Primary vs. Secondary Sources¶
Core Idea¶
Primary vs. Secondary Sources is an epistemic classification in which (1) evidence is sorted by its causal and temporal proximity to the phenomenon being studied, (2) materials produced by or directly in contact with the phenomenon at the time of its occurrence are classified as primary (diaries, original recordings, experimental logs, direct testimony, raw sensor output), (3) materials that analyze, synthesize, or interpret primary materials after the fact are classified as secondary (scholarly articles, review papers, edited volumes, summary reports), and (4) the classification governs how each class is used in argument — primary materials supply evidence whose reliability is evaluated at the point of production, secondary materials supply interpretation whose reliability is evaluated against the primary record they claim to interpret, as Howell and Prevenier (2001) systematize for working historians. [1]
The distinction originates in 19th-century German source criticism, where foundational work established systematic rules for assessing documentary evidence. This framework emerged from recognition that not all evidence encountered bears equal relationship to the phenomenon of study; a contemporaneous diary carries different evidentiary weight than a later commentary on that diary, even if both documents survive intact. The classification thus organizes research materials into an accountability chain: secondary sources derive their credibility from their faithful reading of the primary record, and that reading can be audited by returning to the originals.
How would you explain it like I'm…
Hearing It Yourself vs. Hearing About It
Original evidence vs. later analysis
Original sources vs. interpretive sources
Structural Signature¶
A two-layer (or N-layer, with tertiary and beyond) epistemic hierarchy where the proximity of a source to the phenomenon determines its evidentiary role, as Tosh (2010) develops in framing primary materials as inputs to interpretation and secondary materials as post-hoc synthesis that can itself become input to further interpretation. [2] The structural primitive is the distinction between the evidentiary layer and the interpretive layer, with the classification placing each source on one or the other. The hierarchy is phenomenon-relative: a scholarly article is secondary relative to the historical event it studies, but primary relative to the historiographical tradition it participates in.
This N-layer structure becomes apparent in complex research: an oral history interview (primary as testimony) may cite a participant's memoir (secondary to the historical events, primary to the oral tradition being recorded). A systematic content analysis of such interviews becomes tertiary, and meta-analytical work comparing multiple content-analysis studies becomes quaternary. Each layer depends upon faithful representation of the layer beneath it; collapse at any point breaks the accountability chain that makes revision possible. Modern historiography and data science both recognize this: historians speak of source layers; engineers speak of raw data, derived tables, and reported metrics. Both recognize that moving a source up the hierarchy without explicit justification converts assumed-reliable evidence into assumption.
What It Is Not¶
The classification is not a reliability ranking — primary sources can be biased, partial, or fabricated, and secondary sources can be more reliable than any single primary source when they triangulate across many, a point Bloch (1949) makes in treating witness testimony as material to be interrogated at the point of utterance rather than presumed truth-bearing. [3] It is not identical to first-hand vs. second-hand testimony in legal contexts, though it overlaps; legal best-evidence rules track epistemological concerns but organize them by hearsay doctrine rather than layering. It is not fixed across questions — the same document can be primary for one inquiry and secondary for another, a phenomenon-relativity that Berkhofer (1995) treats as the central problem of source criticism. [4]
It is not the same as Triangulation (#281) — triangulation combines sources (primary and secondary) to cross-check claims; the primary/secondary classification is a precondition for triangulation rather than its equivalent. It is not the Hermeneutic Circle (#265) — the hermeneutic circle is about the interpretive back-and-forth; the primary/secondary distinction is about the layering of materials that the circle operates on, an infrastructural-versus-interpretive split that Langlois and Seignobos (1898) instituted as the foundational division between external (source) criticism and internal (interpretive) criticism. [5]
Broad Use¶
Academic research across all evidence-based disciplines, journalism (source-hierarchy judgments), law (hearsay rules, best-evidence rule), intelligence analysis (raw-intel vs. analytical-products distinction), scientific publication (raw data vs. meta-analyses), archival practice, data engineering (raw data vs. derived tables), and software engineering post-mortems (log telemetry vs. incident narratives). The journalistic instantiation is articulated by Kovach and Rosenstiel (2014), who treat verification against primary sources — direct witnesses, original documents — as the discipline that distinguishes journalism from commentary. [6]
The classification appears with regularity in domains far removed from academic history. In epidemiology, raw case-report data (primary) differs structurally from meta-analyses across case studies (secondary). In organizational retrospectives, incident logs (primary) are distinguished from narrative explanations of causality (secondary). In oral history projects, unedited recordings or transcripts (primary) are distinguished from published narratives that synthesize interviews (secondary). The consistent pattern across domains suggests the primary/secondary distinction names a structural property of knowledge systems themselves: the separation between point-of-production evidence and post-hoc interpretation.
