The second round of contest extends the timescale. The opponents concede that what lies before them can be checked, yet they press further: might the so-called common world be merely the version the victors preserved? Past events did not occur because records of them exist, yet those who come after often can only know them through the traces that remain. Between the events and the events we are able to know stand preservation, classification, retelling, damage, and interpretation. The difficulty of historical controversy arises both from the fact that the past can no longer be reproduced as it was, and from the fact that each generation anew decides which traces deserve entry into common memory.
Imagine a wholly fictional town archive. Decades ago, the riverside market moved from its old site to a new one. Today the townspeople argue over why the old site closed: one account holds that the building was dangerous; another holds that the management committee wanted to hand the land over to a new commercial project; still others recall that both reasons existed at once.
The archive holds several sets of meeting minutes, repair invoices, letters from merchants, two photographs, and a town history written later. The materials are not complete. One set of minutes is missing an annex; the repair invoices cover only some buildings; the letter writers may not have known all of the committee's considerations. If someone called any single document "the complete history," what he omits is not marginal detail but the work that leads from traces to conclusions.
This round of historical debate does not use a fictional story to insinuate any particular event, nor does it treat the story as empirical research. Both sides must jointly honor one distinction: past events, surviving materials, the facts those materials support, the source relations among materials, interpretive frameworks, and value judgments cannot be merged into a single layer within the sentence "history is constructed."
Events Did Not Wait for the Archive to Occur
That the old market once closed is an event posited in advance by the thought experiment. Whether the archive preserves the minutes does not alter the fact that the residents moved out at the time. Even if RC describes determinacy as a secondary construction of observational convergence, it does not follow that those who later kept the records could arbitrarily decide what happened in the past. This restriction comes from the theory's own question of scale. That observers participate in reality does not mean that any single observer monopolizes observation at all levels and across all periods. The residents, the buildings, the transactions, and the bodily consequences of the past all once participated in the process. Today's narratives can change how the past enters the present, but they cannot in reverse make a relocation that already occurred vanish from all its processual relations.
The more accurate formulation is this: later generations' public knowledge of an event depends on the materials available. Deleting a letter may make an experience hard for people today to prove, yet this still does not mean the experience described in the letter never happened. Social acknowledgment and the occurrence of events must be kept apart; otherwise the power to control archives gets inflated into omnipotent control over the past itself.
The error in the opposite direction is just as common. Some say the event objectively occurred, and therefore materials and narratives do not matter. Yet collective decisions require reasons, the pursuit of accountability requires connections, and later generations need to know which accounts have greater support. That events occur independently of archives does not mean we can bypass the archives and possess the full picture of the events directly.
Historical knowledge is therefore under a double constraint: one cannot treat records as the past itself, and one cannot declare the past at will while detached from records. Events constrain interpretation, while materials are the finite channel through which later generations approach that constraint.
A Record Is First a Thing That Still Exists Afterward
The meeting minutes are a document. We can examine the paper or digital version, the heading, the date, the format of writing, and the place of storage. These are facts about the record itself. Whether it accurately presents the discussion of the meeting requires further comparison.
The minutes state "all present agreed to relocate," which can prove that a certain version of the document contains this sentence; it cannot yet, on its own, prove that everyone present in fact agreed at the time. There may have been omissions, summarizing conventions, or later alterations. Raising these possibilities does not amount to proving the document a forgery; it merely marks that an inferential distance still separates the existence of the text from the content of the event.
A merchant's letter saying the committee ignored repair suggestions can support the claim that "someone once made this accusation." To further establish when the suggestions arrived, who had authority to handle them, and whether they were discussed, one must look for delivery records, replies, or other sources. The letter writer's status as a firsthand witness increases the value of certain statements, but even a witness occupies only one part of the event.
A photograph shows a crack in a wall, yet on its own it may not yield when the crack formed, the structural risk, or the motive for the closure decision. An image is not an interpretation-free window onto the event. The angle of the shot, the date, and the identification of the object all belong to source documentation, and judgments of risk require relevant professional materials.
Treating records first as objects blocks a common slide: because a document has official form, its content automatically becomes the event; or because a document contains a standpoint, its content is entirely invalidated. The form of a source affects credibility judgments, but it does not replace item-by-item verification.
Source Relations Matter More Than Source Count
The town history, a commemorative volume, and later articles all state that the market closed because the building was dangerous — apparently three mutually supporting sources. If all three cite the same set of meeting minutes, they constitute first of all a chain of transmission, not three independent confirmations.
