Lakhasly

The debate over the definition of computer science is as old
as our discipline itself. This paper tries to approach this subject from a point of view that has not been widely explored.
Namely, what do our students think the definition of computer science is? The paper starts with an examination of the
very nature of definitions.
2. Definitions
The first thing to remember when developing a definition is
that definitions are a human invention. A definition is an
arbitrary process that serves to include some cases and to
exclude other cases [4]. In other words, a definition must not
simply tell us what something is, it must also tell us what it
is not. This dual function of a definition is difficult to
achieve and there is often a struggle between a narrow definition and a broad definition. A narrow definition will be
specific and make clear whether or not a case will be included in the concept. The problem with a narrow definition is
that some legitimate cases of the concept may be excluded
by the definition. On the other hand, a broad defirfition will
include all relevant cases but might also include some concepts that are marginally legitimate at best.
An example of a narrow definition of computer science
is one given by Edsger Dijkstra. In his definition, Dijkstra
prefers to limit our discipline to an intellectual exercise. In
1987, he wrote "...the core challenge for computing science
is hence a conceptual one: what (abstract) mechanisms we
can conceive without getting lost in complexities of our own
making." [3] The problem with Dijkstra's definition is that
it ignores implementation issues. These issues are very
important to many computer scientists.
A good example of a broad definition is the one given
by Newell, Perils, and Simon in 1967. Their definition simply states that "computer science is the study of computers."
[8] While simple and elegant, this definition would include
many eases that most people would not consider being cornputer science. The Computing Sciences Accreditation Board
(CSAB) has tried to improve on this definition. CSAB
defines computer science as "a discipline that involves the
understanding and design of computers and computational
processes." [1] However, this definition is also very broad
and would allow many extraneous cases to fall within the
umbrella of computer science.
Finally, an example that can be considered both narrow
and broad is given by Long, et al. In their 1987 paper, they
wrote that the central intellectual role of computer science is
"the study and application of languages and methods for
making precise and understandable descriptions of things."
[7] It is narrow in the sense that it leaves out important computer science topics such as ethics in computing. It is broad
in that it allows almost any topic from science to be included. What is science if not an attempt to use methods to make
precise and understandable descriptions?
In addition to the function of a definition, there are
many types of definitions. Some of the more common types
are symbolization, similarity, spatial relations, temporal relations, and causation. Symbolization is when an object is
defined by physically pointing to an example of the concept.
If you are trying to define "apple" and you point to an apple,
you are symbolically defining the term "apple." Similarity is
when a concept is defined by examining similarities among
objects. If you are trying to define "red" you might point to
a red shirt, a red car, and a red house. When you use relations (spatial, temporal, or causation) you describe a concept
by its relation to other concepts.
In an interim report from the ACM/IEEE-CS
Curriculum 2001 task force, computer science is defined as
"an integrated field of study that draws its foundations from
mathematics, science, and engineering." [9] This is an
example of a relational definition. It assumes that one can
define mathematics, science and engineering and it implies
that whenever these three disciplines intersect you have
computer science. Another problem with this relational def