{"id":"ded283c9-0817-4f94-be82-f4e34d0b55c2","arxiv_id":"1907.09612","paper_version":1,"verdict":"UNVERDICTED","confidence":"LOW","novelty_score":3.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Uses network models on fragmentary records of Carolingian royal envoys to document structural evolution and increasing local-aristocracy recruitment across reigns.","lead":"This paper applies social network analysis and statistics to historical records of missi dominici, royal agents in the Early Middle Ages, to test ideas about how their organization and recruitment changed over time. A smart generalist might read it to see whether quantitative network methods can extract trends from very incomplete medieval documents.","discovery_kind":"new_application","skeptic_critique":{"model":"grok-4.3","headline":"Fragmentary historical records may not supply an unbiased sample, undermining network metrics and recruitment trend tests","rationale":"The reader’s weakest_assumption is precisely the load-bearing point; the abstract-only limitation is now superseded by the full-text availability, but the data-bias issue remains the single most direct threat to the central empirical claims. A positive sensitivity result would support moving toward CONDITIONAL acceptance; a negative result would keep the verdict UNVERDICTED or lower.","tokens_in":1621,"tokens_out":381,"duration_ms":9958,"concrete_test":"Recompute the key network metrics (e.g., degree distributions, community structure, transition matrices between reigns) and the recruitment-trend regression after (a) randomly deleting an additional 20–30 % of edges/nodes to simulate further missingness and (b) deleting edges preferentially from low-status or peripheral locations; if the headline evolution or amplification effect loses statistical significance under either perturbation, the original results are sensitive to the missing-data mechanism.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The strongest claim rests on network analysis detecting structural evolution and agent-transition changes across reigns, plus statistical confirmation of increasing local-aristocracy recruitment. Both rest on the assumption that the observed assignments and relations form a sufficiently complete and unbiased sample of the underlying population of missi activities. The abstract itself flags “difficulties related to the analyzing of missing data provided by fragmentary historical records,” yet the reader’s weakest_assumption correctly identifies that no explicit correction, imputation model, or sensitivity analysis for non-random missingness is referenced. If missingness correlates with geography, social status, or reign (e.g., central vs. peripheral assignments less likely to survive in charters), then both the reported network evolution and the temporal trend in recruitment could be artifacts of the recording process rather than genuine historical signals.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.3","summary":"The paper applies social network analysis and statistical methods to data on missi dominici during the Early Middle Ages. It claims that network analysis reveals evolution in network structure across the period and changes in agent transitions between reigns, while statistical exploration confirms an increasing tendency over time to recruit agents from the local aristocracy. The abstract explicitly notes difficulties arising from missing data in fragmentary historical records and the challenges of modeling a complex multimodal political network.","tokens_in":1764,"tokens_out":455,"duration_ms":14309,"significance":"If the central claims hold after addressing data limitations, the work would provide quantitative support for existing historiographical views on the evolution of Carolingian central administration and recruitment patterns, demonstrating a productive intersection of network science with medieval history. The explicit acknowledgment of missing-data difficulties in the abstract is a point of transparency that strengthens the manuscript's credibility even if the current analysis does not fully resolve them.","major_comments":[{"comment":"Abstract (final paragraph): The reported statistical confirmation of an 'amplification, with time, of a tendency to recruit the agents among the local aristocracy' is load-bearing for the second main claim, yet the text provides no explicit correction, imputation model, or sensitivity analysis for non-random missingness (e.g., charters more likely to survive for central vs. peripheral assignments). Without such checks, the observed temporal trend cannot be distinguished from recording bias.","section":"Abstract"},{"comment":"Network analysis section (implied by abstract claims on structural evolution): The detection of changes in network structure and agent-transition frameworks across reigns relies on the observed assignments and relations forming a sufficiently complete sample; the manuscript does not report any robustness tests (e.g., edge-removal simulations or subsample analyses) that would quantify how missing links affect the reported structural metrics.","section":"Network analysis"}],"minor_comments":[{"comment":"The abstract mentions 'several models of social network analysis' and 'statistical exploration' without naming the specific algorithms, centrality