Reference entries — structured provenance guidance
This reference groups descriptive entries used in the provenance record. Each section below is written to explain, in archival terms, how a specific part of tracing is recorded and preserved. The language is neutral and descriptive: entries document observed properties, the anchors preserved, and the rules applied when linking and reconstructing sequences. Readers may use these entries as a consistent guide to interpret origin and linkage evidence across repositories and teams.
Event origin
Event origin provides a factual account of the first recorded manifestation of an action. The entry lists a stable identifier for the initiating artefact or entry, the recorded timestamp or bounded time interval when precise timing is unavailable, the originating agent or system where observable, and any immediate contextual notations captured at first recording. The origin entry distinguishes primary observations from subsequent annotations and records the archival carrier for the initial artefact, for example a paper form, a captured message file, or a log entry. When initial records are incomplete, the origin entry documents the available evidence and notes which properties are inferred; inferred properties are clearly marked and linked to the specific primary evidence that supports them. The purpose of the origin entry is to anchor attribution: subsequent linking and reconstruction reference the preserved origin properties to maintain a consistent chain of traceable items across time and teams.
Context anchors
Context anchors are the descriptive elements that preserve the conditions and immediate relationships framing an event. Anchors include captured metadata fields (for example, device or system id, capture method, geolocation where appropriate), proximate records that share temporal or spatial relations, and contemporaneous annotations added by observers. Anchor entries record who added each annotation, the time it was added, and whether the annotation is an observation or an interpretive note. Anchors are preserved in unmodified form whenever possible and stored as structured descriptors to support automated matching and human review. By maintaining explicit anchor records, the provenance record enables later comparison of entries across distinct repositories and prevents inadvertent overwriting of original source material. Anchors also include contextual lineage for annotations themselves so readers can separate primary signals from interpretive layers accrued over time.
Linkage rules
Linkage rules describe the explicit criteria used to associate one entry with another. Each rule lists the matching fields (for example stable identifier patterns, actor role equivalence, timestamp windows, and contextual tag overlaps), the precedence applied when multiple candidate links are available, and the label taxonomy used to characterise the relation (referential, temporal adjacency, or inferred causal connection). Rules are expressed in clear language and where appropriate augmented with structured notation that captures the exact fields and comparison operators used. Applied-link records include the inputs consulted, the rule version in use at the time, and an explicit reference to any inferred steps. This level of documentation permits readers to reapply the same rules against the original inputs and to observe whether link decisions are reproducible. Where inference is used, the record explains how interpretive steps were taken and what evidence supports them, maintaining auditability of link creation and revision over time.
Sequence reconstruction
Sequence reconstruction explains the rules and provenance of steps used to assemble discrete events into an ordered timeline. The entry defines the ordering criteria, explains how concurrent or ambiguous timestamps are handled, and documents tie-break procedures when ordering cannot be derived unambiguously. The reconstruction entry distinguishes between sequences that are directly observed and those that are inferred, and it records the inputs and intermediate reconciliation steps for each inferred ordering decision. For concurrent entries, the record notes whether ordering is treated as indeterminate or whether a defined tie-break rule was applied, and it provides the version identifier of that tie-break rule. All reconstruction steps reference the primary inputs used so readers can trace how an order was assembled and which parts of a sequence are certain versus interpretive. This approach preserves auditability and enables independent reassessment of ordering decisions by later reviewers.
Reference consistency
Reference consistency covers identifier templates, versioning of amended entries, and reconciliation procedures for duplicate or conflicting references. The entry specifies stable identifier formats and how amended entries are linked to prior versions without altering the original artefact. When two references appear to point to the same underlying entity, reconciliation steps are recorded explicitly: criteria used to assert equivalence, the evidence inspected, and the audit path that documents who performed the reconciliation and when. The record includes guidance for provisional or experimental identifiers to prevent premature conflation with stable references. By preserving both stable identifiers and reconciliation decisions, the provenance record enables different observers to map disparate labels to the same underlying entity across time while retaining a traceable audit trail of interpretive actions and versioned links.