Fasti research and evidence

The chosen course

Use structured PIR records to discover candidates, modern studies of each office to establish chronology, and revised inscriptions or papyri to verify the claims. Targeted page reading and OCR then address specific gaps. Missing evidence remains visible at each step.

The supplied files labelled “1938” show first-edition PIR material from 1897–1898. A filename cannot establish an edition. Two scans have searchable Google wrapper text but no body text in the three probed pages. Copies of the same edition or article count as one scholarly source.

Compare collection strategies and checked access

implemented · 17 files + 8 further local file leads

Establish editions and usable access

Verify supplied files against title pages and body probes; retain checksums, printed/PDF page mappings and bibliographic dependency groups. Keep original names. Three 1938-labelled files are first-edition material; two wrapper-only scans need body OCR. Further local works are access leads only.

Accept when: Every admitted source has a checked edition and exact passage; duplicate scans never count as independent evidence.

implemented · entire 15,061-record local PIR snapshot

Generate reviewable candidates

Extract narrow role abbreviations within semicolon-delimited career/occupation clauses. Preserve offsets and fragment hashes, unknown dates, qualified names and unresolved jurisdictions. Keep complete clauses and bibliographic context in private evidence. Publish candidate metadata separately from admitted claims.

Accept when: No candidate enters fasti automatically; no origin/lifespan date is borrowed, no consul date fills an undated proconsulship, and changing source bytes invalidates the receipt. Recall outside recognised forms remains explicitly unmeasured.

implemented · one new governor + five conventus case studies

Verify a cross-source pilot

Admit C. Iulius Severus, proconsul of Asia 152/153, with the article’s explicit PIR² I 573 identification. Keep PIR’s undated Asia clause separate from its circa-139 consulship. Record the assize/presence distinction and uncertain district grouping.

Accept when: Source-local identity supports a specific authority key; exact passage and date basis are visible. Article and PIR citing it count as one dating argument.

partly-implemented · 97–120 CE; then consecutive 25-year batches to 235

Extend the chronological backbone

Check Degrassi pages against images and later revised fasti, with ordinary/suffect roles, iteration and competing readings retained. Revisit 13 CE and 23 BCE independently. Use CIL fasti, military diplomas and annual acta for suffects; missing suffects have no fixed annual denominator. Continuation checked 2026-10-28: ordinary headings 97–105 CE transcribed from Degrassi pp. 29–31 with source-page receipts and selected later checks. Continue later years in bounded source batches; suffects and ordinary/suffect role disputes require separate admission.

Accept when: All names, roles and year headings in each batch checked; exact source receipts and independent correction search logged. Claims of completeness name the edition and scope, never all historical office-holders.

partly-implemented · Asia 130–170; Syria 50–80; Dacia 106–180, separately

Build governor fasti by office and jurisdiction version

Start from existing Magalia claims plus PIR candidates. Collate Thomasson/Piso and regional studies; locate cited inscriptions in revised digital editions and images where available. Adjudicate legatus Augusti, proconsul, praefectus and procurator separately; never classify all procurators as governors. Continuation: Dacia Porolissensis selected attestations are now admitted as separate source records with unresolved period geometry; complete tenures remain unadjudicated.

Accept when: Each admitted row has identity, exact office title, jurisdiction version or unresolved label, date basis and dependency group. First/last attestation is not an office endpoint. All conflicting dates persist as alternatives.

partly-implemented · Piso II entries 73–76: seven diploma attestations and one source-qualified building-inscription office record

Exploit an accessible missing source

Selected printed 111–123 inspected. Quintilianus CIL III 836 is admitted under the source-reported 157 with explicit modern rank/jurisdiction attribution. A 2025 anonymous diploma was checked and does not supply a named 155/156 attestation. Continue regional entries and independent documents without converting proposals or first/last attestations into tenure.

Accept when: Do not infer title date/author/editor from the filename; each extracted entry is image-checked and rank-classified. Targeted OCR is used only where text fails.

partly-implemented · one college or one Tarraco priestly index block at a time

Fill priestly fasti from appropriate sources

Obtain the annual lists/biographies absent from the local Rüpke extract. For Hispania, inspect CIL II²/14 indices and targeted entries, Étienne’s conventus-cult extract and later cult studies. Separate municipal, conventus and provincial priesthoods; distinguish Augustales, actual priesthoods and ancillary personnel. Cycle continuation adds source-local Judaean high-priest removal and appointments from Josephus; this institution remains distinct from Roman colleges and does not fill the missing Rüpke annual rosters. Source batch reviewed 2026-10-29: Tacitus 3.58/3.71 links Maluginensis’ priestly attestation to a provincial request and refusal, with explicit consular-frame dating. No actual Asian governorship, full priestly term or original enactment year is inferred.