measures, or regression specifications used; adding these details would improve reproducibility.","section":"Abstract"}],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":{"model":"grok-4.3","summary":"We thank the referee for these constructive comments highlighting the need for greater attention to missing-data issues. We respond point by point below and will make partial revisions to address the concerns while remaining faithful to the limitations of the fragmentary historical sources.","responses":[{"response":"We agree that the absence of formal sensitivity analysis or imputation for non-random missingness weakens the ability to rule out recording bias as an explanation for the observed trend. The manuscript already flags the difficulties of fragmentary records in the abstract and conclusion but does not perform the specific checks suggested. In the revised version we will add a dedicated subsection on potential biases, including simple sensitivity checks that compare the temporal trend under alternative assumptions about underreporting of peripheral assignments, and will further qualify the statistical claim. A full imputation model is not feasible without introducing unsupported assumptions about the survival of charters.","revision_made":"partial","referee_comment":"[Abstract] Abstract (final paragraph): The reported statistical confirmation of an 'amplification, with time, of a tendency to recruit the agents among the local aristocracy' is load-bearing for the second main claim, yet the text provides no explicit correction, imputation model, or sensitivity analysis for non-random missingness (e.g., charters more likely to survive for central vs. peripheral assignments). Without such checks, the observed temporal trend cannot be distinguished from recording bias."},{"response":"The referee correctly identifies that no explicit robustness tests for missing links are reported, even though the abstract acknowledges data incompleteness. In the revised manuscript we will incorporate edge-removal simulations and subsample analyses within the network analysis section to quantify the stability of the detected structural changes and agent-transition patterns. These additions will directly address the concern while respecting the small size and sparsity of the underlying historical dataset.","revision_made":"partial","referee_comment":"[Network analysis] Network analysis section (implied by abstract claims on structural evolution): The detection of changes in network structure and agent-transition frameworks across reigns relies on the observed assignments and relations forming a sufficiently complete sample; the manuscript does not report any robustness tests (e.g., edge-removal simulations or subsample analyses) that would quantify how missing links affect the reported structural metrics."}],"tokens_in":1292,"tokens_out":472,"duration_ms":23829,"standing_objections":[]},"desk_editor":{"model":"grok-4.3","letter":"This paper takes the records of Carolingian missi dominici and applies ordinary social-network measures plus some statistical tests. It finds that the network evolved across reigns and that recruitment increasingly drew from local aristocracy rather than central figures. The specific dataset and the reported temporal pattern are the new pieces; they are not in the earlier literature the abstract cites. The authors are straightforward that the sources are fragmentary and that multimodal modeling is tricky, which is better than pretending the data are clean. That honesty is the main strength. The work stays grounded in the actual charters and assignments instead of forcing a grand narrative. The soft spot is exactly the one the stress-test note flags. The abstract itself calls out missing data, yet nothing indicates they tested whether the gaps are random or tied to geography, status, or reign. If peripheral or lower-status assignments are less likely to survive in the record, then both the network evolution and the recruitment trend could be artifacts of preservation rather than real administrative change. Without sensitivity runs, imputation, or even a simple comparison of observed versus expected coverage, the statistical confirmation of the local-aristocracy tendency rests on an unexamined assumption. The math and methods themselves look routine and correctly applied to what data they have; the citation pattern is narrow but appropriate for the subfield. This is a narrow case study for historians of Carolingian administration who already work with network ideas or for digital-humanities people who want one more quantified medieval example. It will not move broad theory, but it adds a concrete data point. A serious editor should send it to peer review so referees can press on the missing-data handling and see whether the authors can add robustness checks. The core effort is real and the topic is legitimate; the paper just needs that extra layer before the claims can be taken as reliable.","headline":"The paper runs standard network and statistical tools on missi dominici assignment records and reports a temporal shift toward local-aristocracy recruitment plus network-structure changes across reigns, but the fragmentary sources make those trends hard to trust without missing-data