Clarity¶
Separating evidence from interpretation is a precondition for evaluating either on its own terms. The classification makes the separation explicit, so that an argument's dependencies on primary materials and on secondary interpretations are both traceable and auditable, a research-practice norm Booth, Colomb, and Williams (2016) codify by requiring writers to distinguish primary, secondary, and tertiary sources and to ground claims in the layer their evidence actually occupies. [7]
Clarity here operates at multiple levels. First, at the level of evidence-evaluation: a document from the period of study invites different scrutiny than a study about that period, because each carries different assumptions about authorial intent and perspective. Second, at the argumentative level: distinguishing where a claim rests on primary evidence versus where it rests on secondary interpretation allows readers (and historians themselves) to identify which components of an argument can be independently tested. Third, at the institutional level: disciplines that lose the ability to trace secondary claims back to primary originals accumulate unchecked interpretive drift, where citations cluster around received wisdom rather than evidence.
Manages Complexity¶
A scholar facing a large corpus can sort it by this classification to locate where independent evidentiary weight sits (primary layer) versus where interpretive commitments accumulate (secondary layer), which makes the corpus navigable and enables revisionist work (see #261) that modifies the secondary layer while preserving accountability to the primary layer. The archival method underpinning this sort is articulated by Duranti (1989), whose revival of diplomatics treats each primary document's authenticity, provenance, and form as the precondition for any secondary use of it. [8]
Complexity management becomes critical in large-scale research. Consider a historian studying long-term institutional change across 200 years: the corpus might include thousands of primary documents (letters, contracts, administrative records, archeological artifacts) and hundreds of secondary interpretations spanning multiple historiographical periods. The primary/secondary sorting enables this scholar to: (a) identify which claims rest on direct documentary evidence; (b) recognize when secondary literature refers to a primary source the historian themselves can examine; © detect interpretive traditions that have accumulated around a primary document through multiple layers of citation; (d) make strategic decisions about where to invest archival effort (examining a heavily-cited primary source directly) versus where secondary synthesis suffices. Without this sorting, the corpus becomes a maze where competing narratives cannot be traced to their evidential grounds.
Abstract Reasoning¶
Displays the general structure of knowledge as layered representation: a base of direct-contact materials, an overlay of interpretations, and an accountability relation between them. This is the same structure found in scientific practice (raw data + analysis), software engineering (production telemetry + aggregated dashboards), and reproducible-analysis practice (version-controlled base + versioned derivation). Jenkinson (1922) gives the canonical formulation in his theory of archives: the archivist's duty is the impartial preservation of authentic primary records so that any later interpretive use can be audited against unaltered originals. [9]
At the most abstract level, the primary/secondary distinction names an asymmetry in any knowledge system: some materials produce claims (primary sources, raw data, direct observation), while other materials interpret or synthesize those productions (secondary sources, analysis, reports). This asymmetry is not merely organizational; it reflects a fundamental epistemic difference. A primary source is what it is regardless of how we interpret it; a secondary source is an interpretation, and its value depends entirely on its fidelity to the materials it interprets. This means the two layers require different evaluative methods. For primary sources, we ask: What are the conditions under which this was produced? What biases, constraints, or motivations shaped its creation? For secondary sources, we ask: How faithfully does this reading represent the primary materials? Where does interpretation become speculation?
