ARCHIVAL APPRAISAL · SELECTION · VALUE · RETENTION · TRANSFER · MEMORY

How Archival Appraisal and Selection Work

Archives do not preserve the past in full. They preserve selected evidence of the past. Appraisal is the discipline that decides which records deserve enduring preservation, which should be kept temporarily, which may be sampled, and which can be responsibly destroyed.

Every archive is shaped twice: first by what people and institutions create, and again by what later stewards choose to keep.

This makes appraisal one of the most consequential acts in archival work. A conservation treatment can often be revisited. A catalogue description can be corrected. A record destroyed under an appraisal decision may never be recoverable.

The problem is difficult because archivists decide in the present about possible value in the future. They cannot know every future research question, social concern or legal need. Appraisal therefore relies on evidence, policy, documented reasoning, institutional mission and a willingness to preserve uncertainty rather than pretending selection can be perfect.

The short answer

RECORDS CREATED
  → BUSINESS / SOCIAL CONTEXT
  → RETENTION NEEDS
  → LEGAL / RIGHTS / ACCOUNTABILITY VALUE
  → EVIDENTIAL VALUE
  → INFORMATIONAL / RESEARCH VALUE
  → REPRESENTATIVENESS
  → UNIQUENESS / DUPLICATION
  → COST / RISK / FEASIBILITY
  → APPRAISAL DECISION
  → TEMPORARY RETENTION | SAMPLE | TRANSFER | PRESERVE PERMANENTLY | DISPOSE
  → DOCUMENT REASON
  → REVIEW OVER TIME

1. Appraisal is not judging whether a record is interesting

Interest is subjective and changes with time. Archival appraisal instead asks whether records have continuing value beyond their immediate operational use and whether preserving them supports the mission of the archive.

A mundane ledger can become historically important because it documents thousands of transactions. A famous person’s routine note may be less significant than a complete series showing how an institution actually worked.

2. Records begin with primary value

Records are first created to support administration, law, finance, operations, evidence or communication. This immediate usefulness is sometimes called primary value.

When that active purpose ends, some records still retain enduring value. Appraisal examines that second life.

3. Evidential value shows how an institution functioned

Records can document authority, structure, functions, policies, decisions and processes. They show not only what an institution said publicly, but how it actually operated.

Minutes, policy files, major case records, organisational charts, audit material and decision papers can therefore have value as evidence of institutional behaviour.

4. Informational value reaches beyond the creator

Records may contain information about people, places, events, communities, environments, technologies and social conditions that remains useful long after the creator’s operational need has ended.

A school register, migration record or planning map may be valuable to researchers for questions the original creator never anticipated.

5. Legal and rights value can last for decades

Some records protect rights, obligations, ownership, citizenship, pensions, land, contracts or institutional accountability.

These records can deserve long retention even when they are not likely to become prominent historical research sources.

6. Research value is difficult to predict

Future scholars ask new questions. Records once considered administrative debris can later become important for social, environmental, demographic or technological history.

Appraisal therefore needs humility. It cannot preserve everything, but it should avoid narrow assumptions about what future researchers will consider significant.

7. Appraisal is shaped by mission

A national archive, university archive, corporate archive, community archive and literary archive have different collecting mandates.

A record can be important and still fall outside a repository’s mission. Selection requires the right home, not merely a yes-or-no judgement.

8. Provenance helps appraisal happen at the right level

Archivists often appraise bodies of records in relation to the functions and creators that produced them, rather than evaluating every sheet independently.

This functional perspective helps identify which records best document significant activities and which are repetitive by-products.

9. Function can matter more than office name

Organisations reorganise. One function may move across departments over time.

Appraising only by current office structure risks breaking continuity. Functional appraisal asks which records document the activity regardless of which unit performed it.

10. High-level policy records and frontline records reveal different truths

Executive records show intention and authority. Frontline records show implementation and lived consequences.

A balanced archive often needs both. Preserving only senior decision papers can create an elegant but incomplete institutional history.

11. Representativeness matters

Large repetitive series may contain millions of similar case files. Appraisal asks whether the entire series needs preservation or whether a documented sample can preserve sufficient evidence of the population.

