64 Baker Street is an independent historical reference project focused on the Special Operations Executive (SOE), its personnel, operations, resistance networks, locations, equipment, and surviving documentary record.
Our research standard is based on a simple principle:
Every significant historical claim should be supported by evidence appropriate to that claim, and the strength of our wording should reflect the strength of that evidence.
The history of secret organizations presents unusual research problems. SOE agents operated under aliases and codenames, records were deliberately restricted, files were sometimes destroyed or dispersed, wartime reports were written with incomplete information, and later accounts can contradict both one another and contemporary documentation.
Our research process is therefore designed to distinguish between what is documented, probable, disputed, approximate, and unknown.
1. Begin With a Research Question
Research should begin with a clearly defined question rather than with a conclusion that we intend to prove.
Examples include:
- When was an agent recruited by SOE?
- Which section or circuit did a person serve with?
- Which agents participated in a particular operation?
- What was the objective of an operation?
- When and where did a mission take place?
- What documentary evidence survives?
- Which source first established a particular claim?
- Do contemporary records agree with later histories?
A focused question helps determine what kind of evidence is needed.
It also reduces the risk of collecting only sources that confirm an existing assumption.
2. Start With the Archival Landscape
Before relying on individual documents, we try to understand where relevant records are likely to be held.
The National Archives advises researchers to consult research guides, learn how catalogues are structured, and consider records held by multiple archives rather than assuming that a single keyword search will reveal everything relevant.
For SOE research, this is particularly important because material can be distributed across several collections.
The National Archives identifies the HS series as the principal body of Special Operations Executive records. Personnel files are mainly found in HS 9, while related intelligence, military, diplomatic, and aviation material can occur elsewhere.
An article should therefore not assume that absence from one catalogue search means that no record exists.
3. Our Source Hierarchy
We generally work from the strongest available evidence outward.
Tier 1 — Primary Archival Records
Examples include:
- SOE personnel files;
- operational reports;
- contemporary correspondence;
- mission reports;
- training assessments;
- signals and communications records;
- official wartime memoranda;
- photographs with institutional provenance;
- aircraft records;
- government files;
- resistance reports;
- contemporary diaries and letters.
Primary sources often provide the closest surviving evidence to an event.
They are not automatically correct.
A contemporary document may contain errors, incomplete intelligence, deliberate deception, cover identities, assumptions, or information that was later superseded.
Primary evidence must therefore be interpreted rather than merely repeated.
4. Institutional and Official Sources
High priority is also given to material published or catalogued by institutions such as:
- The National Archives;
- Imperial War Museums;
- National Army Museum;
- national archives of countries in which SOE operated;
- recognized university archives;
- military museums;
- government collections;
- established historical institutions.
Institutional catalogues can provide important provenance, dates, descriptions, and collection references.
However, we distinguish between a catalogue description and the actual contents of the underlying record.
A catalogue entry may describe a file without making every page of that file available.
We should not imply that an archival document has been directly examined when only its catalogue description has been consulted.
5. SOE Personnel Files
Personnel files are among the most important sources for biographies on 64 Baker Street.
The National Archives identifies HS 9 as the principal series containing SOE personnel files. These records can include service information, medical reports, photographs, assessments of suitability and performance, and reports relating to an individual’s activities.
For an agent profile, researchers should attempt to establish, where evidence allows:
- full name;
- alternative names;
- date and place of birth;
- nationality;
- military or civilian background;
- SOE section;
- recruitment;
- training;
- rank;
- operational role;
- codename or alias;
- network or circuit;
- deployment dates;
- operations;
- capture, escape, death, or repatriation;
- decorations;
- post-war life;
- archival references.
Not every field will be known for every person.
Unknown information should remain unknown rather than being filled with assumptions.
6. Exact Archive References Matter
Archive references should be recorded as precisely as possible.
The National Archives recommends including full catalogue references in research notes because accurate citation enables other researchers to identify the same document, verify information, and continue the research.
Where available, our working notes should record:
Institution → Series → File → Item → Page/Folio
For example:
The National Archives, HS 9/xxxx/x
We preserve spacing, punctuation, and reference formatting as presented by the archive.
If a reference has not been verified in the archive catalogue, it should not be published as though it were established.
7. A Catalogue Result Is Evidence of a Record, Not Necessarily Its Contents
This distinction is fundamental.
A Discovery entry showing that a file exists may establish:
- the file’s title;
- archival reference;
- date range;
- holding institution;
- access status.
It does not necessarily establish every statement contained inside that file.
For example, The National Archives catalogue identifies HS 9 as the SOE personnel-file series and provides item-level records, but many individual documents have historically required consultation at Kew rather than online access.
We therefore differentiate between:
Catalogue verified
and
Document examined
where that distinction matters.
8. Cross-Checking Important Facts
Important facts should ideally be supported by more than one independent source.
