Begin at the beginning...and go on till you come to the end: then stop.
-Lewis Carroll
The time was drawing near, Walter Bigelow thought. Just a few more adjustments, and his great ambition would be fulfilled.He stepped back from the Time Distorter and studied the complex network of wires and tubes with an expert's practiced eye. Twenty years, he thought. Twenty years of working and scrimping, of pouring money into the machine that stood before him on the workbench. Twenty years, to save Abraham Lincoln's life.
These two paragraphs are the beginning of Robert Silverberg's short story, The Assassin (story:assassin). A reader is immediately thrust into a narrative world ([#!read:gerrig1!#]), a specific creation of a human being with a specific purpose. The reader enters this world and attempts to comprehend it as they would the world around them. Ultimately, Silverberg is telling a story; to comprehend it, a reader must understand the concepts being presented. The author expects that this will happen and presents the narrative world in a piece-meal fashion, giving the reader crucial information bit by bit. This has to occur; while the story being told is a single entity, a text is forced to linearize this structure. The act of comprehending a story does not stop, however, with how the author decides to present the elements. The reader, meanwhile, is actively processing the text in an attempt to ``make sense'' of it. Often, predictions are made, some of which are supported, but many of which are discarded. The reader attempts to integrate the material coming from the story with what they know--in the case of this example, this includes what they know about Silverberg stories, about assassins, about ``time drawing near,'' and so forth. And, they attempt to utilize their own experiences as they seek comprehension. We have all had great ambitions and can, therefore, empathize with Mr. Bigelow.
While the first paragraph is a fairly standard introduction to the world of the story, the second paragraph forces a different sort of comprehension on the reader. A Time Distorter does not seem like a device one could acquire at the local electronics store--indeed, it does not seem like it could exist at all. The line about saving Lincoln is further trouble--the story does not seem to be set in the 1860s, yet the reader knows that this is the time of Lincoln's assassination. Of course, these troubles are resolved if the reader allows the possibility of a machine which would permit Mr. Bigelow to travel in time; to go ``back'' to the point of Lincoln's death and prevent it. At the same time, however, the reader is aware that such a time machine does not exist. In order to overcome these understanding problems, the reader must be willing to suspend their disbelief. If a time machine could exist, then the portion of the story read up to this point would ``make sense'' in a reasonable fashion; in other words, it would be coherent. If the reader refuses to suspend their disbelief at this point, the story will be more difficult to read. Since the various elements of these first two paragraphs have not become pieces of a single conceptual framework, the reader will be forced to maintain several possible interpretations as they search for the ``correct'' one (in this case, the one which the author intended).
This suspension opens its own Pandora's Box--does the reader pause now, put the story aside, and ponder the outcome of accepting the existence of a time machine? The story has barely begun; yet the reader, if they were so driven, could stop reading it and spend the next several months considering the ramifications of time travel. An alternative would be for the reader to simply accept time travel and to allow Silverberg to spin the tale in the fashion which he desires. At the moment, this is still possible; there is nothing in the story which forces one to consider precisely how the time travel will take place, only that it must for certain story elements to exist. Therefore, a reader can elect to continue reading.
In reality, most readers of the story will not have an awareness of this precise set of questions and thoughts. After all, most readers in today's world are familiar with the concept of time travel, even those who are not science fiction fans. It is a concept which has permeated much of our culture, in one form or another. Still, a majority of stories will contain concepts that are novel to the reader attempting to comprehend the text. Some of these stories are science fiction, some are general fiction, and some are actually news articles reporting on a new idea. In all such cases, however, a reader must understand the novel concept if the text is to be successfully comprehended. Creative reading is needed.
This dissertation research explores issues underlying the reading process, particularly those issues which enable developing an explanation of creative reading. This means that my work does not concentrate on a wide range of processes which can be seen in other examples of reading research--for example, my work does not make any comment on eye movement while reading, interpreting the printed word in a meaningful fashion, lexical retrieval, and so forth. My research has focused on the higher-level, conceptual aspects of reading. Specifically, the focal issue which my work addresses is the question of how a reader can comprehend a text which contains novel concepts? The major claim of my research is my attempt to answer this question: A READER NEEDS TO BE CREATIVE IN ORDER TO COMPREHEND REAL-WORLD TEXTS, WHICH OFTEN CONTAIN NOVEL CONCEPTS THAT MUST BE UNDERSTOOD. However, this claim is too non-specific to be illuminating; what needs to be done is to break the claim down into smaller theses which are testable. Some of these claims will be shown by example and by argument; others will have their validity demonstrated through the use of a computer model, ISAAC (Integrated Story Analysis And Comprehension), which provides a testable instantiation of the theory. The entire hierarchical organization of the core claims of my research is outlined in in Figure 1; refer to that diagram as I present the claims being put forth in this dissertation.
In brief overview, there are two major divisions to the claims--one set relating to the reading task breakdown and the issues which it raises and the other set relating to the specific details of the creative understanding process and its specific issues. The reading task claims include:
In addition to the reading and understanding claims, there is a third set of claims which my work supports--this set involves ways in which a model and theory of reading can be evaluated. Evaluation is a difficult task in any artificial intelligence undertaking; for a somewhat ill-defined intelligent behavior, such as reading, it is extremely difficult to perform principled evaluation. I have developed a methodology which enables me to evaluate my model in an empirical fashion. However, since the claims I make about evaluation are not directly related to the issues which must be dealt with in order to develop a theory of creative reading, I have not included this set of claims in my hierarchical diagram.
