An interpretation is a function
that associates the
constants of KIF with the elements of a conceptualization. In order
to be an interpretation, a function must satisfy the following two
properties.
First, the function must map constants into concepts of the
appropriate type. It must map object constants into objects in the
universe of discourse. It must map function constants into functions
on the universe of discourse. It must map relation constants into
relations on the universe of discourse. Notice that we allow for
functions and relations of variable, finite arity. The function must
map logical constants into one of the boolean values
or
(which may or may not be members of the universe of discourse).
is an object constant, then
.
is a function constant, then
.
is a relation constant, then
.
is a logical constant, then
.
must ``satisfy'' the conditions and axioms given in this
chapter and the remaining chapters of this document. As a start, this
includes the following conditions.
Every interpretation must map every numerical constant
into
the corresponding number
(assuming base 10).

Every interpretation must map the object constant bottom into

.

Every interpretation must map the logical constant true into

and the logical constant false into

.


Note that, even with these restrictions, KIF is only a ``partially interpreted'' language. Although the interpretations of some constants (the basic constants) are constrained in the definition of the language, the meanings of other constants (the non-basic constants) are left open (i.e. left to the imaginations of the language users).