Methodological Foundations for Probabilistic Geo-Temporal Timelines

Methodological Foundations for Probabilistic Geo-Temporal Timelines: Probabilistic geo-temporal timelines reconstruct historical agents' movements using Bayesian inference over fragmented evidence, assigning probabilities to locations and periods rather than binary assertions. Applied to early modern intellectuals (1550–1750), this methodology addresses archival gaps via evidence weighting, likelihood ratios, and digital tools.

Epistemological Challenges

Surviving evidence favors institutional records (e.g., university matriculations, court rolls), biasing against itinerants like Franciscus Mercurius van Helmont. Biases include confessional disparities (Protestant vs. Catholic archives), war destruction (Thirty Years' War), and esoteric secrecy. Gaps average 1–5 years for comparable figures (Leibniz, Spinoza).

Evidence Hierarchy

Weights (0.0–1.0):

Autograph/dated letter: 1.0 (irrefutable presence). Contemporary witness: 0.9 (e.g., diary entry). Legal/inquisitorial record: 0.75 (movement constraints). Printer gravity: 0.8 (typesetting oversight implies location). Network gravity: 0.6 (associate proximity). Textual inference: 0.4 (MSS access).

Justifications per Van Helmont Geo-Temporal Timeline Dataset examples.

Bayesian Workflow

P(H|E) = [P(E|H) P(H)] / P(E), where H is location hypothesis, E evidence.

Set priors from cohort baselines (e.g., 0.3 mobility in Protestant zones). Compute likelihood ratios, incorporating argument from silence (high P(E|H) absent → low posterior). Thresholds: >0.75 assert; 0.5–0.75 plausible; <0.5 uncertain. Handle dependencies via conditional probabilities.

Geolocation and Gap Management

Use gazetteers (World Historical Gazetteer) for toponyms to coordinates. Gaps via maximum entropy (uniform priors absent evidence); interpolation rules: last/next known with transit priors.

Digital Tools

Recogito for toponym tagging; Nodegoat for networks; QGIS/Kepler.js for visualizations (fuzzy polygons for uncertainty). TEI encoding: <location cert="medium" certainty="0.65">Rome</location>.

Applications

Reconstruct peripatetic esoteric paths; correlate with EMS. Case: Van Helmont's Leiden 1676 (P=0.99) disrupts static models.

Related Concepts

Network Gravity in Intellectual History, Argument from Silence in Bayesian Historiography, Peripatetic Esotericism, Lost Esoteric Archives.