Research Notes and Working Files

Research notes and working files are the hidden intellectual infrastructure of a collection. The objects may survive and a catalogue may name them, but the files often preserve why a collector believed an attribution, how a variant was distinguished, which provenance claims remain uncertain, what evidence was consulted and what work was still unfinished.

Estate planning must therefore do more than keep the papers and digital files. It must enable another responsible person to find them, understand their structure, connect them to the objects, distinguish evidence from hypothesis, protect confidential material and make lawful decisions about preservation, publication, transfer or disposal.

The governing question

Do not ask only, “Where are the collector's notes?” Ask: “Can another person understand, authenticate, preserve and use those notes without the collector being present?”

Collector scenario

A confident catalogue entry with no visible chain of reasoning

A deceased collector's database describes an item as “first printing, second state”. The heirs can repeat that phrase, but the supporting spreadsheet, comparison photographs and expert correspondence are stored elsewhere under private abbreviations. Nobody knows which version is current or whether the conclusion was tentative.

The object has not necessarily lost all value, but its attribution is now harder to defend. A well-prepared archive would preserve the current conclusion, confidence level, evidence relied upon, contrary evidence, superseded versions and the next verification step.

What belongs in the research archive

The category is much wider than formal notebooks or finished scholarship. It includes physical papers, digital working material and knowledge embedded in filing systems, labels and metadata.

Physical

Paper and object-linked records

  • Notebooks, card indexes and acquisition diaries
  • Annotated catalogues, books, photographs and correspondence
  • Variant sheets, measurements, sketches and condition notes
  • Draft entries, restoration histories and unresolved-question lists

Digital

Files, systems and online material

  • Documents, spreadsheets, databases and collection exports
  • Email folders, messages, cloud documents and saved web sources
  • Photographs, audio, video, OCR outputs and presentations
  • Scripts, metadata, website drafts and obsolete-software backups

Embedded

Meaning hidden in the arrangement

  • Filenames, folder structures and colour coding
  • Box labels, shelf order and numbered stickers
  • Comments in spreadsheets and database fields
  • Dealer shorthand, private nicknames and cross-references

Do not tidy before you understand

Reorganising folders, combining notebooks, detaching sticky notes, deleting duplicate-looking images or renaming files can sever the relationships that give the archive meaning. Document the existing structure before changing it unless an urgent conservation or security risk requires intervention.

The five kinds of value

Evidence

Support for claims about an object

Receipts, correspondence, annotated photographs, measurements, auction references and contemporaneous notes may support attribution, provenance, condition history or prior ownership.

Interpretation

The collector's specialist reasoning

Variant systems, maker chronologies, rarity classifications and attribution methods explain how conclusions were reached, even where later research revises them.

Operations

Instructions that help an executor act

Working files may identify specialists, loans, items that must remain together, unfinished catalogue entries, missing certificates or objects requiring urgent review.

Rights

Copyright, confidentiality and ownership

The physical papers, the collector's text, third-party photographs, private correspondence and jointly produced research may each carry different ownership or access rules.

Archive

Historical value beyond the objects

A research archive may document a field, dealer network, manufacturer, collector community or unpublished group of objects and may interest a museum, library or specialist society independently of the collection.

Market value is only one part of the judgement. A file can have little resale value yet remain essential to identifying estate property, proving ownership, returning a loan, explaining a restoration or preventing a tentative claim from being presented as fact.

Separate evidence, interpretation and speculation

Working notes naturally mix observation, judgement and possibility. The estate risk arises when those categories become indistinguishable. A research lead written in shorthand can later be mistaken for a settled conclusion and repeated in a valuation, insurance claim or sale description.

Observed or documented

Evidence

A measurement, dated invoice, visible inscription, identified photograph or attributed statement whose source can be checked.

Reasoned conclusion

Interpretation

A judgement drawn from evidence, such as a probable maker, earlier packaging state or likely production sequence.

Research lead

Speculation

A possibility still requiring work. It may be useful, but it must not become a confident catalogue or sale description merely because the collector is no longer available to qualify it.

Recommended conclusion status

Use a visible status such as confirmed, strongly supported, probable, possible, unresolved, rejected or superseded.

Pair the status with the evidence relied upon, date of assessment, person making the judgement and next verification step.

Build a research archive guide

An executor should not have to become a specialist collector. A short guide should explain what the archive is, where it lives, how it is organised and which records are trustworthy. This guide is often more useful than a larger but unexplained body of carefully stored material.

