Complexity Reduction - Principles, Methods and Challenges
In the ninth sub-programme of the WIN-Kolleg, interdisciplinary research will be conducted into whether, how and why the reduction of complex facts and situations is necessary for scientific knowledge and results and what consequences this reduction of complexity has for the relationship between science and society. Interested applicants will find more detailed information and exemplary research areas for the sub-programme in the detailed programme description. Young scientists are invited to submit proposals that go beyond these examples and fit into the theme "Principles, Methods and Challenges of Scientific Complexity Reduction".
Programme description
With the topics "How do collectives decide?" and "Stability and instability of states", the seventh and eighth seasons of the WIN-Kolleg explored how, across many fields, generally valid criteria for the decision-making of collectives and the stability of complex systems can be found. Following on from this, the ninth season will now explore the question of whether, how and why the reduction of complex facts and situations is necessary for scientific knowledge and results and what consequences this reduction of complexity has for the relationship between science and society.
Our daily experience as well as our scientific actions are determined by a constant reduction of complexity. When using a technical device, when creating a scientific model, when translating between languages, when analysing historical developments or when grasping complex systems and facts, ambiguities, uncertainties and contradictions are methodically excluded to a large extent by various principles of complexity reduction in order to be able to arrive at valid results. However, this raises a number of fundamental questions:
How are "universally" valid statements about formations and processes possible in a world that is seemingly so irreducibly marked by differences? Is complexity only given by the fact that epistemological and empirical means are not sufficient to grasp it? What role do (disciplinary) languages play in this? Is complexity a continuum that can only be reduced by interrupting the continuum at one point? Is complexity therefore unavoidable and are scientific means sometimes insufficient to capture it? What are the consequences? Can research only be done in teams? Do we need artificial intelligence to solve the problems? Are findings the end of theories?
The digitalisation of many aspects of life as well as multiscale modelling in the natural and engineering sciences should suffice as examples of this complex problem. Every day we experience an enormous increase in the complexity of the electronic devices and digital processes that surround us. Complex user interfaces can lead to a digital divide in different populations and artificial intelligence enters with the claim to learn complex models, but often with the loss of transparency and comprehensibility. In theoretical computer science, complexity theory is an established subfield for dividing the resource requirements of algorithms into "equally difficult" classes. In practical computer science, data complexity is an important aspect of Big Data and one goal of software engineering is to master the complexity of large software systems. These informatics complexities often pose insurmountable limits to the solution of scientific and engineering problems. The complexity of the underlying models does not allow for efficient simulation and the complexity of many sensor data exceeds the available analytical capabilities. In the description of complex properties of material science or physical problems, where a reduction of complexity is generally not possible, universal behaviour occurs in the limiting case of large times or dimensions, which can be exemplary for other scientific fields.
The example of the digital field shows that complexity reductions are necessary in science, but that their manifestations are also controversial and can certainly entail dangers. The problems are similar in many other scientific fields. The proposed framework topic raises a wide range of questions about the understanding, classification and control of complexity. Possible research fields within the call could be:
- Digitisation and AI
- Relationship between history and natural sciences
- Environmental issues, climate change, sustainability, natural heritage
- Remembering and forgetting
- Conflict research
- Transmission of historical knowledge
- Edition and translation of texts
- Etc.
Applicants are also invited to submit proposals for other research projects that fit into the theme of "Principles, Methods and Challenges of Scientific Complexity Reduction".
The question of how far a reduction of complexity can go in order to avoid falling into populism or fundamentalism must always be taken into account.