A transmission chain still has significance. It can show when an interpretation entered public narrative, how it was repeated, and what status it later acquired. But if the question is the structural risk at the time, repeated publication cannot add new on-site observation to the minutes. Conversely, that two independent records use identical wording does not mean one must have plagiarized the other. They may both cite a common announcement, or they may have formed similar expressions within the same administrative language. Source criticism should not merely compare similarity of sentences; it must also examine knowable dates, paths of access, and shared upstream sources.
When materials are incomplete, source relations can often be expressed only with differing degrees of confidence. We can confirm that the town history's footnote explicitly cites the minutes; we can conjecture that the commemorative volume was influenced by the town history; and we may have no idea where another article came from. Writing "probably dependent" as "plagiarism proven" lets criticism outrun the evidence.
RC's analysis of conceptual consensus reminds us that repeated interpretation can shape the experience and judgment of later generations. But breadth of circulation and historical veracity remain different questions. An account can circulate widely while lacking independent materials, or circulate narrowly yet enjoy comparatively strong documentary support.
What Can an Absence Actually Prove
No formal agreement between the committee and the commercial project party can be found in the archive. Those who favor the structural-risk interpretation say this proves no land deal existed; skeptics say it is precisely because someone destroyed the agreement that nothing can be found. Both conclusions exceed what mere absence can bear.
That no material has been found may be because the matter never happened, or because documents were never drawn up, never filed, later damaged, misclassified, or fell outside the scope of retrieval. Different possibilities require different evidence. One cannot treat a list of possibilities as though their probabilities were already equal, still less pick one item at will and treat it as fact.
Absence can still carry evidential significance. If the rules of the institution required that transactions of a certain kind generate multiple records, and if records of other transactions from the period in question survive intact, then the persistent absence of one record calls for explanation. It may lower the credibility of a certain claim, with a strength that depends on the conditions under which the record should originally have appeared and survived.
Likewise, that a consecutively numbered series of minutes happens to lack exactly the one from the week of the relocation demands separate explanation more than a general loss of materials across an entire decade does. An anomalous pattern can generate a direction for investigation, yet anomaly alone cannot designate agents and motives.
Published writing should use verbs proportionate to the materials: "the record shows," "the source claims," "not yet found," "inconsistent with known procedure," "omission may exist." These expressions are not weak rhetoric; they preserve for the reader the distance from evidence to conclusion.
Memory Is Both Material and a Process of Its Own
One old merchant recalls that the managers at the time promised everyone could return after repairs. Another resident has no memory of such a promise. The conflict among memories is no reason to discard oral materials wholesale; nor does vivid narration exempt it from checking against time and other materials.
Oral accounts can preserve experience that documents never recorded, including fear, expectation, informal communication, and the consequences of decisions in daily life. They can also show how later generations understood the past. Even where a particular detail cannot be confirmed, that understanding itself may be part of the community's history of memory. But "someone now remembers it thus" and "this is indeed how it was then" must be written separately. Memory can be affected by later narratives, repeated retelling, and present positions; this possibility is not an imputation against the teller's character. All records have conditions of formation; the conditions of oral records especially need to be made explicit.
The interviewer also participates in the formation of the material. How questions are posed, whether candidate answers are offered, and how recordings are edited all affect the final text. If the archive keeps only edited summaries while deleting the original recordings and the editorial notes, later generations can hardly distinguish the telling from the editor's summary.
Respecting a witness does not require accepting every interpretation she offers, and verifying details should not become a demand that she prove the entirety of her suffering before deserving to be heard. Historical writing can acknowledge experience, test verifiable facts, and hold to limits regarding what cannot be confirmed — all at once.
Classification Determines Which Materials Meet One Another
The archive files repair invoices under "public safety," merchant letters under "individual grievances," and land planning under "development projects." Classification makes retrieval possible, and it can also keep materials that were once related separated for a long time.
A later researcher who consults only "public safety" will see a comparatively complete narrative of structural risk; one who consults only "commercial development" will obtain another picture altogether. Both derive from real documents, yet each acquires a different sense of completeness from the boundaries of classification. This does not mean classification necessarily distorts. Without a catalog, materials might be harder still to find. The questions are whether the principles of classification are visible, whether cross-indexing is allowed, whether users can know which related materials were placed elsewhere, and whether the catalog can be revised when new connections emerge.
Language's bidirectional representation becomes concrete here: a label helps the archivist process materials, and it also shapes which connections the later researcher sees first. A label is not simply pasted onto already-finished objects; it participates in forming the retrievable structure of history. But that classification influences observation does not mean classification determines the entire meaning of the materials. A letter filed under "individual grievances" may still contain facts of public repair. Researchers can reread and propose new connections. As long as the originals and their source relations remain checkable, the power of classification is not absolute.