Guide to my collection research and working files

  • Purpose: what the archive records and why it matters.
  • Scope: subjects, periods, makers, franchises or object groups covered and excluded.
  • Locations: physical cabinets, boxes, drives, devices, cloud accounts, email folders, websites and off-site copies.
  • Organisation: folder logic, colour codes, abbreviations, status marks and filename conventions.
  • Authority hierarchy: which source should prevail when records conflict.
  • Specialist contacts: expertise, relationship, possible conflicts and whether contact has been agreed.
  • Restrictions: confidential, personal, disputed, security-sensitive, embargoed or third-party material.
  • Desired outcome: keep with the collection, transfer separately, donate, publish, digitise, restrict or destroy after appraisal.

State which record wins when sources conflict

  1. Signed or original provenance documents
  2. The current master collection database
  3. Dated object-level research files
  4. Published catalogues or articles
  5. Older spreadsheets and exports
  6. Informal notebooks and loose working notes
  7. Unsorted material awaiting appraisal

Create a master index that links research to objects

The index is the bridge between the archive and the collection. It should be compact enough to maintain and strong enough to show what exists, where it is, what it concerns, whether it is current and what restrictions apply.

Recommended master index fields
FieldPurpose
Research IDStable identifier for the file or file group
Object IDLink to the relevant collection object or group
Title and formatPlain-language description and whether it is a notebook, PDF, email folder, spreadsheet, recording or other format
LocationPhysical box, device, drive, account or repository
Creator and dateWho made or accumulated it and when
StatusActive, complete, superseded, restricted or awaiting review
ReliabilityConfirmed, strongly supported, probable, possible, unresolved, rejected or superseded
Rights and accessCopyright owner, confidentiality class and known restrictions
Related filesCross-references to raw sources, images, correspondence and later conclusions
Priority and destinationPreservation urgency and intended beneficiary, archive, sale or destruction route

Stable identifiers outlast storage locations

Use durable identifiers such as COLL-00426 for an object, RES-00426-A for a research file and IMG-00426-017 for an associated image.

“Top drawer, left side” is a location, not an identity. The identifier should remain constant when the object or file moves.

Preserve the archive's internal logic

Original order and provenance can carry information that individual documents do not. The sequence of papers, the grouping of images and the relationship between a notebook and a box label may explain how the collector worked or how conclusions evolved.

Layer 1

Raw or primary material

  • Original correspondence and invoices
  • Unedited photographs, scans and recordings
  • Measurements, source exports and documentary evidence
  • Catalogues and reference material used in the research

Layer 2

Processed material

  • Cropped or enhanced images
  • Transcripts and OCR outputs
  • Edited audio and normalised spreadsheets
  • Extracted quotations and working comparisons

Layer 3

Interpretation

  • Current conclusions and confidence levels
  • Draft catalogue descriptions
  • Attribution reports and narrative research
  • Rejected and superseded hypotheses

A comprehensible digital structure

COLLECTION_RESEARCH/
├── 00_READ_ME_FIRST/
├── 01_MASTER_CATALOGUE/
├── 02_OBJECT_RESEARCH/
├── 03_PROVENANCE/
├── 04_VARIANTS_AND_TAXONOMY/
├── 05_MARKET_AND_VALUATION/
├── 06_CORRESPONDENCE/
├── 07_PHOTOGRAPHS/
├── 08_PUBLICATIONS_AND_DRAFTS/
├── 09_AUDIO_VIDEO/
├── 90_RESTRICTED/
├── 98_SUPERSEDED/
└── 99_UNSORTED_INBOX/

The precise folder names matter less than consistency and explanation. A useful filename normally includes the object or research identifier, subject, document type, date and version. Names such as “final2”, “use-this-one-revised”, “scan00043” or “IMG_7294” conceal meaning and increase the chance that the wrong file will be treated as authoritative.

Keep superseded conclusions, but label them

Research history can be valuable. Retain older versions where useful, but show the version number, revision date, status, author, change note and the record that superseded it. An executor should never have to infer which “final” file is current.

Protect physical and digital working files differently

Physical files

Preserve papers and their context

  • Keep material dry and away from pipes, damp walls, pests and direct sunlight.
  • Use labelled boxes and maintain a box list.
  • Retain original envelopes and folders where they convey provenance.
  • Photograph arrangements before moving them.
  • Avoid household tape, lamination, erasure and automatic removal of old labels.

Digital files

Preserve access, format and dependencies

  • Maintain a working copy, separate local backup and protected off-site copy.
  • Export important data from proprietary systems into accessible formats.
  • Keep original formats where practicable.
  • Document required software, plugins, fonts, linked folders and licence keys.
  • Test restoration and confirm that another authorised person can follow the process.