Knowledge Transfer¶
Mapping Primary vs. Secondary Sources into data-engineering "source of truth" hierarchies:
| Primary/Secondary component | Data-engineering analogue |
|---|---|
| Primary source | Raw event log, immutable source-of-truth |
| Secondary source | Derived table, aggregated dashboard, reported metric |
| Classification governs use | Primary for accountability, secondary for communication |
| Accountability relation | Secondary must be reproducible from primary |
| Phenomenon-relative status | What's primary for one query is secondary for another |
| Revision at interpretive layer | Re-derivation without touching raw logs |
The transfer paragraph: a well-designed data platform separates an immutable "bronze" layer (raw events, append-only) from "silver" and "gold" layers (cleaned tables, aggregated metrics, reported KPIs). The same primary/secondary discipline appears in the replication crisis literature: the Open Science Collaboration (2015), in attempting to reproduce 100 published psychology findings, found that secondary reports could only be evaluated by returning to primary materials and protocols, and that fewer than half of the original effects replicated when the primary record was actually re-examined. [10] The silver and gold layers are secondary sources in the historian's sense — they interpret and synthesize the primary evidence for communication, but their value depends on their reproducibility from the bronze layer.
Teams that lose the ability to regenerate silver and gold from bronze have collapsed the two layers and lost the capacity for legitimate revision. This is the same failure mode as a historiographical corpus that consists entirely of textbooks without archival access: interpretation can no longer be tested against evidence, and the knowledge system becomes brittle. Wigmore's (1937) chart-method for analyzing judicial proof anticipates this failure mode: every inferential step in an argument must terminate at an item of primary evidence, and chains that loop only among secondary inferences are flagged as evidentially ungrounded. [11] Engineering practice and historiographical practice converge on the same structural requirement — keep primary and secondary distinct, and preserve the accountability relation between them. Modern organizations that preserve "immutable audit logs" (primary) alongside "dashboard reports" (secondary) implement the historian's discipline in real time.
Examples¶
Formal/Abstract¶
A social historian studying the 1918 influenza pandemic in New York consults mortality registers from municipal archives (primary), personal diaries from affected families (primary), and contemporary newspaper reporting (primary for the reporting itself, secondary for the underlying medical events). The historian then references recent epidemiological reconstructions (secondary) and a prior monograph on the pandemic (secondary relative to the events, primary relative to the historiographical tradition). The structural challenge — adjudicating between competing primary witnesses with different production conditions — is the same one Renfrew and Bahn (2016) treat in archaeology, where artifact authentication, stratigraphic context, and provenance must be settled before any secondary synthesis of cultural meaning is attempted. [12]
The classification governs the argument's structure: empirical claims are anchored in primary materials; interpretive moves are defended against the primary record the secondary literature cites. The historian working through competing narratives of the pandemic's origins returns to hospital records, death certificates, and contemporaneous medical correspondence rather than accepting the synthesized account of a secondary monograph. When the secondary literature conflicts, the distinction allows the historian to ask: which author examined the primary sources directly, and which relied on earlier syntheses? This question is unanswerable without the primary/secondary distinction. The same layering applies to visual primary sources: Drucker (2014) argues in Graphesis that charts, diagrams, and images are not transparent reports of underlying data but are themselves primary artifacts whose production conditions must be read before any secondary inference is made from them. [13]
Applied/Industry¶
An incident commander writing a production-outage retrospective distinguishes between raw telemetry (primary: structured logs, distributed traces, alert history, deploy records) and narrative artifacts (secondary: the Slack transcript of the war-room, the on-call engineer's reconstructed sequence, the initial hot-fix commit-message explanation). The retrospective's claims about root cause must remain reproducible from the primary telemetry; the secondary artifacts are used to reconstruct the human decision-making sequence but cannot themselves carry root-cause weight independent of the primary record. This is the same chain-of-custody discipline Saferstein (2017) treats as foundational in forensic science: physical evidence retains its primary-source standing only if every transfer between collection, analysis, and courtroom is documented and unbroken. [14]
A retrospective that collapses the two — treating the initial on-call story as root-cause — is making the same error as a historian treating a textbook summary as evidence for the event it summarizes. When teams stop preserving raw logs and retain only narrative summaries, they lose the ability to revisit the incident later with new questions. A new hypothesis about why the outage occurred cannot be tested if only the story survives, not the telemetry. The epistemic stakes are the same as those Coady (1992) develops in his philosophical analysis of testimony: derivative knowledge through reports remains warranted only as long as the chain back to direct witnesses is preserved and inspectable. [15]
Mapped back to the six-component structural signature: Both examples demonstrate how the classification separates the evidentiary layer (mortality registers, structured logs) from the interpretive layer (monographs, narrative reconstructions). Both show phenomenon-relativity: the newspaper article is primary as a document produced in 1918 but secondary as a historical event. Both require preservation of the accountability relation—secondary claims must be reproducible from primary sources. Both illustrate how revision becomes possible only when the layers remain distinct.
