Discussion

The reading framework provided here is intended to be just that--a framework. More precisely, it acts as a workbench on which I can implement my theory and model of   creative understanding (Chapter 6). I believe that it is imperative that one implement the complete reading theory, utilizing published stories, in order to do research on any piece of it. As the next chapter shows, there are aspects of the reading process which directly act to constrain the creative aspects of the understanding process. In order to accommodate this idea, I have to present a reading theory sufficient to allow the creative understanding process to occur. This does not mean that I do not make specific claims about the reading process. My ``highest'' level claim about the reading process is the point that   THE BEHAVIOR KNOWN AS READING IS MADE UP OF A COLLECTION OF SMALL TASKS, as demonstrated in this chapter by describing the tasks which make up the process.

Furthermore,   THE REQUIRED TASKS WORK LARGELY INDEPENDENTLY BUT ARE ORGANIZED WITH A HIGHER-LEVEL STRUCTURE. This claim is supported by the large amount of literature concerning the reading process and by the abilities of the model which implements this approach (as described in Chapter 8). The higher-level structure are the supertasks which I have described in this chapter, and I conclude that   ORGANIZING THE TASK SET INTO SUPERTASKS ALLOWS TASKS WITH SIMILAR PURPOSES TO COORDINATE THEIR EFFORTS. As the reading elements are not my primary focus, this claim is another one which is largely supported by prior evidence from the literature.

Since the supertasks are already working largely autonomously, one possible implementation would be for complete autonomy. However, this results in supertasks doing redundant work--as I pointed out earlier, supertasks discover information which can aid other supertasks in their processing. So,   COMMUNICATION IS REQUIRED BETWEEN THE SUPERTASKS. Without this communication, supertasks become too autonomous. Thus,   INFORMATION DISCOVERED BY ONE SUPERTASK NEEDS TO BE AVAILABLE TO OTHER SUPERTASKS. This form of communication is implicit and is achieved by shared memory in the ISAAC model. There is also a system claim that   EXPLICIT COMMUNICATION REQUESTS PASS BETWEEN THE SUPERTASKS IN ORDER TO FACILITATE COORDINATION OF PROCESSING. In ISAAC,         these requests are in the form of the expectations, anticipations, suggestions, and requests which are passed among the supertasks. Notice that both of these last claims are tied to the model I have built. The higher level claim which encompasses them both, that communication is required, is a theoretical one. Different instantiations of the theory could implement this idea in different fashions.