If we examine the two sample stories, Zoo and Lycanthrope, we can see that both contain numerous examples of objects, including agents. The characters in the stories are all examples of agents--these include Professor Hugo and the children from Zoo and Ed and Norm from Lycanthrope. In both examples, the agents are involved in numerous situations in which they initiate activity. The stories also contain numerous objects which do not initiate action and which are therefore not agents. The spaceship is an example; the book on witchcraft is another. These objects exist in the stories and are manipulated by the agents. In all of these examples, the objects I have been discussing could exist in the ``real'' world; some are simply more common than others. There are objects and agents in the stories which do not exist in the world; for example, commocs are not a unit of currency in our experience and horse-spider creatures have never been seen. These, then, are both examples of novel entities which the stories are presenting to the reader. To comprehend the stories, both novel and recognized entities need to be handled by the reasoner--part of how this is done will be discussed in Section 4.4; part of this will be discussed in Chapter 6 when I discuss the details of the creative understanding process.
While recognizing that an object exists is important to the reasoner, it does not lead to understanding of the object--there must be a way in which to describe the characteristics of that object; I call these characteristics attributes. All objects possess an attribute known as a function. The function is the current role an object is playing. For example, a chair's function may be to allow someone to sit on it. Naturally, objects have more than one possible function (for instance, one may use a chair as a battering ram, if needed); so, a reasoner also knows a set of possible functions for a given object. This is not the complete set of functions which a particular object may fill; this complete set could contain an extremely large number of members.
In order to understand an object's function, a reasoner needs to understand how that object can accomplish the given function. Again, consider the chair and the function of ``sittability'' it possesses. A chair fills this role by having a surface large enough to be sat on, by being stable, and so forth. These functional characteristics make up the primary attributes of an object. Thus, a reasoner has an explanation of how an object performs its current function; the characteristics of the object which allow the function to be fulfilled make up the object's primary attributes. A reasoner will also know a number of other facts about the object and its various characteristics, facts which have nothing to do with the current function. For instance, a reasoner may know that a particular chair is blue, is made of wood, is located in the kitchen, and so on. Attributes which do not contribute to the understanding of the object's function are referred to as secondary attributes The primary and secondary attribute distinction is only with respect to the current function of the object. If the function was to change, then the division between primary and secondary attributes may also need to change. Similarly, if a reasoner considers a different set of primary attributes or a different set of secondary attributes, the function of the object may need to be altered. As another example, the weight of a chair is generally unimportant to its role as a object one may sit on (with the caveat that the weight must be large enough to produce stability). If a reasoner considered a new function, such as acting as a battering ram, the weight would become more important, therefore needing to be a primary attribute.