This article is part of a series of essays I’ve written for an Introduction to the Philosophy of Science course at KTH on the fall of 2017.
In order to answer the question of whether computer science can be viewed as a science or not, one should rather begin by defining what science is. We can safely settle for the simple definition of science being a set of tools and methods humans have devised to gain a better knowledge of the world. These tools and methods often lead us to develop theories, which are truths that may be validated or refuted by ways of objective experiment[1].
Since computer science has long been viewed as a strictly applied discipline, it has been rather hard for the past decades to categorize it as a science. Compared to physics or biology, it often lacked a scientific process during its discoveries, and its final objective was never discovering hidden truths about natural objects[2]. While computer science made use of mathematical theories from the likes of Turing or Djikstra to find better ways to solve its challenges, it didn’t involve the typical scientific method that required experimentation in order to validate sets of theories.
After all, if computer science allowed us to do stuff (with notable success), why would we need to deal with more abstractions? The answer is surprisingly simple: Without a set of theories and models, how could computer science could ever be limited ? Could we call spreadsheet calculations or the study of coding styles a part of computer science ?
Fortunately, great progress has been made so far in order to limit computer science to a more precise definition.
So rather than seeing computer science as a simply computer-related discipline, or even worse: the study of computers, it may be more relevant to classify it as a science related to computing.
From beautiful dunes created by simple information processes local to each grain of sand [3], to complex DNA formed by encoding mechanisms [4], computing can be found everywhere.
So instead of answering whether computer science is a science or not, the question has evolved into whether computing can be viewed as a science. As computing is a discipline that studies natural objects, the answer can be easier to answer. Whether it’s automation theory, encoding processes or storing information, the discipline becomes limited to studying information processes and is therefore easy to classify as science.
One would think that answering our main question can be considered enough, but others may argue that we should not merely strive to know if computer science constitutes science or not, but whether we should try to make it one ?
The challenge is clearly not as easy as answering a simple question, but it’s possible that it can be riped with benefits to the community. What can be so bad about experimenting more instead of holding subjective experiences as holy truths[5]? Or demonstrating once and for all certain falsehoods and never looking back ?
Surely progress will be slowed down by applying these more rigorous approaches, but by defining our discipline as science, shouldn’t we start rising up to its standards ? If computer science has allowed for great technological gems to be unraveled, one could only dream about what they can become when polished with the sharp scrutiny of science.
[1] https://www.sciencebuddies.org/science-fair-projects/science-fair/steps-of-the-scientific-method
[2] Quote from “Richard Feynman, Feynman Lectures on Computation, 1970”
[3] Abstract from https://link.springer.com/chapter/10.1007/3-540-36494-3_56
[4] https://www.sciencedaily.com/releases/2012/09/120912093204.htm
[5] Shoulds computer scientists experiments more ? Walter F.Tichy