This is especially important for:
- dates of missions;
- identities;
- codenames;
- arrests;
- executions;
- casualty figures;
- responsibility for operational failures;
- claims about betrayal;
- controversial decisions;
- statements attributed to historical figures.
Multiple websites repeating the same claim do not constitute independent confirmation if they all ultimately copy the same source.
Where possible, we trace a claim back toward its earliest identifiable evidence.
9. Contemporary Records vs Later Histories
Contemporary documents and later historical scholarship serve different purposes.
A wartime report may preserve details unavailable elsewhere but may have been written before all facts were known.
A later historian may have access to records unavailable to wartime participants but may also rely on interpretation.
Neither category automatically overrides the other.
Where later scholarship revises an earlier account, we examine the evidence supporting the revision.
10. Memoirs, Interviews, and Oral Histories
Personal recollections can provide exceptionally valuable information about clandestine operations.
They may preserve:
- conversations;
- emotional experiences;
- interpersonal relationships;
- descriptions of training;
- impressions of operations;
- details absent from administrative records.
But memory is not a recording device.
Accounts written decades after an event can be affected by hindsight, later publications, fading memory, self-presentation, or information learned after the war.
Memoirs should therefore be attributed appropriately.
Instead of writing:
The operation happened this way.
we may write:
In his later memoir, X recalled that…
when the memoir itself is the evidence.
11. Conflicting Evidence
Conflicting sources are expected in intelligence history.
When credible evidence disagrees, researchers should identify:
- what each source claims;
- when each source was created;
- who created it;
- whether the author had direct knowledge;
- whether one account depends on another;
- whether new evidence became available later;
- which interpretation best fits the available documentation.
If uncertainty remains, the published article should preserve that uncertainty.
We do not force an artificial conclusion merely because a clear answer would be more convenient.
12. Names, Aliases, and Codenames
SOE research frequently involves multiple identities.
Researchers should distinguish between:
- legal name;
- birth name;
- married name;
- maiden name;
- operational alias;
- cover identity;
- codename;
- network name;
- radio callsign;
- nickname.
These terms should not be treated as interchangeable.
Alternative spellings should be retained when historically relevant because they can help readers locate archival records.
13. Foreign Names and Transliteration
SOE operated across many linguistic environments.
Names and places may therefore appear in:
- English;
- French;
- Polish;
- Dutch;
- Norwegian;
- Czech;
- Serbo-Croatian;
- Greek;
- Italian;
- German;
- Burmese;
- Malay;
- Thai;
- and other languages.
Where multiple transliterations exist, we normally use the form most recognizable in reliable English-language historical scholarship while recording important alternatives when useful for research.
Historical spelling should not automatically be “corrected” without understanding its context.
14. Dates
Chronology is one of our most important verification checks.
Researchers should differentiate between:
- exact dates;
- approximate dates;
- date ranges;
- dates assigned retrospectively;
- dates recorded differently in competing sources.
If a document states only May 1944, we should not invent a specific day.
Likewise, phrases such as:
probably,
around,
by late 1943,
or
according to one account
should be retained where warranted.
False precision can be more misleading than an honest approximation.
15. Places and Geographic Evidence
Operational geography often requires special caution.
A historical source may identify:
- a country;
- region;
- village;
- railway line;
- safe house;
- landing field;
- drop zone;
- exact building.
Those represent very different levels of geographic precision.
Coordinates should only be published when evidence justifies them.
A modern map marker placed in the center of a village should not be described as the exact site of an operation unless that position has been verified.
16. Operations and Missions
An operation article should attempt to establish:
- official or commonly used operation name;
- date;
- location;
- participating organizations;
- objectives;
- planners;
- agents or units involved;
- method of insertion;
- timeline;
- immediate result;
- wider consequences;
- casualties where applicable;
- surviving archival record.
Operational success should not be judged solely from whether a physical target was destroyed.
Strategic consequences, resistance support, intelligence value, political effects, reprisals, and subsequent operations may also matter.
The National Army Museum’s overview of SOE demonstrates the breadth of its wartime work—from sabotage and support for resistance movements to activities in Europe and Asia—making context essential when assessing individual missions.
17. Networks and Resistance Organizations
SOE rarely operated in isolation.
Research concerning a resistance circuit should attempt to distinguish:
- the SOE network or circuit;
- local resistance groups;
- political organizations;
- military formations;
- liaison missions;
- other Allied organizations.
Names and organizational boundaries could change over time.
A resistance organization described one way in a British report may have understood its own identity differently.
Whenever possible, local scholarship and archival sources from the relevant country should supplement British documentation.
18. Casualty and Arrest Figures
Numbers associated with clandestine warfare can be particularly difficult to verify.
We avoid automatically repeating:
- rounded numbers;
- commemorative estimates;
- figures from unsourced websites;
- totals that combine different categories.
When sources differ, the article should state the range or explain the disagreement.
This is especially important for arrests, executions, civilian reprisals, resistance membership, and claimed sabotage results.