Access risk

The files exist but nobody can reach them

  • Unknown passwords or two-factor authentication routes
  • Closed email or cloud accounts
  • Encrypted drives without recovery instructions
  • Subscriptions, domains or hosting allowed to lapse

Format risk

The files can be found but no longer opened

  • Obsolete software or proprietary databases
  • Missing fonts, plugins, linked image folders or lookup tables
  • Old media, operating systems or licence keys
  • Corrupted exports or incomplete migrations

Integrity risk

The archive survives but its meaning is altered

  • Synchronisation propagates accidental deletion
  • Folders are renamed or reorganised
  • Metadata and annotations are stripped
  • Apparently duplicate files are deleted without comparison

Synchronisation is not automatically a backup

A synced deletion, corruption or ransomware event may propagate to every connected device. A trustworthy backup is a separate recoverable copy that has been tested, not merely a service reporting that synchronisation succeeded.

Email, messages and online research

Valuable evidence may exist only in email, specialist forums, marketplace messages, private chats, social groups or shared cloud folders. These services can disappear, remove attachments, close accounts or prevent an executor from accessing material through ordinary credentials.

Preserve durable research outside the platform

  • Save key exchanges with sender, recipient, date and thread context.
  • Preserve attachments and record the platform on which the exchange occurred.
  • Note confidentiality and whether publication permission was given.
  • Record a proper citation and, where lawful, retain a reference copy or archived link.
  • Do not rely solely on bookmarks or an executor impersonating the deceased.

Access must be possible without weakening lifetime security

Record the existence of devices, cloud services, password managers, encryption, recovery codes, two-factor authentication routes, domains and paid subscriptions. Do not place live credentials directly in a will or ordinary archive guide.

Use a separate secure access arrangement agreed with the collector's solicitor and chosen representatives. Possession of a password does not by itself establish legal authority to access an account.

Confidentiality, living people and publication risk

The collector's own death does not make every document suitable for release. Research files may contain personal information about living people, private addresses, sale terms, allegations, security arrangements, ownership lists, disputed claims or statements supplied in confidence.

Suggested confidentiality classifications
ClassMeaning
OpenMay be shared without special review
Estate use onlyFor executors and professional advisers
RestrictedAccessible only to named people or roles
EmbargoedClosed until a stated date or event
Review before releaseRequires legal, archival or specialist assessment
Destroy securelyNo continuing evidential, legal or research purpose after authorised appraisal

Treat allegations and reputational material as a specialist threshold

Unverified accusations about dealers, collectors, makers or experts should be isolated, clearly labelled and reviewed before circulation. A provisional note can create legal and reputational harm when stripped of its context.

Physical ownership, copyright and permission are separate

Inheriting a notebook does not automatically transfer every right in its contents. The estate may own the physical volume and the collector's original text, while a correspondent or photographer retains copyright in letters or images. Joint research, commissioned cataloguing, employee-created material and database contributions may introduce further rights.

Questions to settle during life

  • Who inherits the physical archive?
  • Who inherits the collector's copyright?
  • Who may complete or publish unfinished work?
  • May a beneficiary license images or database content?
  • May an institution digitise and provide access?
  • What attribution, restrictions or embargoes must apply?

Collaborative research

  • Distinguish sole, joint, licensed and third-party material.
  • Record who controls the master database and shared photographs.
  • Agree whether a surviving collaborator may continue the project.
  • Document publication, attribution and revenue-sharing arrangements.
  • Prevent one participant deleting the only shared cloud copy.

Do not confuse research subjects with estate property

A research archive may contain photographs of other people's objects, confidential ownership registers, records of previously sold items, borrowed documents or prototypes held temporarily. The index should mark each referenced item as currently owned, formerly owned, on loan, third-party owned, anonymously held, institutional, missing or destroyed.

Maintain a third-party property register

Record the owner, item or file, date received, purpose, return terms, permitted uses, storage location and contact details. Executors must be able to distinguish estate property from material merely held by the collector.

Unfinished research needs an editorial decision, not a blanket instruction

Partial catalogues, draft books, unattributed objects, unsorted scans and recorded interviews may deserve completion, preservation or privacy. “Publish everything” is rarely adequate because drafts can contain errors, confidential prices, unattributed quotations, unresolved accusations and copyrighted material awaiting permission.

