Information Technology Reference
In-Depth Information
According to Locke, a universal is formed through the cognitive process of abstraction. In this pro-
cess we view particular instances and abstract from those particulars certain properties that they hold
in common. Thus, a universal is one collection of common properties held by a set of particulars. At
first glance this sounds like Aristotle's view of universals. And, yet, Locke rejected the Aristotelian view
that the classes exist in the world. We do not, according to Locke, find objects and their features neatly
divided by nature into objectively delimited classes. Instead, limitless similarities and differences are
there for us to perceive but how we perceive them, how we select them, how we use that information to
form classes, and how we place objects in different classes is up to us to decide. Locke disagreed with
Aristotle by saying that we select the features we use to frame a universal rather than simply taking what
is there. Yet, he still believed that universals were explicitly defined. That is, the abstraction process,
according to Locke, was rational and could be explained objectively. David Hume went a step beyond
Locke and said that the construction of universals was not only done in the mind of the observer, but,
according to Hume, this abstraction process occurs below the level of rationality.
According to Hume, universals are constructed by noticing similarities and differences between
objects, just as Locke had asserted. However, this process is not guided by abstraction objectives as
Locke had claimed. Instead, according to Hume, it is a result of organizing our experience at a precon-
scious level. Any explicit definition of what constitutes a given universal is done after the universal is
defined in our minds and incorporated into our language. If the meaning of universals is constructed
at a preconscious level, then an obvious question is raised - is it always possible to explicitly define the
characteristics of universals that exist both in the minds of people and in common everyday language
? Wittgenstein didn't think so !
Wittgenstein took an extreme position with regard to universals. To him, universals not only exist
solely as concepts in the mind, but they exist in the mind as poorly formed concepts. According to
Wittgenstein, explicit definition of certain classes is not possible. In fact, many classes that we use
are so poorly formed, that the only thing that the particulars in them have in common is that they are
all members of the same class. He illustrates the point by comparing class membership with family
resemblances. Members of a family may look like each other, yet it may be difficult to define a set of
features that they all share. Some have the same nose. Some the same chin. Some the same eyes. But
there is no set of features common to them all. Classes that are held together by family resemblances
defy any attempts to construct well defined classes categories or universals.
The Pro BLeM of un IVer SAL And Infor MATIon Mode LIng
The philosophical positions we just discussed represent a range of assumptions that information modeler
may hold. It is unlikely that these assumptions are explicit. Yet they can be seen in various approaches
to modeling. For example, naïve information modelers often adopt the Platonic approach. They study
the domain until the entity classes emerge from some intuitive process as though they had apprehended
the World of Forms with their intellects. Platonic modelers cannot tell you how they derived the classes
although they seem fairly certain, intellectually, that the classes are correct. They are likely to be naïve
realists believing that the classes exist in the domain and that they have discovered them. However,
the existence of the World of Forms was even difficult for Plato to justify and the Platonic information
modeler is on similarly shaky ground.
Search WWH ::
Custom Search