Structural Tensions and Failure Modes¶
T1 — Classification collapse under pressure. When primary materials are lost, destroyed, or never produced, secondary materials are pressed into primary service, often without explicit acknowledgment of the layer shift. The resulting arguments carry interpretive weight that the original classification would not have licensed. Example: the absence of surviving testimony from enslaved people in early American archives creates a gap that scholars fill with secondary commentary on slave narratives that do survive, which then becomes cited as if it were primary evidence of enslaver intent. The layer shift is not acknowledged, and the interpretive distance multiplies.
T2 — Primary-source fetishism. Treating primary materials as automatically more reliable ignores that they are produced with their own framings, biases, and omissions. A single primary source can be more misleading than a careful secondary synthesis of many. The classification orders the materials by proximity, not by reliability. A 15th-century chronicle written by a court historian to glorify a king (primary) may be less trustworthy than a modern historian's synthesis of multiple chronicles and archeological evidence (secondary).
T3 — Phenomenon-relativity unacknowledged. A document's status as primary or secondary depends on the phenomenon under study. Pretending the classification is fixed produces errors when a researcher's framing question shifts — the same archive can support legitimate primary use for one question and only secondary use for another. A historian studying 18th-century attitudes toward disease treats a medical treatise as a primary source for intellectual history but must treat epidemiological claims in that same treatise as secondary (needing to be checked against patient records).
T4 — Secondary-layer self-reference. Citation chains that never return to primary materials produce the illusion of evidentiary grounding while remaining entirely within the interpretive layer. "Everyone cites this claim, so it must be true" is a failure mode in both historiography and journalism; the remedy is returning to the primary source at the base of the citation tree.
T5 — Evaluative method confusion. Treating primary and secondary sources as evaluable by the same standards fails to account for their different modes of failure. A primary source fails by being misleading or incomplete; a secondary source fails by being unfaithful to its primary sources. Applying reliability standards designed for one layer to the other creates category errors: we cannot ask "Is this primary source accurate?" without first specifying "accurate to what?" But we can ask "Does this secondary source faithfully represent the primary materials it cites?" A research culture that conflates these evaluation modes loses the ability to distinguish textual infidelity from evidentiary gaps.
T6 — Operationalization across disciplines. The primary/secondary distinction, while structurally universal, operationalizes differently in each domain. In legal contexts, primary sources include witness testimony and documents produced at the time of events; in scientific contexts, primary sources are original experimental reports; in oral history, primary sources are recorded interviews. A scholar moving between fields may unconsciously apply one discipline's operationalization to another's materials, misclassifying documents or reconstructing accountability chains that do not map across domains. This is not a flaw in the distinction but a sign that the distinction requires domain-specific instantiation.
Structural–Framed Character¶
Primary vs. Secondary Sources is a hybrid on the structural–framed spectrum, leaning toward the framed end with a substantial inherited frame. Part of it is a bare pattern — ranking evidence into layers by how close each piece sits to the phenomenon it concerns — and part of it is the epistemic vocabulary and assumptions of historical scholarship that give the classification its point.
