiguity for readers, is link each ambiguous word in a definition not to a target word but to a target lexical unit, defined as a pairing of word with a sense. Thus, ("cat", cat-domestic-animal) is a lexical unit; "cat" is a word and cat-domestic-animal is the concept.

What a dictionary could also do is place disambiguating words in brackets. Thus:

  • cat: domestic animal (including horses and bacteria) that is X

That would be very clumsy. What a dictionary could do instead is make sure each word is marked for primary sense. OED does not do that since it is a historical dictionary, listing senses in the order in which they originated. Many other dictionaries list senses by frequency, and thus, the first listed sense can be treated as primary sense. Thus, the definition would ideally use each word in its primary sense. However, unless a sense is expressly designated as primary, this method provides no stability: sense frequency changes, and once a word would be moved up in the entry because of frequency, it would break all the definitions that depend on that word.

Wikidata provides astounding tools for concept clarification. Since, each concept is a Wikidata item and the defining statements are created as structured statements. Thus, Wikidata defines concepts using concepts, rather than defining words using words. However, it is far from easy to use for mere mortals, and as a result, most entries for general concepts lack the defining statements. The name Wikidata does not reveal that this capability is present. This is in part since Wikidata, being a knowledge graph, serves two rather distinct purposes: concept definition and data entry about individuals. The latter is what one would typically find in a relational database, say in a table about individual companies.

See also[edit | edit source]

Further reading[edit | edit source]