View source for Network Gravity in Intellectual History

Network gravity models intellectual proximity as inverse-square attraction proportional to node masses (influence, output) and inverse to distance (geographic, confessional). In early modern [[Republic of Letters]], it infers locations from associate clusters, treating scholars as masses in socio-epistemic space.

=== Definition ===
Gravity G_ij = (m_i m_j) / d_ij^2, where m is node mass (e.g., publication count, correspondence volume), d distance (km or affiliation divergence). Bayesian update: P(location_i | associates) ∝ ∑ G_ij.
Priors from baselines ([[Six Degrees of Francis Bacon]] datasets); masses weighted by centrality (PageRank).

=== Workflow ===

Build graph from prosopographies (e.g., CBDB for Chinese parallels, EMLO for Europe).
Segment temporally (decades).
Infer position: Maximize ∑ G for unobserved nodes.
Validate via negative evidence (e.g., silence in high-gravity clusters implies absence).

Implementation: NetworkX/PyMC for inference; Gephi for visualization.

=== Applications ===
Rank diffusion hubs ([[Sulzbach Kabbalistic circle|Sulzbach]] EMS covariate). Comparative: [[Christian Knorr von Rosenroth]] (high local gravity) vs. [[Franciscus Mercurius van Helmont]] (dispersed). Predict archival yields in low-gravity peripheries.

=== Example: Van Helmont ===
Ragley 1670s: High gravity from Conway/More (m=0.8), pulls P=0.95 despite gaps. Sulzbach 1667: Knorr mass elevates posterior from 0.5 to 0.85.

=== Related Concepts ===
[[Esoteric Maneuverability Score]], [[Methodological Foundations for Probabilistic Geo-Temporal Timelines]], [[Socio-Epistemic Networks]].