Why Computer Science Is a Science.

3 arguments for why computer science is truly a science.

Date: June 7, 2026

If you have taken or have taken computer science, you have probably asked yourself whether computer science (CS) is a science. What is science anyway?

Ancient and medieval philosophy defined science as the knowledge of causes. Thus, a discipline that uses reason to understand causality is a science. The etymology of the word science is revealing; the word science comes from the Latin verb scio, which means I know. Evidently, science has to do with knowledge. Knowledge is the possession of truth, where truth is, borrowing from Aristotle, the agreement of what is in the mind with what is reality. Thus, science is the study of causes in reality (nature).

Today, science is largely considered to be what the positive (hard) sciences study. Positive or hard meaning is what can be apprehended directly by the senses. For physics, it is physical matter, and for biology, it is living things. Positive science studies aspects of nature that can be measured: material substances. The so-called scientific method is effective for the study of material substances because material substances lend themselves to experimentation. A scientist forms a hypothesis, which is a falsifiable claim, designs an experiment to test it, and uses mathematical precision to assess the outcome of the experiment.

To show that CS is a science, I have to argue that CS studies some aspect of nature and that it uses the scientific method.

1. Material object argument: CS studies an aspect of nature

I argue that CS studies an aspect of nature. Unfortunately, the name computer science is quite misleading. A quote by Edsger W. Dikistra expresses this accurately: “Computer Science is no more about computers than astronomy is about telescopes.” CS is the study of what is computable, of algorithms. I argue that algorithms are an aspect of nature. Dynamic beings rely on an information-driven process for development. The most formidable example of this is the unfolding of DNA. In sum, nature itself uses algorithms. CS asks questions like, What can we achieve with mechanical step-by-step procedures (algorithms)? What are the limits to designing algorithms? Are there problems we cannot solve using algorithms?

Algorithms do not execute immaterially. All algorithms are material. A core subject of CS is the study of data structures and algorithms. This branch of CS is obsessed with formalising knowledge on how to design algorithms that economise space and time. Space and time. Space and time are part of nature. CS is most clearly seen to interface with nature when it contends with space and time. Arguably, one of the hardest unsolved problems in science is that of solving NP-hard problems.

Since CS studies nature, it often piggybacks on physics and even biology. The success of deep learning originated from the modelling of the biological neuron. Geoff Hinton, hailed as the godfather of AI, started in experimental psychology. He often borrows ideas from neural science for application in deep learning. It is therefore no surprise that some scientific theories exist where computer science ends up bridging with physics. Take, for instance, the application of random matrix theory and spin glasses in deep learning (see Physics meets ML for more). One of the biggest open problems is the search for a scientific theory of deep learning. Scientists are still puzzled about how deep neural networks generalise despite their complexity. The curiosity of physicists and mathematicians is being aroused by the new problems posed by deep learning. There seems to be a secret of nature in this phenomenon, and science is ultimately concerned with finding out such secrets. CS is one more science that attempts to converse with nature; the art of science can be thought of as the art of giving nature a language to speak to us.

2. Formal object argument: CS uses the scientific method

CS uses the scientific method. This is not difficult to argue. CS research, like other sciences, begins with a hypothesis. The scientist then designs an experiment to test the hypothesis. Mathematical and statistical precision is used to understand the experiment and hence decide the fate of the hypothesis. Once the scientist finds an interesting observation, they publish their findings, whereupon other computer scientists reproduce the experiment. And thus CS progresses. Arguably, AI is the fastest-growing science today.