How to read a grimoire

Before you begin

You open a reproduction of a book presented as ancient. Its title page names Solomon, claims that the text comes from Hebrew, and gives a city and year of printing. Inside are the names of spirits, figures and instructions. You can read those sentences and recognise many of their words. You do not yet know what document you are looking at.

This lesson teaches you how to resolve that problem. You will separate what a book claims, what can be observed in the copy, and what its interpreters propose. The result will be a critical reading record and a reference framework for the book’s components: its voices, lists, images, forms of authority and relationships with readers. You will be able to explain what you are using and what you can conclude from it.

You do not need to know Latin, older forms of French or palaeography. The relevant passages are explained in English. You do need to pay close attention to distinctions: an approximate date is not the same as a printed date; a manuscript is not the same as its edition; a suggestive comparison is not a demonstrated chain of copying.

The main case study is the Grimorium Verum. You will also consult the description of a manuscript in Wellcome Collection, compare forms of organisation with the Arbatel, and work through teaching case files. Historical documents and invented practice examples are identified separately. No teaching example is presented as an archival discovery.

What you will be able to do

• Identify the particular witness used in a reading, recording its institution, reference and the form in which you consulted it.

• Classify a statement as a document’s claim, catalogue information, an observation or an interpretation.

• Assess a proposed relationship between versions using specific evidence and an alternative explanation.

• Write a conclusion that retains relevant uncertainties without treating them as a reason to abandon investigation.

How to work through this lesson

Divide your study into five sessions. In the first, read sections 1–4 and complete the reflection prompts. In the second, work through sections 5–9. In the third, study the structure and historical content covered in sections 10–15. Use section 16 for reference. In the fourth, complete the case file in section 17 before reading its model answer. In the fifth, complete the workshop in section 18 and compare your answers with section 19. As an initial guide, allow four to six hours for reading, consulting sources and writing; adjust the time to your needs.

The aim is not to memorise a list of names. You should be able to explain why a claim goes too far, what evidence supports it, and what further check would improve it.

1. The grimoire as a document

A working category

In this lesson, we use grimoire to mean a book or body of texts transmitting knowledge and instructions regarded as magical within a tradition. This is a working definition. It does not make every book classified in this way an example of a uniform genre, or assume that its producers used the word themselves.

The definition identifies a field of study. We must then distinguish particular materials. One volume may combine prayers, lists, diagrams, recipes and commentary. Another may select only part of that repertoire. A shared attribution to Solomon does not establish that both transmit exactly the same work.

Nor should you automatically equate a magic book with a book used in practice. A copy might circulate as a working tool, a collector’s item, a curiosity, a commodity or material for controversy. Choosing between these possibilities requires evidence. The presence of an instruction shows that it was written or printed, not that anyone carried it out.

Three questions to keep separate

The first question is textual: what does the document say? The second is historical: who produced, altered, acquired or read it? The third concerns its claims: what kind of evidence would we need to establish that what the document promises or recounts actually happened?

These questions may lead to different answers. You can confidently identify a promise of effectiveness without knowing whether any reader tried to test it. You can also document a note recording use without determining what happened afterwards. Study becomes more precise when it preserves these distinctions.

Imagine, as a teaching example, a note reading ‘Tested on Tuesday’. That phrase does not interpret itself. Before treating it as evidence that a procedure was performed, you must examine who wrote it, which passage it refers to, and whether it copies an earlier note. Even if it were a personal annotation, it would document a claim that a test took place. It would not, by itself, establish the cause of a result.

Circulation also transforms content

Owen Davies studies how manuscripts and printed books jointly transmit learned magic. His 2023 analysis encourages us to consider the selection, translation and reuse of materials alongside their supposed lines of origin. Printing did not replace manuscript copying at a stroke. ^[7]

For your reading, this has a practical consequence: do not arrange documents solely by medium. A manuscript may copy a printed book. An edition may incorporate passages from several manuscripts. The medium helps describe the object; relationships between witnesses must be argued on other grounds.

2. Identifying what you are actually reading

Work, witness and copy

The work is the textual unit you are trying to recognise across its manifestations. A witness is a document preserving a form of that text. A copy is the individual object you can locate. In much research, you will use ‘witness’ to emphasise its value for comparison and ‘copy’ to emphasise its material identity.

Think of two copies of the same printed edition. They share much of their text and layout. Yet one may preserve a leaf missing from the other, and only one may contain annotations. If you describe those annotations as features of ‘the edition’, you attribute to every copy what you have observed in just one.

The distinction also works in the opposite direction. Two different objects may reproduce the same typographical error because they belong to the same print run. That error does not establish that the second owner copied the first. The explanation may lie in the edition’s production, before the copies went their separate ways.

How the material reaches you

A digitisation presents images of an object. A transcription offers a reading of its signs. A translation interprets the text in another language. A critical edition compares witnesses and makes decisions about readings. Each resource allows certain checks and limits others.

You need not reject a translation because it mediates the text. You need to know which questions it can answer. If you are investigating how a term is translated, the translation is your object of study. If you are investigating its spelling in a manuscript, you will need to return to the image or to a transcription that explains its principles.

