The Power of Network Pluralism: Multi-Perspective Modeling of Heterogeneous Legal Document Networks

Titus Pünder, Corinna Coupette

Abstract

Traditional network science often seeks a single, definitive graph to represent complex systems. However, in domains like law, any single model is an act of perspective-taking that can obscure critical structural nuances. This paper introduces "Network Pluralism," a framework that treats networks as diverse perspectives within a multidimensional "network space." By applying this approach to the German legal system, we demonstrate that standard "importance" metrics are highly sensitive to modeling choices. Our multi-perspective analysis reveals that traditional citation counts often overemphasize procedural "boilerplate" over substantive law, and that shifting granularity allows for a more accurate mapping of legal relationships.

Introduction: The Problem of Perspective

In the legal world, the answer to almost any question is "it depends." Network science, however, usually models complex systems—such as the law, the brain, or social groups—as a single graph. This approach often flattens inherent hierarchies and leads to incomplete results. In legal networks, for instance, a citation might refer to an entire act, a specific section, or a single paragraph.

If researchers only look at document-level citations, they risk concluding that the most "important" laws are procedural rules (e.g., court cost regulations) simply because they are cited ubiquitously. This "procedural dominance" masks the substantive legal heart of the matter. To address this, we propose Network Pluralism: a shift from single-model analysis toward a multi-perspective approach that maps the "dimensions of variation" in a domain.

Dimensions of variation in legal network modeling, illustrating the different perspectives (Rules, Documents, Institutions, and References) that define the network space.

Methodology: Navigating the Network Space

To implement Network Pluralism, we developed a conceptual and technical workflow to navigate the "network space":

  1. The Base Network: Instead of an aggregated graph, we construct a raw, non-aggregated "base network." This is a directed bipartite graph connecting two "forests" (hierarchical trees). One forest represents the judiciary (courts → panels → decisions → paragraphs), and the other represents legislation (statutes → sections → sub-sections).
  2. Dynamic Counting Algorithm: To handle varying levels of citation granularity (e.g., citing a high-level statute vs. a specific sub-paragraph), we developed a "Dynamic Counting" algorithm. It "broadcasts" references down to the most granular level and "aggregates" them back up on demand, ensuring no information is lost regardless of the chosen resolution.
  3. Multi-Network Analysis: We derive a multiplicity of networks (60 in our case study) by varying granularity, filtering by court type, and using different projection methods, such as co-occurrence of laws within the same paragraph versus the same decision.

Our study utilizes a massive dataset of German federal legislation (6,815 statutes and 91,758 sections) and a subset of ~74,000 court decisions from the highest federal courts issued since 2010. We utilized the QuantLaw package for sophisticated, rule-based reference parsing to build these multi-layered representations.

Results and Discussion

Our analysis reveals that "importance" is inherently perspectival. A law’s centrality changes significantly depending on how it is counted. For example, two-thirds of the most-referenced laws at a low threshold ($k=1$) disappear from the top ranks when requiring at least two references per decision ($k \ge 2$).

Furthermore, we found that "learning by difference" is a powerful tool. By contrasting a "coarse" network (decision-level) with a "fine" network (paragraph-level), researchers can filter out procedural noise. While procedural norms dominate the coarse view, they vanish in the fine-grained view, revealing the substantive legal neighbors. We also identified "dominant" norms—those essential to many other laws but independent of them—by analyzing the bidirectionality of references. The authors visualized this "rank sensitivity," showing how the "most important" nodes shift as you change the minimum reference threshold, proving that a single "top 10" list is often an artifact of arbitrary parameter choices.

Conclusion

Network Pluralism argues that researchers should stop defending a single model and instead map the dimensions of variation in their domain. While multi-network analysis is computationally more expensive and requires more complex data handling, it provides a far more robust understanding of heterogeneous systems. Future work will involve applying this framework to other domains, such as biology or citation science, and utilizing hypergraphs to better capture references to groups of norms.

Summary updated

Abstract

Insights are relative - influenced by a range of factors such as assumptions, scopes, or methods that together define a research perspective. In normative and empirical fields alike, this insight has led to the conclusion that no single perspective can generate complete knowledge. As a response, epistemological pluralism mandates that researchers consider multiple perspectives simultaneously to obtain a holistic understanding of their phenomenon under study. Translating this mandate to network science, our work introduces Network Pluralism as a conceptual framework that leverages multi-perspectivity to yield more complete, meaningful, and robust results. We develop and demonstrate the benefits of this approach via a hands-on analysis of complex legal systems, constructing a network space from references across documents from different branches of government as well as including organizational hierarchy above and fine-grained structure below the document level. Leveraging the resulting heterogeneity in a multi-network analysis, we show how complementing perspectives can help contextualize otherwise high-level findings, how contrasting several networks derived from the same data enables researchers to learn by difference, and how relating metrics to perspectives may increase the transparency and robustness of network-analytical results. To analyze a space of networks as perspectives, researchers need to map dimensions of variation in a given domain to network-modeling decisions and network-metric parameters. While this remains a challenging and inherently interdisciplinary task, our work acts as a blueprint to facilitate the broader adoption of Network Pluralism in domain-driven network research.
BibTeX
@misc{punder2025power,
  author = {Titus Pünder and Corinna Coupette},
  title = {{The Power of Network Pluralism: Multi-Perspective Modeling of Heterogeneous Legal Document Networks}},
  year = {2025},
  eprint = {2512.05679},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2512.05679},
}

Loading related content...

Figure