Michał J. Gajda
Abstract:Philosophy of science attempts to describe all parts of the scientific process in a general way in order to facilitate the description, execution and improvements of this process.
So far, all proposed philosophies have only covered existing processes and disciplines partially and imperfectly. In particular logical approaches have always received a lot of attention due to attempts to fundamentally address issues with the definition of science as a discipline with reductionist theories.
We propose a new way to approach the problem from the perspective of computational complexity and argue why this approach may be better than previous propositions based on pure logic and mathematics.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is the limitations of existing philosophy of science in describing and explaining the scientific process. Specifically, the author believes that the existing philosophy of science (such as logical positivism, empiricism, falsificationism, etc.) has its value in some aspects, but they fail to comprehensively cover all scientific activities and are insufficient in dealing with computational complexity and predictive ability.
### Main problems
1. **Limitations of existing philosophy of science**:
- Existing philosophical methods in science (such as logical positivism, empiricism, falsificationism, etc.) can only partially cover the existing scientific processes and disciplines and cannot provide a unified framework to describe all scientific activities.
- In particular, logical methods, although attempting to define science through reductionist theories, are often too simplistic to deal with complex real - world problems.
2. **Definition of computational science as a new scientific paradigm**:
- With the development of computing power, computational science has gradually become a new form of science, which combines data science and empirical modeling of natural phenomena. Therefore, there is an urgent need to provide a reasonable philosophical definition for computational science.
- The author believes that computational science is not only a new method of program construction, but also a completely new perspective of the philosophy of science.
3. **Predictive ability and computational complexity of scientific theories**:
- The author proposes a new philosophical view - **Predictivism**, emphasizing the importance of the predictive ability and computational complexity of scientific theories.
- Predictivism believes that the core of scientific knowledge lies in providing probability predictions, not just through experimental verification or falsification. Even if some theories have been falsified, if they have low computational costs and good predictive effects, they can still be regarded as valid scientific knowledge.
### Research questions
To explore these problems, the author proposes the following research questions:
- **RQ1**: What are the characteristics of science compared with other human activities?
- **RQ2**: Can we define science so that it covers all current scientific activities?
- **RQ3**: What are the differences in the applicability of different philosophies of science in different fields and scientific activities?
### Solutions
The author proposes a new method based on computational complexity to re - examine the philosophy of science, emphasizing the following aspects:
- **Predictivism**: Scientific theories should focus on predictive ability rather than relying solely on experimental verification or falsification.
- **Computational complexity**: Consider the computational complexity of theories and choose those theories that can provide the best predictions under given computational resources.
- **Hierarchical structure of theories**: Consider scientific theories as a hierarchical structure, in which simpler theories can provide useful predictions at lower computational costs, even if they are not as accurate as more complex theories.
Through this method, the author hopes to provide a new framework for the philosophy of science, which can not only explain the diversity of existing scientific activities, but also meet the new challenges brought by computational science.