A useful identification

A working record should allow another person to locate your source. Record the title as you use it, institution, shelfmark or identifier, part consulted, type of resource and access date. If there is an edition or translation, note who prepared it.

Do not fill unknown fields with unacknowledged guesses. ‘Author unidentified’ and ‘date proposed by the catalogue’ are useful information. An identified gap shows what remains to be investigated; an invented fact conceals the problem.

3. Reading a title page without taking it at its word

The case of ‘Memphis, 1517’

Joseph H. Peterson’s digital edition of the Grimorium Verum reproduces a title page mentioning a translation from Hebrew and describing Plaingière as ‘Jésuite Dominicain’. Its imprint also gives Memphis, Alibeck and 1517. These are claims made by the document, not yet the results of historical verification. ^[1]

Begin by transcribing what matters without silently correcting it. If you start by writing ‘published in 1817’, you erase the discrepancy you need to study. If you write ‘published in 1517’, you turn the claim into a fact. Your initial note might read: ‘The title page states 1517; the chronological validity of that date needs to be checked.’

A title page deserves attention because it constructs the book’s authority. It may associate the contents with a prestigious figure, a respected language or a claimed antiquity. But you should not attribute a particular intention to the printer simply because the presentation seems suspicious. Explaining who chose those elements, and why, requires further investigation.

Building the argument

The historical documentation of the Society of Jesus dates its papal approval to 27 September 1540. Boston College provides a digital edition of that year’s Formula of the Institute, linked to the apostolic letter Regimini militantis Ecclesiae. ^[3]

You now have two pieces of information to relate: a title page claiming 1517, and an institutional designation incompatible with literally dating the title-page state under examination to that year. The conclusion rests on a specific chronological contradiction, not the book’s antique appearance.

The scope of the conclusion matters. You can reject uncritical acceptance of 1517 for that state of the document. You cannot arrive at 1817 by adding or subtracting a convenient number. Nor have you dated every formula transmitted within the volume. The title page, textual body and materials used may have different histories.

The date 1540 should not mechanically become an exact date for the printed book either. It provides a reference point for detecting the institutional anachronism. A detailed study of the term’s emergence and use, together with the copy’s material composition, would allow the argument to be refined. This lesson does not replace that work with an automatic rule.

What ‘1517 [1817]’ means

Peterson’s bilingual page transcribes a bibliographical description attributed to the British Library for the copy with shelfmark 8632.aa.5. It contains the form ‘1517 [1817]’. Here we are consulting that transcription; we are not claiming to have examined the library’s current record directly. ^[2]

The two years serve different purposes: one reproduces a stated date; the other supplies bibliographical dating. Preserve that distinction in your notes. A date in square brackets does not, by itself, explain the evidence behind its proposal.

A careful formulation would be: ‘The title page states 1517. The bibliographical description transcribed by Peterson adds 1817 in square brackets. The identified anachronism prevents uncritical acceptance of the stated date, but does not suffice to establish the alternative year.’

An incomplete argument and how to repair it

The argument ‘the Jesuits did not exist in 1517, so the book dates from 1817’ contains a leap. Its first part challenges one date; the second introduces another that needs evidence of its own. Repairing the argument means separating those operations, citing the origin of each, and acknowledging what remains to be checked.

Nor is ‘the whole book is a fake’ an adequate conclusion. A false or inaccurate detail on the title page does not make the document disappear as a historical source. The object still provides information about its production and reception. What changes is the confidence you can place in one particular claim.

4. Reading the catalogue alongside the document

A second dating problem

Wellcome Collection describes L’art Magique under the reference MS.983 and dates it to around 1709. Its catalogue notes that the title page contains ‘MCCCIX’, possibly in error for 1709. It also notes that the general title does not accurately describe all the contents brought together in the volume. ^[4]

This case is useful because it requires you to retain three levels: an inscription, the identification of a problem, and a proposed interpretation. If you record only ‘1709’, you lose the reasoning. If you record only the inscribed date, you conceal the catalogue’s qualification. Your record should preserve both and explain their relationship.

The difference can be read precisely: MCCCIX represents 1309; 1709 would be written MDCCIX. This helps explain the discrepancy described, but does not resolve it through automatic correction. The catalogue adds that the date recurs in the title of another part of the volume. The proposed error must remain identified as a proposal, however reasonable it seems. ^[4]

An approximate date is not an awkward way of expressing an exact one. It tells us about the available degree of precision. Removing ‘around’ increases the certainty of the claim without adding any evidence.

The title is not the contents

The description of MS.983 identifies several textual units within the volume. You must therefore distinguish the title of the collection from the particular text you are using. A name written at the beginning may introduce a compilation without describing every part of it. ^[4]

This affects citation. If your observation concerns a section, locate that section. Saying ‘the manuscript states’ may suffice in an initial conversation, but not in a comparison where it matters which part transmits each reading.

Suppose a modern edition selects the first unit from a composite volume. It would be a mistake to attribute to that selection everything the catalogue says about the whole volume. It would also be a mistake to conclude that an omitted unit was never there. Editorial selection and material contents answer different questions.