Accept when: Exact title and institutional level are source-supported; membership, inauguration and last attestation have separate claims. Reconstructed calator patrons remain hypotheses. No unexamined CIL title becomes a named priestly roster.

partly-implemented · 51 Haensch rows; 47 inherited coordinate matches

Version seats, districts and spatial relations

Use checked Appendix IX as an undated 1997 synthesis. Add province creation/reform, seat validity and district membership from separately read passages and later Haensch studies. A new projection preserves raw analytic wording and separates polarity from uncertainty; kaum is qualified negative.

Accept when: No ancient interval is inferred from the book date, no unlocated seat gets coordinates, and presence/assize/findspot/seat roles remain distinct. Any polygon has a source, valid period, uncertainty and revision receipt.

partly-implemented · Asia/Aristides; Egypt/P.Yale I 61; Tarraconensis

Model institutions and events before mapping them

Keep session, judicial district, resident association, cult institution and editorial corpus area distinct. Record ordered episodes without drawing a known travel path. Reconcile Lewis’s Arsinoe wording with BL 7.282/BL 8.514 and the March 2026 papyri.info correction; current primary text still needs retrieval/inspection.

Accept when: Each assize claim has evidence for judicial activity, not mere presence. Unknown group membership and irregularity remain explicit; no universal annual circuit is inferred from one province.

planned-access-work · 235–305 PIR overlap; 284–395 PPRET pilot; later cohorts separately

Use period-specific authorities

For late antiquity obtain and inspect PLRE and Consuls of the Later Roman Empire. PPRET’s 2026 introduction describes a bounded set of prefectural inscriptions, not all governors. Track changing office structures, East/West recognition and post-consular dating.

Accept when: Authority scope, source revision and edition dependencies are declared; no generic PIR endpoint is extended beyond its coverage and no prefectural sample represents every governor.

implemented-and-ongoing · each admitted batch and each release

Audit selection bias and reproducibility

Report source coverage, extraction success, identity/date/role disputes and source dependencies separately. Compare a manually checked positive/negative sample before widening extraction rules. Test date borrowing, negative evidence, same-name collisions, chronology seams and no-script/mobile access.

Accept when: Regression gates and source hashes pass; measured extraction precision/recall is reported only when a declared gold sample exists. No automated score substitutes for historical adjudication.

A stage marked implemented describes the bounded deliverable named there. Full imperial coverage and independent specialist adjudication remain outstanding. Download the programme and acceptance criteria.

What PIR can supply

The retained JSON snapshot has the same Git object checksum as the official repository file checked on 27 October 2026. Its data-file commit is from 6 May 2021. The live API reports the same total, and six selected ordinary/addendum records were checked. This establishes repository identity and sample access; it does not establish equality of every live field.

Ordinary book entries sampled expose identity and reference information; two public addenda also expose career notes. The raw repository contains additional fields and internal comments. The workbench publishes derived candidate metadata and keeps full clauses and bibliographic context in local review evidence.

Two future-dated modification values are excluded from freshness assessment. The latest non-future record date in this snapshot is 9 January 2020. A recent website footer or catalogue update does not make this a new prosopography.

Official PIR · Official repository · BBAW on completion of PIR² in 2015 · Snapshot and access audit

Reuse what Magalia already holds

The existing campaign layer contains 20 selected governor persons. It supplies identity and command leads; its reported command ranges still require the same admission checks as new research. The retained PIR URL crosswalk supports exact reference lookup. Republican office assertions and hypothetical I.Sicily name clusters retain their distinct source scopes.

Existing data inventory, counts and reuse limits.

Candidate review queue

Each row points to one career or occupation clause. Recognised office words and place labels guide review; they do not establish a governorship, date or identity. Date expressions remain search tokens. A date in another clause is never attached.

The abbreviation rules are deliberately narrow. Unmatched spellings, date prose and unsearched source fields remain a discovery gap. No measured precision or recall is claimed.

Seats and jurisdiction context

Haensch’s Appendix IX is a comparative synthesis of provincial centres. These 51 rows have no ancient validity intervals. Forty-seven names have inherited coordinate matches; four remain unresolved. This table can guide a provincial enquiry, while each dated jurisdiction and seat association still needs evidence for its period.

Britannia’s “kaum” is retained as a qualified negative in the new analytic projection. Empty cells remain unassessed, explicit negatives remain negative, and Ø means not applicable. The earlier parser’s positive classification is preserved alongside the corrected reading for audit.