Record an outcome for each significant unfinished project

  • Whether the work should be completed and by whom.
  • Which file is the current draft.
  • Whether publication is authorised.
  • How uncertainty and contrary evidence must be presented.
  • Whether editorial alteration is acceptable.
  • Who may receive income or control licensing.
  • Whether the material should instead remain private.

Research can strengthen value, but only when it reduces uncertainty

Value strengthened

Research substantiates the object

  • Confirms a rare variant or maker
  • Establishes significant provenance
  • Links pieces into a complete group
  • Documents exhibition, publication or restoration history
  • Demonstrates originality and reduces buyer uncertainty

Little value added

The material cannot be used confidently

  • Opinion is undocumented or internally inconsistent
  • Files cannot be connected to the object
  • Sources are obsolete, inaccessible or unattributed
  • Rights or confidentiality prevent lawful use

Confidence reduced

The archive reveals problems

  • Restoration, missing components or disputed attribution
  • Weak or contradictory provenance
  • Earlier descriptions that no longer withstand review
  • Evidence of ownership or authenticity disputes

Unfavourable evidence should not be suppressed. Concealment creates greater legal and reputational risk than a transparent record of uncertainty, restoration or changed attribution.

Private terminology and negative findings are transferable knowledge

Create a glossary

Explain private labels such as “Type IIb”, “early red”, “Doug copy”, “factory set” or “A-state”.

  • Definition and distinguishing features
  • Date the term was introduced
  • Relationship to published terminology
  • Whether other specialists accept the classification

Record what was ruled out

Negative research can save successors years of duplicated work.

  • Sources and archives searched
  • Dates, search terms and people contacted
  • Objects compared
  • Hypotheses rejected and reasons for rejection

Myth versus reality

Myth

The collection catalogue contains everything that matters.

Reality

Catalogues usually record conclusions. Working files preserve the evidence, reasoning, uncertainty and abandoned alternatives behind them.

Myth

My family will understand my filing system.

Reality

Private abbreviations, colour coding, nicknames and implicit links are rarely self-explanatory to a non-specialist.

Myth

Digital files are safer than paper.

Reality

They can be lost through account closure, corruption, encryption, obsolete formats, failed storage or undocumented software dependencies.

Myth

Whoever inherits a notebook may publish it.

Reality

Physical ownership, copyright, confidentiality, third-party rights and permission to publish are separate questions.

Myth

Draft notes are harmless because they were never published.

Reality

Drafts may contain personal data, allegations, confidential prices, incomplete citations or claims the collector intended to verify.

Myth

More paperwork always adds value.

Reality

Documentation helps when it is attributable, intelligible and connected to the object. Disorganised or unsupported notes may increase uncertainty instead.

Collector action hierarchy

Immediate priority

  • Identify every physical and digital research location.
  • Create a plain-language read-me or archive guide.
  • Name at least one person who understands the filing system.
  • Back up irreplaceable files and test that they open.
  • Separate password access arrangements from ordinary research records.
  • Mark confidential, hazardous and third-party material.
  • Record loans and link major files to stable collection identifiers.

Next priority

  • Create a master index and glossary of private terminology.
  • Distinguish evidence, interpretation and speculation.
  • Mark authoritative, superseded and unfinished versions.
  • Export important proprietary databases into accessible formats.
  • Preserve important correspondence outside online platforms.
  • Review copyright, collaboration and publication arrangements.

Advanced planning

  • Record a structured walkthrough or oral-history interview.
  • Arrange archival, conservation or digital-preservation appraisal where warranted.
  • Discuss possible deposit or donation with a suitable institution before naming it in a will.
  • Set publication instructions, access restrictions and embargo periods.
  • Assign the collector's copyright deliberately.
  • Test whether another person can retrieve, understand and explain the archive unaided.

Executor's first-response checklist

Secure

  • Do not dispose of papers or permit unsupervised removal.
  • Secure computers, drives and mobile devices.
  • Preserve folder structures and photograph shelves, boxes and desk arrangements.
  • Protect important subscriptions, cloud services and domain renewals where lawfully possible.

Identify

  • Locate the archive guide, master index and collection-management systems.
  • Find backups, encryption information and current research projects.
  • Identify borrowed, jointly owned or confidential materials.
  • Contact the named specialist or professional adviser.

Stabilise

  • Copy vulnerable digital storage using appropriate methods.
  • Avoid opening or changing obsolete files unnecessarily.
  • Isolate mouldy or water-damaged papers and retain associated labels and containers.
  • Record every movement, intervention and access decision.