Interpretation Places Materials into Causal Relations
One interpretation says the dangerous building caused the market's closure; another says land-development interests pushed the committee to emphasize the danger; a third holds that repair difficulties, fiscal pressure, and development opportunities together altered the choice. The task of interpretation is to show how multiple facts relate to one another, not merely to arrange the materials in sequence.
Causal interpretation requires temporal order, but occurring earlier does not equal causing the later outcome. That the repair report predates the closure decision can constitute a relevant condition; to judge its role in the decision, one must still see how the report entered the meetings, how the decision makers responded, and whether alternative proposals and contrary materials existed.
Motive in particular cannot be inferred solely by working backward from outcomes. That a commercial project ultimately used the old site does not suffice to prove that every supporter of relocation at the time acted for the project's benefit. Structures of interest can generate tendencies, yet specific intentions still require words and deeds, documents, or other evidence. At the same time, the absence of unified intent does not mean structural influence does not exist. Multiple actors each pursuing budgets, efficiency, and development may, taken together, keep draining the repair option of resources. Historical interpretation can articulate such mechanisms without a single mastermind, but it cannot on that ground assign every detail to structure and thereby escape falsification.
Theory here offers questions, not answers. RC's hierarchy of power can prompt us to examine who could define risk, allocate funds, and preserve documents. It cannot determine the motive of the relocation merely from "evaluative authority was concentrated," still less carry out factual investigation for real history beyond the fiction.
Value Judgments Should Not Masquerade as Historical Fact
Even if the materials amply prove that the old market closed because of danger, people today may still argue over whether the decision at the time was justified. Some prize the avoidance of accidents; some hold that an affordable transition should have been provided first; some criticize the exclusion of the merchants from the decision. The clearer the facts, the fewer the value disagreements need not be — only, the disagreements can land more precisely on consequences and responsibility.
The sentence "this was the price progress required" contains a value judgment: that a certain development was worth certain people bearing the loss. The sentence "this was sheer plunder" may likewise contain both fact and evaluation. Writers should indicate which materials support claims about where benefits flowed, and which words express normative judgment of that flow. This is not a demand that historical writing maintain a value-free stance. The very choice of research question is bound up with concerns, and one may critique concentrated harms explicitly. The requirement is that critique display its factual basis and its inferences, rather than fill gaps in the materials with moral intensity.
Value judgments must also face counterevidence. If the relocation genuinely averted grave danger, critics must incorporate that gain; if the compensation offered was usable by only a few, defenders cannot merely list policy names either. Acknowledging unfavorable material does not weaken a position; it lets the position know what trade-offs it bears.
The levels of historical interpretation can thus unfold in order: what the materials are; what they directly support; what bridging conditions are still needed; what competing interpretations exist; what principles the value judgments adopt; what new materials would change the conclusion. Each level can hold a provisional answer; one need not wait for omniscience to begin writing.
The Seams of Power Between Materials and Interpretation
Since public knowledge afterward depends on records, those who control the conditions of preservation can — by not filing documents, shortening retention periods, changing classifications, or restricting access — make certain kinds of experience hard to qualify as history. This capacity is a real institutional possibility, not something to be waved off as "theory misapplied."
The darkest inference runs: there is no opposition in the archive, therefore consensus already existed at the time. An absence of records is first produced by the rules of preservation, then taken as evidence that no dissent existed. Those affected lose not only their materials; they must have this very loss prove that they never objected. This structure does not require burning all documents. So long as the formal minutes fixedly record resolutions without recording principal objections, so long as merchant letters are destroyed as short-term business, and so long as the later town history draws only on the formal minutes, public history can grow tidier layer by layer. Each step can be called routine administration; together they produce selective memory.
To establish that someone in reality deliberately tampered with records still requires concrete evidence. This chapter only indicates how the mechanism could hold: asymmetric preservation rules, shorter lifespans for dissenting materials, higher weight accorded to official sources, and later compilers who no longer examine the absences. Intent and structural consequence must be investigated separately.
Repair cannot proceed by fabrication in reverse either. Knowing that a certain kind of voice may have been omitted does not mean we may write, on behalf of people in the past, the words they might have been hoped to say. The more honest course is to mark the gaps, seek other sources, and restrict conclusions of the form "there was no opposition."
