Knowability as continuity: a topological account of informational dependence

Alexandru Baltag,Johan van Benthem

2024-01-13

Abstract:We study knowable informational dependence between empirical questions, modeled as continuous functional dependence between variables in a topological setting. We also investigate epistemic independence in topological terms and show that it is compatible with functional (but non-continuous) dependence. We then proceed to study a stronger notion of knowability based on uniformly continuous dependence. On the technical logical side, we determine the complete logics of languages that combine general functional dependence, continuous dependence, and uniformly continuous dependence.

Logic,Logic in Computer Science

What problem does this paper attempt to address?

### The Problem the Paper Attempts to Solve

The paper attempts to address the issue of how to model informational dependence in empirical investigations. Specifically, it focuses on how to effectively describe the dependencies and logic between variables when only limited precision measurement results are available.

### Background and Motivation

1. **Concept of Dependence**:
- In both the physical and social worlds, there are various forms of dependencies between objects or phenomena, ranging from complete functional determination to weaker correlations.
- Dependence can be seen as a relationship of information transmission. If variable \( y \) depends on a set of variables \( X \), then given the values of \( X \), we can know the value of \( y \).

2. **Challenges in Empirical Inquiry**:
- In empirical science, the values of variables are usually determined through imprecise observations or measurements, rather than ideal precise values.
- Therefore, traditional dependence logic becomes unrealistic in this context, requiring a new approach to model such dependencies.

### Main Contributions

1. **Topological Model**:
- The paper introduces a topological model to handle the uncertainty of empirical variables and observational results.
- Open sets in a topological space represent possible outcomes of experiments or measurements, and propositions are represented by arbitrary subsets.

2. **Continuity and Information Transmission**:
- Continuity is defined as a dependency relationship of information transmission. If an approximation of the values of \( X \) can arbitrarily closely approximate the value of \( y \), then this dependency is continuous.
- The paper defines global and local continuous dependencies and studies their logical properties.

3. **Uniform Continuity and Stronger Knowability**:
- Uniform continuity is a stronger dependency relationship, requiring the value of \( y \) to be consistently approximated at all points.
- The paper proposes a logical system to describe uniform continuous dependencies and proves its completeness and decidability.

4. **Independence and Functional Dependence**:
- The paper explores the complete independence and topological independence between sets of variables \( X \) and \( Y \).
- Topological independence is compatible with global functional dependence, meaning that even in completely correlated cases, approximate observations may not provide any new information.

### Conclusion

By introducing topological models and the concept of continuity, the paper provides a new approach to handling dependencies in empirical investigations. This approach is applicable not only to the fields of physics and mathematics but also to social sciences and computer science, offering new perspectives for understanding information transmission and knowledge acquisition.