Information Technology Reference
In-Depth Information
the architecture of neural networks. Obviously, the computational representation
of cognitive processes really boils down to developing computer models of these
complex biological and psychological processes. However, due to the complexity
of the object studied, this modelling approach seems particularly convenient, as
models can not only be created, but also researched experimentally using com-
puter simulations. The adequacy of these models is confronted with the neurobio-
logical knowledge of the structures and functions of nervous system parts and
elements, constantly collected and extended thanks to new research techniques of
brain biology. This approach has meant that a 'cognitive science revolution' begun
in the development of science, resulting in the creation of a pure science of cogni-
tion - cognitive science.
One of the reasons for the development of cognitive science was the recogni-
tion of the fact that the problem of cognition turned out to be much more complex
than originally thought, and getting to grips with its intricacies implied a new split
of research work between those collecting empirical facts and those who inter-
preted them. What started distinguishing cognitive science from other sciences
(philosophy or psychology), was its multidisciplinary nature. The term 'multidis-
ciplinary' was introduced to characterise fields of science (such as cognitive sci-
ence) which are founded in many other fields and which significantly contribute to
the development of these fields. This term was introduced to mark a difference
from 'interdisciplinary', which appears when some well-defined research or engi-
neering problem requires the cooperation of representatives of various scientific
disciplines, each of which contributes the tools typical for his/her discipline, but
does not receive much in return 2 . The situation was different in researching cogni-
tion and understanding processes, where detailed disciplines contributing to the
above multi-disciplinary approach receive contributions to their own development
and growth from the common resource to a greater extent than they contribute
ready solutions to that resource.
Looking back, it has been observed that today's science can distinguish and re-
search only small fragments of the complex set of problems associated with cogni-
tive science. Such isolated cognitive science problems as perception, imagination,
memory, learning, conceptual thinking, understanding and many others have been
researched - but each one separately. Every one of the above forms of cognitive
activity of the mind is now the field of interest of a separate research discipline,
and in addition, each one can be analysed at many levels. The most widely used
structure of these levels has been described by David Marr [69], who distin-
guished three primary levels:
the computational (theoretical) level,
the level of representation and the algorithm (the interface between the theory
and the empirical),
the hardware (mainly computer) implementation level.
2 Such work was done to, for example, construct the nuclear bomb, space rockets or analyse
the human genome [91], [92].
Search WWH ::




Custom Search