Evidence room

How an accusation became a case

A guide to reading witch-trial material without treating the charge as proof.

ARCHIVE / PRIMARY RECORDSources checked 9 October 2026 · interpretation and access limits stated below.
Follow the source trail (3)

Start with a person, not a costume

A witch-trial document is a record of people making decisions about other people. That is the most useful place to begin. A charge, an examination, a witness statement and a verdict do different jobs. They may be bound into the same archive collection, but they are not interchangeable. Before asking whether a story is dramatic, ask what kind of document you are holding and who had the authority to create it. The National Archives brings original material together in teaching collections about accusation, trial and community. Their value lies in the chance to examine the machinery of a case rather than read the past as a sequence of spooky anecdotes.

Use the person’s name whenever the record makes it available. Describe someone as accused of witchcraft rather than silently accepting the label as an identity. A person who appears in a hostile allegation has not necessarily chosen to practise a religion, cast a spell or describe themselves in any particular way. That distinction becomes especially important when modern identities enter the conversation.

Separate the stages

Make four headings in your notes: allegation, reported evidence, procedure and outcome. Under allegation, write what was said to have happened. Under evidence, note who said it, whether they claimed to have seen it, and whether another person repeated the same account. Under procedure, record what the document actually tells you about questioning, testimony or a decision. Under outcome, write only the outcome established by the source. If an outcome is missing, leave it missing.

This simple division prevents a common slide. A complaint about illness becomes a statement that someone caused illness; an accusation becomes a confession; a confession becomes a voluntary admission; an arrest becomes an execution. Each change adds something that the earlier statement did not establish. Your job is not to supply those missing steps because they make a satisfying narrative. Your job is to see which steps can be traced.

Ask who is speaking

A witness, a clerk, an official, a printer and a later historian can all contribute words to what now looks like one text. Note those layers. A surviving account may preserve someone’s speech through the vocabulary of an institution. A printed report may arrange events for readers rather than reproduce every exchange. Even a contemporary text has an audience and a purpose. Being close in date to an event makes a source valuable; it does not make its author detached.

Read descriptions of bodies, behaviour and reputation with care. The description may reveal the observer’s assumptions more clearly than the accused person’s own experience. Avoid extracting a colourful detail and presenting it as the whole life of someone you otherwise know almost nothing about. A small amount of uncertainty is more respectful than an invented biography.

Comparison without a universal story

Ask the same questions of several records, but do not expect identical answers everywhere. Which institution was involved? What dates does the material cover? What did the relevant decision-makers regard as evidence? Who could speak, and whose voice is absent? The word witchcraft can travel across places and periods while the procedures around it change. A general explanation should help you ask better local questions, not make local questions unnecessary.

Similarly, do not use a single case to diagnose an entire community. A record of conflict may preserve an unusual moment precisely because it reached an institution. Ordinary cooperation and quiet disagreement are less likely to occupy the same files. Notice what an archive’s selection makes visible and what it cannot show.

A careful first reading

Choose one document, not an entire famous trial, for your first session. Write its title, archive reference, date and access link. Summarise its claim in two sentences, then write two things it establishes and two things it does not establish. Keep your paraphrase separate from any exact wording. Do not modernise an unfamiliar word into a conclusion until you have checked its context.

End by writing a small next question: whether another statement survives, whether the printed account names its author, or whether a verdict is recorded elsewhere. That is useful progress. You have created a trail another reader can follow, rather than a dramatic story they must take on trust. These are reading practices, not legal conclusions about any individual case.

Put the question to work

Record the source, the claim and one check still needed. Keep your interpretation visibly separate from the record.

Continue at the desk ↗