Information Technology Reference
In-Depth Information
America before writing
El Otoño del Patriarca
; and so on. Within the framework
provided by MEXICA-impro the Full Expansion of Knowledge might be achieved
by running several times the system and constantly interchanging the role of Leader
and Follower between the two agents. This solution needs to be tested. But the impor-
tant point here is to illustrate the type of research that a system like MEXICA-impro
allows.
13.5 Conclusion
The Engagement-Reflection computer-model of writing provides detailed descrip-
tions of every process involved in the model along with meticulous accounts of the
relationships that such processes establish with each other. This in depth description
of even the smallest mechanisms implicated in the system is a typical characteristic
of this type of models. The requirement of expressing in such specific terms all the
concepts behind an idea forces the researcher to ponder aspects that otherwise he
never would have considered. The development of the ER-Computer-Model based
on Sharples' general account of writing illustrates this situation.
MEXICA has proved to be a flexible tool. Because the Dictionary of Story-Actions
and the Previous Stories are defined in text files by the user, it is simple to modify the
agent's knowledge base. In the same way, it is possible to observe every activity within
the system. The program also generates all kind of reports that permit analyzing the
data produced. Thus, it is possible study the so called Unforeseen Dynamics of the
System. All these characteristics provide lots of freedom during the testing of the
system and allow a deep understanding of the model.
However, all this flexibility comes with a price to be paid. Computer models
focus on specific aspects of the phenomenon that they represent. In that sense they
are limited. For instance, MEXICA only generates plots rather than stories; it cannot
explicitly manipulate objects within a story; the system cannot include suspense,
flashbacks, parallel events; it does not take into consideration the effects of the
environment during the creative process; and so on. The list of potential components
that might influence the development and evaluation of a narrative and that MEXICA
does not consider is big. Although this might be seen as a drawback I believe it
constitutes one of the main advantages of computer models: it allows reducing to
practicable dimensions what otherwise would be unmanageable.
It is necessary to recognize that we are dealing with a phenomenon that we hardly
understand. So, it makes sense to decompose this massive complex problem into
more accessible pieces that humans can deal with, like computer models. Even so,
the task is still enormously difficult. That is what makes it interesting.
So, rather than focusing on what is missing we should concentrate on critically
reflecting on the concepts introduced by the model, establish its scope—have a clear
understanding of its strong points and limitations—and, based on that information,
suggest novel directions to progress the research. That is the way a computer model
of creativity must be understood and used.
Search WWH ::
Custom Search