The layered structure does transfer: sorting materials by causal and temporal proximity, with raw firsthand records at the base and after-the-fact analyses built on top, recurs in scientific evidence hierarchies, in journalism's distinction between original reporting and commentary, and in data lineage where raw captures are distinguished from derived products. But the prime is framed by a discipline's account of evidence: what counts as "primary," how proximity confers evidentiary authority, and how sources feed interpretation are notions developed within historiography and tied to its standards of inquiry. The classification carries normative weight about reliability, its home is a scholarly practice rather than a formal relation, and applying it well means adopting an epistemic perspective on which sources to trust. A clear structural ordering keeps it from the far pole, but the evidentiary framing sets it on the framed side of the middle.
Substrate Independence¶
Primary vs. Secondary Sources is a moderately substrate-independent prime — composite 3 / 5 on the substrate-independence scale. Its signature — a proximity-to-phenomenon hierarchy distinguishing direct sources from interpretive ones — is substrate-agnostic in form and recurs across history, journalism, law, and experimental design. The reach across evidence-evaluation contexts is implicit and widely used, but the examples are empty, so the transfer is asserted more than shown. It is a genuinely epistemological pattern with broad applicability, held to the middle band by the lack of explicit cross-domain demonstrations.
- Composite substrate independence — 3 / 5
- Domain breadth — 3 / 5
- Structural abstraction — 4 / 5
- Transfer evidence — 2 / 5
Relationships to Other Primes¶
Parents (1) — more general patterns this builds on
-
Primary vs. Secondary Sources is a decomposition of Classification
Primary-vs-secondary sources is the specific shape classification takes when the entities being sorted are evidentiary materials and the criterion is causal-temporal proximity to the phenomenon under study: primary materials are produced by or in direct contact with the phenomenon at its time of occurrence, secondary materials interpret or synthesize primary materials afterward. It is a structurally-particularized instance of assigning items to discrete categories by explicit criteria, with the added commitment that the classification governs argumentative use — primary materials supply evidence, secondary materials supply interpretation — and that boundary cases require source-criticism.
Path to root: Primary vs. Secondary Sources → Classification
Neighborhood in Abstraction Space¶
Primary vs. Secondary Sources sits among the more crowded primes in the catalog (32nd percentile for distinctiveness): several abstractions describe nearly the same structure, so a description that fits it will tend to fit its neighbors too — transporting it usually means disambiguating within this family rather than landing on it exactly.
Family — Historical Time & Interpretation (11 primes)
Nearest neighbors
- Historical Determinism — 0.82
- Historical Empathy — 0.82
- Historicism — 0.81
- Narrative Construction (in History) — 0.80
- Narrative — 0.80
Computed from structural-signature embeddings · 2026-05-29
Not to Be Confused With¶
Primary vs. Secondary Sources is fundamentally about epistemic layering—the organization of evidence and interpretation into hierarchical tiers based on proximity to the phenomenon under study. It is most readily confused with Synchronic vs. Diachronic Analysis, which also involves temporal organization, but they answer different questions. Synchronic analysis takes a single time-slice and examines relationships within that moment (how do these forces balance at time T?); diachronic analysis examines change across time (how has the system evolved from T0 to T1?). Primary vs. secondary sources, by contrast, ask: "How close was this material produced to the phenomenon?" and "Who produced it—a participant or an interpreter?" A participant's diary from 1918 is diachronic if it tracks change over the year, but primary regardless of its temporal scope, because it was produced in contact with events rather than analyzing them later. A historian writing in 2020 about 1918 is secondary whether she is doing synchronic (analyzing a single moment in 1918) or diachronic (tracking change across decades) analysis. The distinction is orthogonal to the temporal perspective; it concerns the evidentiary relationship between source and phenomenon, not the analytical frame applied to the phenomenon.