Sampling is powerful only when the method is explicit and future users can understand what the preserved sample represents.

12. Random sampling answers one kind of question

Random samples can support statistical representativeness when the record population is sufficiently understood.

They may miss exceptional or historically important cases by design, so random sampling is not appropriate for every archival purpose.

13. Systematic sampling creates regular coverage

An archive might retain every nth file, one month per year or records from defined intervals.

This can be easier to administer than random selection but may introduce bias if the underlying sequence has hidden patterns.

14. Purposive sampling preserves known significance

Archivists may deliberately preserve cases that document rare events, policy precedents, exceptional outcomes, underrepresented groups or major public interest.

Purposive sampling is interpretive and therefore requires especially clear documentation.

15. Appraisal should examine whose voices are present

Formal institutions often create abundant records about powerful actors and comparatively thin records about people with less administrative authority.

Appraisal can reinforce that imbalance if it preserves only records that look formally important. Community records, oral histories, ephemera and private papers may broaden the evidentiary field.

16. Silence is also an appraisal outcome

If records are never created, archivists cannot preserve them. If they are destroyed before appraisal, the archive inherits the silence.

Appraisal therefore cannot correct every imbalance. It can, however, document gaps and make selection decisions more conscious.

17. Uniqueness matters, but duplication can still have value

An exact duplicate usually has less archival value than a unique record. But duplicate copies can contain annotations, different provenance, better physical condition or evidence of circulation.

“Duplicate” should therefore be established, not assumed from similar appearance.

18. Format should not determine value

An email can document a major decision. A handwritten note can be trivial. A database can contain essential evidence.

Appraisal follows function and content before format, then considers preservation feasibility separately.

19. Preservation cost is real

Some records are expensive to preserve because they are fragile, proprietary, encrypted, oversized or dependent on obsolete technology.

Cost can influence appraisal, but high cost should not automatically defeat high significance. It should trigger a transparent risk-and-resource decision.

20. Digital abundance changes appraisal pressure

Digital systems can generate enormous volumes of logs, versions, backups, messages and machine data.

Cheap storage can delay decision-making, but preservation cost accumulates through migration, security, metadata, privacy review and discovery. Digital abundance therefore increases the need for systematic appraisal.

21. Appraisal should happen before systems are decommissioned

A legacy database may hold records of enduring value but become nearly impossible to extract after the software and staff expertise disappear.

Records managers and archivists should identify archival value before decommissioning, not during emergency shutdown.

22. Retention schedules operationalise appraisal

Retention schedules turn appraisal decisions into rules that staff and systems can apply consistently.

A schedule can state that one record class is destroyed after a defined period while another transfers to archives permanently.

23. Appraisal should be documented

Future researchers may need to know not only what survived, but why other records did not.

Appraisal reports, retention schedules and disposal authorities help reconstruct selection history and protect archivists from appearing to have created accidental gaps.

24. Destruction is part of responsible archives

Archives are not morally required to keep everything. Responsible destruction reduces cost, privacy exposure and discovery noise while allowing resources to concentrate on records of enduring value.

The key is authorised, documented and defensible disposal.

25. Deaccession is different from routine disposal

Deaccession concerns material already accepted into a collection. It may occur because the material falls outside mission, was acquired under mistaken assumptions, duplicates another holding unnecessarily, or is transferred to a more appropriate repository.

Because accession creates institutional commitments, deaccession should follow explicit policy and provenance documentation.

26. Appraisal is not permanent certainty

Collection missions, laws, research interests and technologies change. Some appraisal decisions deserve review.

Reappraisal can identify neglected value, duplicate holdings or records that should move to a more suitable custodial environment.

27. Macroappraisal looks outward

Some archival traditions emphasise analysing the functions of institutions and their relationships to citizens and society before deciding which record series best document those functions.

The strength of this approach is that appraisal begins from what society and institutions do, not merely from the records that happen to survive.

28. Documentation strategy looks across many creators

A documentation strategy can ask how best to document a broad topic, community or activity across several repositories and creators.

This prevents every institution from collecting the same obvious records while less visible areas remain undocumented.

29. Community archives change who defines value

Traditional institutional appraisal can overlook communities whose recordkeeping does not match formal bureaucratic structures.

