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

TL;DR

Network Pluralism introduces a framework to treat networks as perspectives, enabling multi-network analysis to produce more complete, robust insights in legal networks. The authors apply it to the German legal system, building a network space from judicial and legislative documents, and introduce a dynamic counting method to resolve structural inconsistencies across granularities. Through two case studies, they show how perspective-aware metrics and projections reveal information that single-network analyses miss, while highlighting limitations and data needs. The work provides a blueprint for domain-driven network research and suggests extensions to higher-order networks and cross-domain applications.

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.

Figures (15)

• Figure 1: Law as a corpus of interconnected rules.Rules are the most granular unit of law, and they differ in their level of abstraction regarding the situations and persons addressed (\ref{['fig:rule']}). Documents, such as statutes, rulings, or contracts, are the carriers of rules (\ref{['fig:documents']}). Institutions in different branches of government (as well actors outside government) produce documents as one of their typical outputs (\ref{['fig:branches']}). References connect rules contained in documents of potentially different types, allowing them to build on and complement each other. This is illustrated for the reference relations between statutes (S), regulations (R), and court decisions (D), which together cover key parts of the legal system (\ref{['fig:interplay']}).
• Figure 2: Reference granularity mirrors hierarchical structure. The legislation in our case study shows that most rules are not just distributed horizontally across sections but also distributed vertically across several levels of sub-sections (\ref{['fig:legislation_depth']}). This elaborate substructure is reflected in the references made by judicial decisions, which refer to legislative units with a high level of detail, i.e., often below the section level, and mostly at the most granular level available (\ref{['fig:reference_depth']}).
• Figure 3: Base network. Our non-aggregated network representation includes hierarchical structure both within documents and within the institutions that produce them.
• Figure 4: Accessing the network space requires domain-informed modeling decisions. First, we can focus our bipartite relation on one of the two document sides, resulting in one investigated side and one descriptive side (\ref{['fig:accessing_network_space:calibrate']}). This decision is often determined by a specific research interest. Next, we can filter (\ref{['fig:accessing_network_space:filter']}), group (\ref{['fig:accessing_network_space:group']}), or aggregate (\ref{['fig:accessing_network_space:aggregate']}) the network to yield one or more bipartite networks tailored to our research question(s) (\ref{['fig:accessing_network_space:bipartite']}). These choices depend on the substantive scope as well as the desired level of abstraction. Optionally, we can further derive two projections from the bipartite network (\ref{['fig:accessing_network_space:projections']}), capturing co-referencing or bibliographic coupling.
• Figure 5: Analyzing edges sets as distributions adds context. Source distributions (grouped by courts) show that overall importance tends to be driven by a specific subset of courts. Target distributions indicate that courts predominantly refer to specific subsections of prominent norms. For the source distributions, each court is represented by a box. The box coloring depicts the court-specific section reference frequency, scaled by the court's total number of references. For the target distributions, the first, slightly separated box represents the section, and all subsequent boxes denote subsections. Here, the coloring corresponds to the share of direct references, highlighting that there is no uniform distribution of references among subsections. Statute abbreviations are resolved in \ref{['appendix:abbreviations']}.