19. Photographs
Images require source verification.
Before identifying a photograph, researchers should try to establish:
- originating archive;
- catalogue number;
- caption supplied by the archive;
- photographer where known;
- date;
- location;
- people depicted.
A caption copied across numerous websites does not become reliable merely through repetition.
When identification remains uncertain, the caption should say so.
20. Historical Maps
Maps are treated as research products rather than decoration.
A reconstructed map should distinguish between:
- documented locations;
- approximate locations;
- inferred routes;
- uncertain locations.
Where useful, the map should explain its source base.
We avoid creating visually precise routes unsupported by historical documentation.
21. Newspapers and Contemporary Reporting
Newspapers can be valuable for:
- public announcements;
- obituaries;
- awards;
- post-war trials;
- political reactions;
- local reporting.
However, wartime newspapers were affected by censorship and limited knowledge of secret operations.
Contemporary newspaper reporting should therefore not automatically override later archival evidence.
22. Wikipedia and General Reference Sources
Wikipedia and other general reference works can be helpful starting points.
They are useful for:
- terminology;
- identifying people;
- locating books;
- discovering archival references;
- identifying alternative spellings.
They should generally not be the final authority for a significant disputed historical claim when stronger sources are accessible.
Researchers should follow references back to their original source whenever possible.
23. Search Engines Are Discovery Tools, Not Sources
Search engines help locate evidence.
Search-result snippets themselves should not ordinarily be used as historical evidence.
Researchers should open the underlying source and evaluate what it actually says.
The same principle applies to automatically generated summaries and knowledge panels.
24. Artificial Intelligence
Artificial intelligence may assist research organization but is not accepted as an authority.
AI can be useful for:
- suggesting search terms;
- organizing notes;
- identifying possible spelling variants;
- summarizing already verified information;
- detecting gaps that merit additional research;
- helping structure a draft.
AI output can also contain fabricated names, dates, quotations, events, and archival references.
The National Archives explicitly advises researchers to verify information supplied by AI and to check AI-generated document references against the archive catalogue.
Accordingly:
No AI-generated archive reference should be published until independently verified.
The same rule applies to quotations, dates, operational details, and biographical claims.
25. Negative Evidence
Failure to find a record is not necessarily evidence that an event did not occur.
Archives are incomplete.
The National Archives notes that intelligence material may have been destroyed, retained by security services, restricted, or never transferred to the archive.
We therefore distinguish between:
“No evidence has been found”
and
“This did not happen.”
Those statements are not equivalent.
26. Research Notes
For substantial articles, working research notes should ideally preserve:
- search terms;
- archive references;
- publication details;
- page numbers;
- URLs;
- access dates where useful;
- unresolved contradictions;
- alternative spellings;
- questions requiring additional research.
This makes future revision significantly easier.
It also allows another editor to understand how an article’s conclusions were reached.
27. Minimum Standard for Agent Profiles
Before publishing a substantial SOE agent biography, we aim to verify as many of the following as reasonably possible:
Identity → Birth → Recruitment → Section → Training → Role → Alias/Codename → Deployment → Network → Operations → Fate → Honours → Archival record
A profile does not need to contain every category.
It should, however, avoid confidently filling gaps that the evidence does not support.
28. Minimum Standard for Operation Articles
An operation article should establish, where possible:
Name → Objective → Date → Location → Participants → Planning → Execution → Outcome → Consequences → Sources
For controversial operations, an additional Evidence and Interpretation section may be required.
29. Minimum Standard for Place Articles
Historical-location pages should seek to establish:
Location → Wartime function → Dates of use → Organization → People associated → Events → Present status → Sources
Where an exact address or coordinate is uncertain, that uncertainty should be stated.
30. Corrections and New Evidence
Research standards do not end when an article is published.
Historical knowledge changes as:
- previously closed files are released;
- collections are digitized;
- new archival references are identified;
- specialist research is published;
- errors in older scholarship are discovered.
We therefore treat publication as the current state of the evidence rather than the final word.
Material corrections should be made when reliable evidence requires them.
31. Reproducible Historical Research
A central goal of our research process is reproducibility.
Another researcher should, wherever practical, be able to follow our references and understand why a factual claim was made.
That is why precise catalogue references matter.
The National Archives emphasizes that proper citation allows researchers to identify records, verify information, and take research further.
The same principle guides 64 Baker Street.
We do not merely want readers to accept what we publish.
We want them to be able to investigate it.
Our Standard
No historical project can eliminate uncertainty, incomplete records, or human error.
The appropriate response is not to pretend those limitations do not exist.
Our standard is to:
find the strongest available evidence, preserve accurate citations, cross-check important claims, distinguish documentation from interpretation, disclose uncertainty, and revise our work when better evidence becomes available.
That is the research standard behind 64 Baker Street and the foundation on which we aim to build a dependable reference to the history of the Special Operations Executive.