Find a province or seat

Province / seat wording	Source and coordinate status	Analytic distinctions

The existing governor map retains its source-bounded markers. These are jurisdiction context, not personal presence or an assize route. No geometry or historical boundary is invented.

Conventus: define the evidence before mapping

Keep judicial sessions, judicial districts, resident associations, cult institutions and editorial corpus areas distinct. A visit to a city does not by itself establish a judicial session. A district name does not supply a boundary.

Aristides’ case and C. Iulius Severus

Modern-study sequence in the context of the proconsular year 152/153. Exact episode dates and travel path remain unknown.

1 · Ephesus

Judicial session

Assize context reported by Dalla Rosa.

2 · Smyrna

Presence at the Dionysia

A judicial session is unestablished; a private visit is a qualified interpretation.

3 · Pergamon

Judicial session

Assize context reported by Dalla Rosa.

Inspect the new governor claim and identity basis. Dalla Rosa 2012, p. 268; PIR² I 573 identified at p. 270 n. 80. The underlying Aristides passages still require separate collation.

judicial-session versus presence

C. Iulius Severus in Asia, 152/153

Dalla Rosa follows Aristides’ case through Ephesus, Smyrna and Pergamon. The discussion identifies assizes at Ephesus and Pergamon. At Smyrna, the governor attended the Dionysia; a judicial session is not established. The private-visit interpretation is qualified.

Dalla Rosa 2012, p. 268; identity p. 270 n. 80, PIR² I 573 · Modern study inspected; Aristides passages not independently collated in this run.

Consequence: Keep three source-reported episodes with differing activity evidence. Sequence is from the discussion; exact travel dates and path are unknown.

judicial-administrative session; textual-place dispute

Egypt, petitions and the prefect’s conventus

Lewis reports 1,804 petitions submitted in two-and-a-fraction days in 209, citing P.Yale I 61, lines 5–7. He explicitly rejects equating that statistic with the whole assize duration. The 2026 papyri.info history favours the Arsinoite reading (BL 7.282) over the city reading (BL 8.514).

Lewis 1981, p. 121, lines 5–7 of P.Yale I 61; papyri.info editorial history 25/28 March 2026 · Lewis page checked; current papyrological text/XML not yet successfully retrieved and inspected.

Consequence: Retain Lewis’s reported Arsinoe location as a secondary claim under review. Do not plot a securely located Arsinoe-city session, derive total assize duration or date a complete prefectural tenure from this passage.

conventus civium Romanorum: resident association

Roman conventus in Chalcidice

The article treats a Roman resident association and its fortunes. Direct testimony of the association is distinct from inferences from Roman names or related families.

Loukopoulou 1996, pp. 143–147 · Five-page study read; individual inscriptions not independently re-edited.

Consequence: Represent a corporate association, membership claims and locations with their evidence. Do not map it as an assize district or infer an immigrant route from names alone.

cult institution and judicial district

Conventus cult and priestly fasti in Hispania

Étienne’s supplied file is a chapter extract. Printed p. 12 distinguishes assembly, judicial circumscription and religious use of that framework. The CIL Tarraco volume titles provide corpus geography, not proof of a particular judicial event or priestly tenure.

Étienne 1958, chapter II, printed p. 12; CIL II²/14.2–4 access probes · Étienne title and opening definition checked; no individual priestly catalogue transcribed.

Consequence: Create separate institution and office levels for municipal, conventus and provincial cult. Target the CIL indices/entries and compare later work before extending the priestly fasti.

territorial/resource context

Mining in the conventus Asturum

The files labelled Díez and Reher point to the same Reher et al. study; Reher Díez is a surname. Its landscape/mining context is not itself evidence for a governor’s visit or judicial circuit.

Reher et al. 2012, Configuring the Landscape: Roman Mining in the Conventus Asturum; two supplied copies · Work identity and contextual relevance checked; no comprehensive office extraction.

Consequence: Deduplicate the work; use district/landscape context only where source-supported. Do not turn production sites into itineraries or modern inferred district boundaries.

Typed case and episode data · P.Yale I 61 and its correction history

Source review

Each assessment states its identified work, reviewed scope and limits. Selected passages do not establish complete reading or independent witnesses.