Primary vs. Secondary Sources is also distinct from Microhistory vs. Macrohistory, though both involve scope. Macrohistory examines large-scale patterns across centuries and continents (the rise of capitalism, the fall of empires); microhistory examines small-scale local contexts (the life of a village over a generation, family dynamics). But this is a choice about resolution and scope of analysis, not epistemic relationship to evidence. A microhistorian uses primary sources (parish records, wills, family correspondence) just as a macrohistorian does. A scholar writing a macrohistorical synthesis of the Spanish Inquisition is producing a secondary source that may rest on decades of microhistorical primary-source work by many researchers. The distinction between micro and macro is about what size system you analyze; the primary/secondary distinction is about how the materials relate to the system you are analyzing. They are independent dimensions: a source can be micro-level primary (a local family's account of their experience during an event) or macro-level secondary (a national synthesis of microhistories).
Primary vs. Secondary Sources should also be distinguished from Triangulation, which combines multiple sources (primary or secondary) to cross-verify claims. Triangulation is a method of validation—does the claim hold up when checked against different evidence streams? Primary/secondary is a classification of materials—which materials are inputs to interpretation, which are outputs? These are complementary but distinct. A researcher might triangulate by comparing two primary sources (two eyewitness accounts) or by comparing a primary source to a secondary interpretation (does the historical monograph faithfully represent the original documents?). Triangulation presupposes primary/secondary classification; it is not a substitute for it. The classification establishes what kind of epistemic work each source does, and triangulation uses that understanding to strengthen claims through multiple corroborating channels. Without the classification, triangulation becomes confused—mixing different epistemic layers without acknowledging the mix, or treating convergence across different types of sources as stronger than it actually is (two secondary sources agreeing with each other carry less weight than each is independently checking against primary materials).
Solution Archetypes¶
Solution archetypes in the catalog that build on this prime — directly (this prime is a source ingredient) or as a related prime.
Built directly on this prime (1)
Also a related prime in 9 archetypes
- Cascade Initiation Bias Diagnosis and Correction
- Comparative Benchmark Validation
- Contingency-Visibility Across Scales
- Independent Verification Oversight
- Knowledge-Warrant Audit
- Layered Record Accumulation
- Recursive Triangulation of Triangulation
- Stratigraphic Time-Ordering Inference
- Traceability Linking
Notes¶
Foundational methodological prime for the historiography block. Cross-links to triangulation (#281, engineering/ethnography methods overlap) and to hermeneutic_circle (#265) as the interpretive operation performed on the classified corpus. Useful bridge into Pass B work on "reproducible analysis" and "evidence hierarchies" as potential cross-domain solution archetypes.
The primary/secondary distinction is not merely a filing system for historical materials; it is an epistemological commitment to the separability of evidence and interpretation. This commitment surfaces in engineering (immutable logs vs. dashboards), law (testimony vs. hearsay rules), journalism (primary reporting vs. commentary), and data science (raw data vs. derived analytics). Wherever knowledge systems persist, this distinction reappears, suggesting it names something fundamental about how humans build reliable claims: by preserving a base layer of uninterpreted materials and an overlay of interpretation, with an explicit relation between them.
The distinction becomes especially valuable when used preventively — when researchers design their projects to maintain primary/secondary boundaries from the start rather than discovering collapsed layers later. This means: preserving raw data or original documents alongside analysis; documenting the conditions under which materials were produced; tracking which claims rest directly on primary sources and which rest on secondary commentary; and building revision capacity into knowledge systems by maintaining the possibility of returning to primaries with new questions.
