Our concern is with the development of tools useful for the construction of semantically intelligent web based systems. We indicate that fuzzy set based reasoning systems such as approximate reasoning provide a fertile framework for the construction of these types of tools. Central to this framework is the representation of knowledge as the association of a constraint with a variable. Here we study one important type of variable, veristic, These are variables that can assume multiple values. Different types of statements providing information about veristic variables are described. A methodology is presented for representing and manipulating veristic information. We consider the role of these veristic variables in databases and describe methods for representing and evaluating queries involving veristic variables.