Review and decide

  • Separate estate-administration evidence from private research.
  • Review living-person data, copyright, donor and collaborator restrictions.
  • Decide which files support administration, should remain with objects, have independent archival value, require restriction or may eventually be destroyed.

When specialist help is warranted

Professional help is not reserved for obviously valuable manuscripts. An archive can be intellectually important even where no single document appears commercially significant. The relevant specialist may be an archivist, paper conservator, digital-preservation practitioner, solicitor, copyright adviser, data-protection adviser or subject expert.

Specialist threshold

  • The archive is extensive, disordered or distributed across many systems.
  • The collector is recognised in the field or holds unpublished primary sources.
  • A museum, university, library or specialist repository may be a suitable destination.
  • Digital files rely on obsolete hardware, software, databases or encryption.
  • Materials are mouldy, water-damaged, fragile or otherwise unstable.
  • Several collaborators or third parties may hold rights in the material.
  • Confidential correspondence, contested provenance or allegations are substantial.
  • Publication, licensing or royalty income may arise.
  • The archive and the objects may be divided between different beneficiaries.

Decide whether the archive follows the objects

Research should usually remain connected to the objects where it decodes numbering, establishes provenance, explains variant distinctions, identifies restoration or documents why a group must remain intact. The complication arises when one archive relates to objects passing to several beneficiaries.

Custodian model

Keep the master archive together

One person or institution preserves the original archive and supplies indexed copies or access to the owners of related objects.

Object-file model

Relevant files follow each object

Beneficiaries receive object-level files while a master index and reference archive remain intact elsewhere.

Reference-copy model

Preserve originals and distribute copies

The archive remains whole, but lawful digital reference copies accompany divided parts of the collection.

Managed model

Create a custodianship arrangement

A designated custodian manages access, publication, digitisation and continuing costs under written instructions.

Do not assume an institution must accept a bequest

Before naming a museum, university, library or specialist repository, ask whether it collects the subject, accepts digital files, can honour restrictions, will catalogue the material, may split or dispose of parts, and whether copyright is included. Agreement in principle should be obtained during the collector's lifetime.

Appraise before selling or destroying

Before sale

  1. Inventory the archive.
  2. Determine physical and intellectual ownership.
  3. Identify confidential and third-party material.
  4. Preserve a reference copy where lawful.
  5. Clarify what the buyer may publish or reproduce.
  6. Document which objects the files accompany.

Before destruction

Exact duplicates, blank forms, corrupt temporary files and routine administrative material may be candidates for disposal after appraisal.

Do not casually destroy records concerning title, provenance, taxation, loans, authenticity, restoration, insurance, copyright, contracts, disputes or personal data requiring secure handling. Keep a destruction log.

A model research-file record

Object ID
COLL-00426
Research file
RES-00426-A
Subject
Identification of printing state
Current conclusion
Probable second state
Confidence
Strongly supported, not confirmed
Evidence
Logo dimensions, printer code, comparison objects and dated specialist correspondence
Contrary evidence
Undated dealer note describing the item as first state
Unresolved question
Whether the printer code was used concurrently
Next action
Compare with publisher archive production ledger
Current master
RES-00426-A_v05.pdf, superseding versions 1–4
Restrictions
Correspondence restricted; conclusions may be shared
Executor note
Do not describe the state as definitive without specialist review

Minimum viable research-transfer pack

A collector who cannot complete a full archival programme should still create a small pack that prevents the most serious loss of meaning.

  1. A short archive guide
  2. A list of physical and digital storage locations
  3. A current master collection export
  4. A glossary of private terminology
  5. A list of specialists and collaborators
  6. A list of unfinished projects
  7. A confidentiality warning and access classification
  8. A copyright and publication statement
  9. A tested backup
  10. A clear note explaining what must not be separated, renamed or discarded

The ultimate test

A research archive has not been successfully transferred merely because it survives.

It succeeds when a responsible successor can find it, open it, understand its structure, connect it to the objects, distinguish fact from hypothesis, identify the current conclusion, evaluate the evidence, protect confidential material, recognise third-party rights and make an informed decision about preservation, publication, transfer or disposal.

Key takeaways

  • Preserve the collector's reasoning, not only the final catalogue conclusion.
  • Use stable identifiers, a master index and a plain-language archive guide.
  • Separate raw evidence, processed material, interpretation and speculation.
  • Keep original order and context until the archive has been documented.
  • Plan separately for physical preservation, digital access, confidentiality and copyright.
  • Test the system with another person while the collector can still correct it.

Continue learning

Related topics