Controlling history has more than one path beyond reducing records. Institutions can also retain vast quantities of duplicated briefings, promotional photographs, and reports of identical format, burying a small number of critical materials in a practically unsearchable collection. Formal openness reaches its maximum while the user's capacity to find declines.
The quantity of information is not a sufficient condition for accessibility. Without catalogs, version relations, and source documentation, several thousand documents may be harder to verify than a deliberately abridged summary. The managers can then say everything has already been made public, shifting responsibility for not finding it entirely onto the public.
Another kind of noise comes from ceaseless re-versioning. Old links expire, file names change, revisions retain no diffs. Each new version may appear more complete, yet the cumulative result leaves citers unable to determine what they originally saw. History is not erased in a single act; it loses traceability through continuous updating.
RC's account of how language can produce information overload through redundant noise has direct critical significance here. But "there is a great deal of material" still cannot automatically prove manipulation. One must examine whether retrieval design, versioning rules, maintenance resources, and known problems have been repeatedly placed in hard-to-reach positions.
Avoiding noise also cannot become an unlimited license for centralized selection. Archivists need to organize, yet they should preserve selection criteria, original materials, and channels of review. Otherwise, tidying in the name of clarity may return once more to the question of who gets to decide which history is worth existing.
The town later wrote the market's relocation into public education. The teaching packet not only states the event but requires learners to accept the interpretation that "rational modernization triumphed over backward operation." Those who raise fiscal interests or the harms of relocation are deemed to be denying the town's collective achievement.
A community can form valuable historical understanding without treating all interpretations as equal. If an account plainly conflicts with reliable materials, educators can point this out. But from "the materials do not support it" to "disloyal" adds a judgment of identity; and from "the shared narrative has a cohesive function" to "cohesion ought to override factual correction" adds a normative leap.
Once interpretation becomes a test of membership, new evidence bears a double cost. It must not only prove the fact but also withstand accusations of betrayal. The original narrative, meanwhile, acquires an exempt status by representing the community. Historical consensus is no longer formed through comparison; it is used to stipulate who is qualified to take part in comparison. This dark path can appear in counter-mainstream groups as well. A group founded on exposing the old narrative may require members to accept every interpretation of oppression and treat those who point to insufficient evidence as defending power. A critical identity likewise cannot grant its own historical judgments permanent exemption.
Conceptual consensus has coordinating and transmissive value; that value must not be bound to loyalty. A person can cherish a community while holding that one of its decisions was wrong; she can also accept a historical fact while rejecting the narrative of heroism or guilt derived from it. Keeping these combinations expressible is what keeps history from being reduced to the testimony of camps.
A later archival catalog divided the merchants of the time into "supporters of development" and "obstructors of development." These two categories may not have existed at the time of the relocation. Later conflicts are projected back onto the past, so that ambiguous, conditional, or unrelated actions all acquire camp significance.
Historical writing inevitably uses later concepts. We cannot return entirely to the vocabulary of the past, and we need concepts to compare different periods. The requirement is to state when a category arose, how far it stands from the self-understanding of the time, and which materials do not fit under it.
If a letter requesting a one-month delay of the move is filed as opposition to development, the letter's concrete demands about timing and compensation are deleted. Conduct is compressed into identity, and identity then interprets all motive on behalf of later generations. This procedure can manufacture a clear history, and it can also leave complex choices without a place.
A darker use is the renaming of the past by a new regime or a new group, sorting former members respectively into lists of those worth commemorating and those to be forgotten. The naming itself does not change what they did, yet it affects later generations' access to materials, to memorial resources, and to standing in public speech. The right to narrate thereby produces real consequences.
To critique this is not to promise a wholly neutral vocabulary. The workable requirement is to preserve original terms, to state analytical categories, to display boundary cases, and to let categories be rebutted by materials. Concepts are instruments of observation, not commands that cut every object into a predetermined shape.
Facing historical gaps, the most direct technical impulse is to record everything. The town decides to preserve all residents' correspondence, locations, and transactions, so that no dispute will ever arise again. Archival completeness thereby becomes a warrant for abolishing private boundaries. This inference is likewise unsound. That something may have research value in the future does not mean there is a right to collect all information today; recording can aid accountability, and it can also expose uninvolved individuals, fix momentary errors, or furnish resources for later control. The value of preservation must be judged jointly with consent, purpose, access, and duration.
The decision processes of certain public offices should bear a heightened duty of recording, because their authority affects others. The private lives of ordinary people do not become public archive in their entirety merely by coexisting with public events. Distinguishing roles and matters allows the materials needed for accountability to remain, without turning every person into a transparent object for future interpretation.