Community-led appraisal can bring different concepts of significance, ownership, consent and access into the selection process.

30. Privacy can influence appraisal

Records containing personal data may have strong research value and strong privacy risk at the same time.

The answer is not automatically destruction. Preservation can be separated from access through closures, embargoes, redaction or controlled use where lawful and appropriate.

31. Access risk is not the same as preservation value

Highly sensitive records may still deserve permanent preservation. A repository can retain them while restricting access until the risk changes.

Appraisal should avoid confusing “cannot open now” with “should not preserve”.

32. Disaster can force emergency appraisal

After flood, fire, cyberattack or storage failure, institutions may face urgent salvage choices.

Pre-existing appraisal and collection priorities help rescuers know what deserves immediate attention.

33. AI can assist appraisal but not own the judgement

AI can cluster records, identify duplicates, extract entities, classify functions and flag sensitive data across enormous record sets.

These tools can accelerate review, but archival value is normative and contextual. Models should support decisions, not silently make irreversible destruction choices.

34. AI appraisal needs explainability

If a machine recommends destruction, archivists need to know which features or rules produced the recommendation.

Opaque selection at scale can create invisible archival bias.

AI SUGGESTION
  → APPRAISAL RULE
  → SAMPLE REVIEW
  → BIAS / REPRESENTATION CHECK
  → HUMAN AUTHORITY
  → DOCUMENTED DISPOSITION

35. Singapore: public-record appraisal and transfer

The National Archives of Singapore identifies public records of historical and national significance for long-term preservation and works with public agencies on records management and transfer.

This illustrates the lifecycle connection: appraisal is not an isolated archival activity at the end, but part of government recordkeeping and authorised disposition.

36. Appraisal shapes national memory

What public agencies transfer, what private donors preserve, what archivists accept and what institutions destroy all shape the documentary record future generations inherit.

No archive can be complete. The ethical obligation is to make selection systematic, documented and open to review.

37. Failure modes

Failure	What breaks
Keep only prestigious records	Everyday systems and marginal voices disappear.
Keep everything	Privacy, cost and discovery collapse.
Appraise by format	Important digital or informal records are undervalued.
Sample without documenting method	Future researchers cannot interpret representativeness.
Destroy because access is sensitive	Potentially important evidence disappears unnecessarily.
No appraisal record	Selection history becomes invisible.
AI destruction without human authority	Bias can erase records at scale.
Appraise only at system shutdown	Context and export options may already be lost.

38. A practical appraisal checklist

Identify the creator and function.
Understand immediate retention obligations.
Assess evidential and informational value.
Consider rights, accountability and legal value.
Examine representativeness and missing voices.
Identify duplication and uniqueness.
Assess preservation feasibility without letting cost become the only criterion.
Use sampling deliberately where appropriate.
Document the rationale.
Connect the decision to retention and disposal authorities.
Review major appraisal frameworks as institutions and technologies change.

39. The deeper model: appraisal is designed forgetting

Memory systems need forgetting. Without selection, everything becomes noise. Without preservation, everything becomes loss.

TOO MUCH FORGETTING → AMNESIA
TOO LITTLE FORGETTING → ACCUMULATION
DOCUMENTED APPRAISAL → CONTROLLED MEMORY

That is the central paradox of archives: deliberate forgetting can be part of preserving memory well.

The archive is not what escaped destruction by accident. A mature archive is what survived because someone could explain why it mattered.

40. Why appraisal matters to civilisation

Civilisations generate far more records than they can preserve permanently. Selection is unavoidable.

The question is therefore not whether society will forget. It is whether forgetting will happen accidentally, invisibly and unequally—or through a disciplined process that preserves enough evidence for future people to question the past.

Source and authority routes

Continue the Archives and Publishing series

Publication control: Wintour House · eduKate Publishing · value, evidence, retention, preservation and archive gates.

World Return: Before destroying a body of records, ask not only whether the organisation needs them today. Ask what future rights, accountability, research and memory would become impossible if they disappeared—and document the answer.

How Archival Appraisal and Selection Work | From Record Value and Risk to Retention, Transfer and Historical Memory