checks.","tokens_in":2195,"tokens_out":454,"would_cite":false,"duration_ms":17251,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":{"model":"grok-4.3","evidence":[],"headline":"Historical SNA of Carolingian missi dominici; no RS-shaped machinery","alignment":"orthogonal","rationale":"Paper applies standard network metrics, MDS, and regression to fragmentary prosopographic data on agent-reign transitions and local-aristocracy recruitment. No J-cost, ratio symmetry, φ-ladder, 8-tick periodicity, or parameter-free derivation appears. RS theorems (reality_from_one_distinction, AbsoluteFloorClosure, AlexanderDuality, Cost.FunctionalEquation, etc.) concern forcing of spacetime/constants from a single distinction; this work lies in a domain RS explicitly has no opinion on.","tokens_in":61360,"confidence":"high","tokens_out":146,"duration_ms":4495,"cache_read_input_tokens":38528,"cache_creation_input_tokens":0},"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"grok-4.3","headline":"Missi dominici political networks evolved in structure across reigns while recruitment from local aristocracy increased over time.","keywords":["missi dominici","social network analysis","early middle ages","political networks","historical network analysis","local aristocracy","Carolingian administration"],"falsifier":"A more complete compilation of missi dominici assignments that shows no measurable increase in local-aristocracy recruitment and no structural change across reigns would falsify the reported trends.","tokens_in":2513,"feed_emoji":"","tokens_out":504,"duration_ms":11910,"temperature":0.7,"pith_summary":"This paper applies social network analysis and statistical tests to records of missi dominici assignments in the Early Middle Ages. It models the agents as a multimodal political network linking people, places, and reigns. The analysis shows that network structure changed across the period and that the pattern of agents moving between reigns shifted. Statistical checks also indicate a growing preference for recruiting these central agents from local aristocracy rather than from the royal center. The work notes persistent difficulties in handling incomplete historical data.","feed_headline":"Missi dominici networks show growing local-aristocracy recruitment","feed_subtitle":"Analysis of assignment records reveals structural evolution and a shift toward local agents across successive reigns in the Early Middle Age","key_machinery":"Social network analysis of missi dominici agents, their assignments, and relations between reigns, enriched by statistical exploration of recruitment patterns.","core_discovery":"The study constructs several models of the social network formed by missi dominici and their assignments, then enriches the models with statistical tests of historiographical hypotheses. Network metrics reveal an evolution in overall structure through the studied period together with a change in how agents transitioned between successive reigns. The same data show an amplification over time in the tendency to select agents from the local aristocracy at their places of assignment.","pith_inferences":["Similar network methods could test whether other Carolingian offices followed parallel recruitment shifts.","The observed local-recruitment trend may imply gradual decentralization of royal oversight that later administrative systems inherited.","Robustness checks against missing-data biases would strengthen claims about temporal amplification of local recruitment."],"forward_implications":["Network structure of the agents changed throughout the Early Middle Ages.","The framework governing how agents moved between reigns shifted over time.","Recruitment of agents showed increasing reliance on local aristocracy as the period advanced.","Fragmentary records create measurable difficulties for multimodal network modeling."],"fun_headline_variants":["Missi networks evolve between reigns","Local recruitment rises among missi dominici","Agent transitions change in missi network","Models track missi assignment shifts over time"],"cache_read_input_tokens":2112,"weakest_assumption_plain":"The surviving historical records form a sufficiently unbiased sample of assignments and relations to yield reliable network metrics and statistical conclusions about recruitment trends.","fun_headline_variants_meta":{"raw":{"variants":["Missi networks evolve between reigns","Local recruitment rises among missi dominici","Agent transitions change in missi network","Models track missi assignment shifts over time"]},"model":"grok-4.3","cost_usd":0.003912,"raw_usage":{"total_tokens":1882,"prompt_tokens":581,"num_sources_used":0,"completion_tokens":50,"cost_in_usd_ticks":39115500,"prompt_tokens_details":{"text_tokens":581,"audio_tokens":0,"image_tokens":0,"cached_tokens":64},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":1251,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":581,"tokens_out":50,"duration_ms":7337,"temperature":1.0,"reasoning_tokens":1251,"cache_read_input_tokens":64,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-05-24T17:20:22.404381+00:00","model_set":{"reader":"grok-4.3"},"falsifier":"A more complete compilation of missi dominici assignments that shows no measurable increase in local-aristocracy recruitment and no structural change across reigns would falsify the reported trends.","supporting_citations":[],"review_version":1}