Identified work	Reviewed scope	Limits / consequence
PIR first edition, 1897; part identification provisional pir1-1897-part1	Generic title and selected body probes reviewed.	Part title remains unverified; no new person or office claim is admitted.
PIR¹ pars II, Dessau, 1897 pir1-1897-part2	Part title and selected body probes reviewed.	OCR remains unreliable; first-edition numbering is not PIR² numbering.
PIR first edition, 1897; not PIR² pir1-1897-part1	Title identifies 1897; part identification remains provisional.	Do not identify this material as the second edition. No person or office claim is admitted by this assessment.
PIR¹ pars II, Dessau, 1897; not PIR² pir1-1897-part2	Title and selected body material reviewed.	Same edition as the other pars II representation, not an independent witness; OCR confuses Latin and Greek characters.
PIR¹ pars III, von Rohden and Dessau, 1898; not PIR² pir1-1898-part3	Title reviewed.	First edition; retain edition-specific identifiers.
PIR¹ pars I, Klebs, 1897 pir1-1897-part1	Title and selected body material reviewed.	Search leads require image checking; OCR is not a verified name authority.
Étienne, Le culte impérial dans la péninsule ibérique (1958), selected chapter etienne-1958	Title and printed p. 12, chapter II Le culte de conventus.	Selected chapter only; distinguish assize, district and cult institution. Historical generalizations require further checking.
Naphtali Lewis, The Prefect’s Conventus: Proceedings and Procedures (1981) lewis-1981	Printed pp. 119–121.	Judicial and administrative circuit; Memphis/Arsinoe alternation is qualified. The 1,804 petitions do not establish the full number or duration of an assize.
Loukopoulou, fortunes of the Roman conventus of Chalcidice (1996) loukopoulou-1996	Printed pp. 143–147.	Roman resident association, not a judicial district. Onomastic inference differs from explicit membership.
CIL II²/14.2 (2011), Tarraco cil-ii-14-2	Bibliographic identification and bounded extraction probes only.	No inscription or roster admitted. Corpus title alone does not attest an assize or priest; target numbered inscriptions and revised records.
CIL II²/14.3 (2012), Tarraco cil-ii-14-3	Bibliographic identification and bounded extraction probes only.	No inscription or roster admitted. Separate monument, edition and office claim; verify revised inscription identifiers.
Dalla Rosa, Praktische Lösungen für praktische Probleme, ZPE 183 (2012), 259–276 dalla-rosa-2012	Corresponding selected article passages reviewed.	Same article as the other representation; not an independent study.
Dalla Rosa, Praktische Lösungen für praktische Probleme, ZPE 183 (2012), 259–276 dalla-rosa-2012	Printed pp. 267–271; p. 270 n. 80.	C. Iulius Severus = PIR² I 573, procos. 152/153. Smyrna presence does not establish an assize there; district membership and regularity remain unresolved.
Reher et al., Configuring the Landscape: Roman Mining in the Conventus Asturum (2012) reher-et-al-2012	Title/body identification; no systematic person/office extraction.	Reher Díez is one person’s surname. Deduplicate by work; landscape/mining context does not locate governor travel.
Reher et al., Configuring the Landscape: Roman Mining in the Conventus Asturum (2012) reher-et-al-2012	Title/body identification.	Same work as the other representation. Verify individual office claims separately from territorial/resource context.
CIL II²/14.4 (2016), Tarraco cil-ii-14-4	Bibliographic identification and bounded extraction probes only.	Target priestly/official indices and individual inscriptions; no uninspected roster admitted.
Haensch, Capita provinciarum (1997), Appendix IX haensch-1997	Printed pp. 752–755.	51 province-seat rows; 47 inherited coordinate resolutions and four unresolved. No ancient validity intervals. Britannia’s kaum is a qualified negative, not an attestation.
Ioan Piso, Fasti provinciae Daciae II. Die ritterlichen Amtsträger (2013 bibliographic identity) piso-dacia-ii	Contents VII; foreword IX–X; entries 73–76, printed pp. 111–123.	Selected entries inspected; seven diploma attestations and source-qualified CIL III 836 are incorporated. Literal office wording and modern presidial/provincial interpretation differ. Origins, kinship and immediate succession remain unadmitted. Imprint/edition identification remains qualified; no full collation or document autopsy.
Acta XII Congressus Internationalis Epigraphiae Graecae et Latinae, I (Barcelona 2007) acta-xii-2007-volume-i	Title; Arnaldi p. 87 and Granino Cecere p. 653 no. 62; paper headings pp. 85/643.	One inscription in two dependent studies. Restorations, recipient uncertainty and document-only post-176 dating retained; no whole-volume or 2014 monograph reading.

Further bibliographic leads

Catalogue identification is separate from passage review. These works remain leads until their relevant entries and evidence are checked.

Source review metadata · Bibliographic leads · Consular year gaps and office programmes

Interactive research metadata could not be verified. The sources and methods remain available above.