Is a Computer Science considered a real science?

[Niaboc67]

I am majoring in Computer Science and have wondered, is computer science a science? as in the physical and natural sciences or is more appropriate to call it engineering? If not is it proper to call someone that has a career in Computer Science a Scientist? I am dealing with the terms and semantics of what the major as a whole exactly is.

Thank You

 

[Greg Bernhardt]

A discipline that applies the scientific method is a science.

http://www.wired.com/beyond_the_beyond/2007/06/computer_scienc/

 

[analogdesign]

The old joke is that anything with "Science" in the name isn't science.

As far as scientific method, there's not much empiricism in Computer Science I'm afraid. But I dunno... I feel like it's a science. Certainly Computer Scientists I have known strike me as scientists.

 

[mpresic]

I am a physicist. Some of my friends got computer science degrees 30 years ago. I always thought computer science was cool. I always regarded it as the only "man-made" science.

 

[Jozape]

Depends on what you consider to be science. Some people seem to believe mathematics does not qualify as science since experiments/observations are not required, while, at least on Wikipedia, mathematics is labelled a "formal science". Computer science is understandably labeled a formal science since "computer science" is similar to mathematics, while physics is considered a "natural science".

 

[StatGuy2000]

This question ultimately hinges on what we define as "science". While I'm not a philosopher of science, from my readings and my general experience, a field of inquiry can be labelled as a science if the following conditions hold:

(1) The field is in some way studying properties of the world (whether natural or "human-made").

(2) The field utilizes the "scientific method" as a core part of its inquiries; that is, the 4 essentials following are in various iterations or orderings:
(a) characterizations (observations, definitions and measurements of the inquiry)
(b) hypotheses (theoretical, hypothetical explanations of observations of measurements of the subject)
(c) predictions (reasoning including logical deduction from hypothesis)
(d) experiments (tests of all of the above)

By the above characterizations, computer science certainly meets all of the above criteria. I would even argue that mathematics meets the above criteria since the development of proving theorems can be considered an experiment/observation.

 

[JakeBrodskyPE]

I tend to subscribe to the definition of science proposed by Richard Feynman. In other words, one must be able to conduct a repeatable, controlled experiment that can validate one hypothesis but not others.

Physics is a science. Chemistry can be a science. Even biology can be a science. I do not see how Computer "Science" meets that definition.

 

[analogdesign]

This is starting to sound like a conversation between undergraduates at a party... "Well, what do you really *mean* by the Science". Thomas Aquinas would be proud. :)

 

[AlephZero]

I tend to subscribe to the definition of science proposed by Richard Feynman. In other words, one must be able to conduct a repeatable, controlled experiment that can validate one hypothesis but not others.

I think that leaves "computer science" in a similar position to mathematics: in principle, the outcome of any "experiment" could be deduced by logic, without actually doing the experiment. That's not to say that "doing experiments" in either subject has no value, of course.

Another way to define "computer science" would be along the lines of "the stuff related to computers that isn't computer engneering, or software engineering.".

Looking at the "Great Principles of Computing Project" referenced in Greg's link, http://cs.gmu.edu/cne/pjd/GP , reminded me of Rutherford's comment that "all science is either physics or stamp collecting" - and most of computer science certainly isn't physics.

 

[WannabeFeynman]

The authors of Structure and Interpretation of Programs didn't consider it a science.

 

[NATURE.M]

I think computer science can be considered more of an applied science, much like engineering.

 

[Vanadium 50]

Why is this career guidance? How does the definition of the field influence your career choice at all?

 

[StatGuy2000]

Why is this career guidance? How does the definition of the field influence your career choice at all?

I agree with you that this isn't really about career guidance. Perhaps this thread be moved to General Discussion, or some other section in PF?

 

[StatGuy2000]

I tend to subscribe to the definition of science proposed by Richard Feynman. In other words, one must be able to conduct a repeatable, controlled experiment that can validate one hypothesis but not others.

Physics is a science. Chemistry can be a science. Even biology can be a science. I do not see how Computer "Science" meets that definition.

Well, first of all, one can conduct repeatable, controlled experiments to validate hypotheses in specific subfields of computer science (e.g. reliability of algorithms, computer experiments, simulation experiments). So by Feynman's definition, at least these subfields can definitely be considered a science.

More generally, with all due respect to Feynman and his enormous contributions to physics (and science more generally), I personally feel that his definition of science is somewhat too limiting, because implicit in his definition is the need to run some sort of physical experiment to validate hypotheses. However, even within certain fields of biology (e.g. evolutionary biology) direct experiments are rarely ever conducted. Instead, evolutionary biologists take the wait of observations over a span of time to validate hypotheses, and therefore still meets the test of the scientific method.

The theoretical foundation on which computer science rests is on the definition of computation and the algorithm and on what phenomena can or cannot be computed or defined algorithmically. This theoretical foundation is an outgrowth of mathematics, and as I described earlier, mathematics is a science, where deduction and proof is the experimental method to validate/invalidate hypotheses.