Total recording is also a fantasy. Any system still has to decide formats, precision, and modes of retention. The claim "everything is preserved" hides these choices, and it may further foster the illusion that data equal life itself. The larger the archive, the more it needs to state what it has not recorded. Anti-forgetting, therefore, should not automatically become anti-privacy. If historical justice demands that anyone permanently surrender all traces, it may create a new dispossession of subjects. The aim of protecting neglected experience must be reconciled with the recorded person's reasonable control over the degree of her own exposure.
Why the Right to Interpret Must Be Distributed, Not Abolished
The archive has professionals responsible for restoration, cataloging, and source documentation, and this division of labor can markedly improve the usability of materials. Criticizing the concentration of interpretive authority does not mean abolishing professional judgment or having every material interpreted by everyone simultaneously.
The questions are whether professional positions can explain their methods, whether dissenters can examine the relevant evidence, whether new materials can alter the catalog and the exhibitions, and whether a path of correction exists once an error has occurred. Authority derives from checkable competence and a scope of responsibility, not from a position's ownership of historical truth.
Distributing the right to interpret is not the parallel display of every claim found online. Material strength, source independence, and quality of inference remain comparable. If misinformation appears in volume, the archive can point out the problems explicitly — only, this judgment too should give its grounds, not merely cite its own identity.
Community members hold lived experience that archivists lack; archivists hold capacities for preservation and retrieval that members may lack. The two positions can correct one another. Romanticizing either side as naturally authentic replicates the error of the single observer.
If RC's multi-scale observation is to have constructive significance here, it should support the mutual entry of materials from different positions, rather than install a self-proclaimed ultimate interpreter representing all levels. The theory cannot prove that power should be distributed evenly to everyone, but it can reveal how a single point of evaluative entry narrows visible possibility.
Correcting History Requires Preserving the Correction Itself
The archive eventually discovers that the town history omitted an annex concerning a temporary repair plan. The new edition adds the material and changes "the only feasible plan" to "the plan the committee chose at the time." This correction does not automatically prove the land-interest interpretation; it does indeed weaken the older edition's narrative of inevitability.
If the new edition simply overwrites the old, later generations may not know when public knowledge changed. Preserving versions, reasons for revision, and scope of impact makes the correction part of history. It acknowledges the old error, and it avoids disguising today's knowledge as having always been so.
Correction also has to reach the places where the old narrative has already taken effect. If exhibitions, textbooks, and decision documents continue to use the old conclusion, while a new annex merely sits in the archival database, the effective public interpretation may not have changed at all. The cost of rectification cannot be handed entirely to those who happen upon the update. Yet correction does not require destroying all earlier achievement. The town history's accounts of the timing, the location, and parts of the procedure of the move may still be reliable. Revising according to scope of impact preserves the continuity of public knowledge better than oscillating between a sacred official history and a total hoax.
Processual completeness here is not the sentence "history never has answers," but rather: answers can state their current grounds, accept targeted counterevidence, and preserve the trail of revision. If corrigibility were to mean never bearing clear judgment, it would equally fail to serve common memory.
Writing Accountable History from Incompleteness
The fictional town can in the end produce a layered account: it is confirmed that the market relocated in a certain year; the meeting minutes list structural risk as the reason; the repair materials show that some facilities did indeed have problems, but the surviving documentation is insufficient to prove that every alternative was infeasible; the land project proceeded afterward, and its specific influence on the original decision remains disputed; the losses the merchants experienced and the transition arrangements can be established by other materials. This account is not imposing, yet it gives each sentence its own range of confidence. It neither writes the unknown as a vacancy for anyone to occupy, nor uses incompleteness to cancel what is already supported. Readers can see where they are agreeing to facts, where choosing interpretations, and where making value critiques.
Historical records always come from finite processes. Preservation omits, language compresses, classification alters connections, and today's questions reorganize the past. But "constructed" does not mean "constructible at will." Materials constrain one another, events leave consequences, and erroneous interpretations meet resistance in new records and in the relations of living.
If the deconstruction of civilization merely proclaims that all history is a narrative of power, it will eventually demolish public verification along with the rest. A stricter deconstruction must point out the seams of narrative while preserving the capacity of materials to rebut narrative. It neither enshrines the archive as the past itself, nor hands the past to the most powerful interpreter.
The dark use occurs precisely when these seams are concealed: silence produced by deletion is written as consent, unreachability produced by noise is written as openness, camps produced by classification are written as the essence of the people of the past, and the value of preservation is stretched into permanent surveillance. To critique them requires not merely another story, but making sources, inference, authority, and correction visible again.