References¶
[1] Howell, Martha C., and Walter Prevenier. From Reliable Sources: An Introduction to Historical Methods. Ithaca: Cornell University Press, 2001. Standard graduate-level introduction to historical method; codifies the primary/secondary distinction and the analytic procedure of moving from source identification to interpretation. ↩
[2] Tosh, John. The Pursuit of History: Aims, Methods and New Directions in the Study of History. 5th ed. Harlow: Pearson Longman, 2010. Develops the layered structure of historical sources and treats phenomenon-relativity (the same document can serve as primary or secondary depending on the question being asked) as a defining feature of historiographical reasoning. ↩
[3] Bloch, Marc. Apologie pour l'histoire, ou Métier d'historien. Paris: Armand Colin, 1949. (English: The Historian's Craft, trans. Peter Putnam, Knopf, 1953.) Classic Annales-school treatment of source criticism: witness testimony must be interrogated at the point of utterance, never presumed truth-bearing. ↩
[4] Berkhofer, Robert F., Jr. Beyond the Great Story: History as Text and Discourse. Cambridge, MA: Harvard University Press, 1995. Argues that source classification is question-relative and that the same archive can support legitimate primary use for one inquiry and only secondary use for another. ↩
[5] Langlois, Charles-Victor, and Charles Seignobos. Introduction aux études historiques. Paris: Hachette, 1898. (English: Introduction to the Study of History, trans. G. G. Berry, Duckworth, 1898.) Foundational positivist methodology: separates external criticism (authentication, provenance, source classification) from internal criticism (interpretation), establishing the infrastructural-versus-interpretive split. ↩
[6] Kovach, Bill, and Tom Rosenstiel. The Elements of Journalism: What Newspeople Should Know and the Public Should Expect. 3rd ed. New York: Three Rivers Press, 2014. Treats verification against primary sources — direct witnesses, original documents, on-the-record statements — as the discipline that distinguishes journalism from commentary. ↩
[7] Booth, Wayne C., Gregory G. Colomb, and Joseph M. Williams. The Craft of Research. 4th ed. Chicago: University of Chicago Press, 2016. Practical research methodology: instructs writers to distinguish primary, secondary, and tertiary sources and to ground each claim in the source layer its evidence actually occupies. ↩
[8] Duranti, Luciana. "Diplomatics: New Uses for an Old Science." Archivaria 28 (1989): 7–27. (Series continued through Archivaria 33, 1991–92.) Revives diplomatics as a method for evaluating the authenticity, provenance, and form of primary documents — the precondition for any defensible secondary use. ↩
[9] Jenkinson, H. (1922). A Manual of Archive Administration. Clarendon Press. Foundational archival-science text establishing the principle that unbroken custody is the basis of archival authenticity, and that any single break in the custody chain compromises the evidentiary value of the entire record. ↩
[10] Open Science Collaboration. (2015). Estimating the reproducibility of psychological science. Science, 349(6251), aac4716. Coordinated replication of 100 published psychology experiments: reproduced significant effects in only 36% of cases despite nominal transparency of original methods, dramatizing that disclosed information without shared data, code, and pre-registration is insufficient to support substantive scrutiny. ↩
[11] Wigmore, John Henry. The Science of Judicial Proof, as Given by Logic, Psychology, and General Experience, and Illustrated in Judicial Trials. 3rd ed. Boston: Little, Brown, 1937. Develops the chart-method for analyzing argument structure: every inferential step must terminate at an item of primary evidence, and chains looping only among secondary inferences are flagged as evidentially ungrounded. ↩
[12] Renfrew, Colin, and Paul Bahn. Archaeology: Theories, Methods, and Practice. 7th ed. London: Thames & Hudson, 2016. Standard archaeological methods text; treats artifact authentication, stratigraphic context, and provenance as primary-source operations that must be settled before any secondary synthesis of cultural meaning. ↩
[13] Drucker, Johanna. Graphesis: Visual Forms of Knowledge Production. Cambridge, MA: Harvard University Press, 2014. Argues that charts, diagrams, and images are not transparent reports of underlying data but are themselves primary artifacts whose production conditions must be read before secondary inferences are drawn from them. ↩
[14] Saferstein, Richard. Criminalistics: An Introduction to Forensic Science. 12th ed. Boston: Pearson, 2017. Canonical forensic-science textbook; treats chain of custody as the discipline that preserves physical evidence's primary-source standing through every transfer between collection, analysis, and courtroom. ↩
[15] Coady, C. A. J. Testimony: A Philosophical Study. Oxford: Clarendon Press, 1992. Philosophical analysis of testimony as a source of knowledge; argues that derivative knowledge through reports remains warranted only as long as the chain back to direct witnesses is preserved and inspectable. ↩