Descriptions can change

Catalogues are research tools maintained by institutions. They may retain older descriptions, incorporate corrections or distinguish alternative interpretations. Recording the link and access date allows you to explain which information you used.

Caution does not mean treating every catalogue statement as an arbitrary opinion. It means recognising its origin, reading its notes and preserving its qualifications. An institutional description usually provides a better starting point than an unattributed label, but it remains a description produced through human work.

A basic record for this case

The record does not claim that you have seen something you know only through the catalogue. This apparently small distinction is essential to traceability.

5. Preparing to compare versions

A focused question

‘Which grimoire is authentic?’ is usually too broad a question. It may conflate antiquity, completeness, religious authority and textual fidelity. Before comparing, turn that concern into an investigable question: ‘Do these two witnesses share a particular omission?’ or ‘In which witness does this arrangement first appear within the material consulted?’

The second formulation includes a limit: within the material consulted. It does not identify the arrangement’s absolute origin. A witness that has not survived or has not yet been identified could change the result.

Start with a manageable unit: a short list, a page reference, or a passage with a recognisable beginning and end. Comparing entire books without recording specific points produces impressions that are difficult to verify. Comparing a unit lets you show someone else what you observed and how you reached your conclusion.

This lesson’s historical case file

Peterson identifies the printed witnesses associated with Alibeck, Blocquel, Bestetti’s 1868 edition and Muzzi’s 1880 edition as relevant. His introduction argues that the Italian editions depend on the French ones. This is the editor’s conclusion: using it requires attribution to him and consultation of his argument when you wish to assess it. ^[1]

Peterson’s digital edition of book three of the Clavicules du Roi Salomon, based on Lansdowne 1202, offers another point of comparison. He identifies the manuscript as a mid-eighteenth-century French witness and notes affinities with the Grimorium Verum. Affinity does not mean complete identity or direct copying. ^[5]

These details allow us to assemble a case file, not to draw every arrow in a genealogy ourselves. Dating a witness and placing it within a related group are only the beginning of the analysis.

What to record before comparing

For each witness, record its location, language, dating and the unit selected. Record whether you are working with an image, transcription or translation too. Two Spanish texts may look identical because the same modern translation has regularised them; that correspondence does not establish that their historical sources are identical.

Keep observations in one column and interpretation in another. In the first, write what is visible: ‘the third item in the list is missing’. In the second, write the proposal: ‘this could be a shared omission’. Mixing the columns makes it difficult to revise the interpretation without changing your record of the evidence.

Do not silently correct forms that seem wrong. The anomaly may be precisely the clue you need. Keep a working transcription and, if you introduce a normalised reading to facilitate comparison, identify it as such.

6. Examining material features and absences

The history of a copy

Provenance documents an object’s ownership history. Folgerpedia distinguishes external evidence, such as receipts or sale records, from internal marks, such as ownership inscriptions or bookplates. Both can help reconstruct that history. ^[6]

Material provenance is not the same as textual genealogy. When the identification is well supported, a book’s having belonged to someone establishes ownership. It does not automatically establish that the person wrote it, read it all or used its instructions.

Imagine a copy containing three names in different places. Before treating them as a succession of owners, investigate whether they mark ownership, dedications, references or copied annotations. A word that looks like a signature may serve another purpose. The hypothesis must fit the mark, rather than the mark being made to fit the story you want to tell.

A missing page has several possible explanations

If a figure is missing, first ask where it is missing: from the physical copy, the reproduction consulted, the transcription or the editor’s selection. These are different situations. An incomplete digitisation can create an apparent absence; a text-only transcription may omit images because of its format.

If the text refers to a figure that does not appear, record the reference and locate the exact point at which it should relate to the figure. Do not immediately insert a similar drawing. A shared theme may be insufficient to establish that this was the intended image.

A useful reconstruction explains which witness supplies the added element and why it is considered relevant. Concealing that operation may lead readers to attribute to the document a combination assembled by a modern editor. Reconstruction itself is not the problem; presenting it as direct observation is.

What to examine in a reproduction

Examine the continuity of numbering, the beginnings and ends of texts, catalogue notes and visible changes in layout. If you identify a discontinuity, describe its form first. Do not immediately equate it with physical loss: it may also result from irregular numbering or a selection of images.

A reproduction is sufficient for some questions and limited for others. The thickness of the writing material, the structure of gatherings or particular ink details may require specialist examination. A good record explains how far the resource used allowed you to go.

7. When a difference becomes evidence

Not all correspondences carry the same weight

Two texts may share a common formula without either depending directly on the other. The correspondence does little to distinguish relationships if it occurs in many witnesses. To assess its value, consider how likely it would be to occur independently.

An unusual combination of changes may be more informative. But ‘unusual’ should not mean ‘surprising to me’. Compare it with other witnesses and ask which copying processes might produce it. An omission between repeated words, for example, may arise more than once without direct dependence.

The Parvum lexicon stemmatologicum brings together specialist vocabulary for studying textual transmission. It is a tool for defining problems more precisely, not a machine that automatically turns similarity into genealogy. ^[8]

Worked example: a reference that no longer works

The following is a teaching example; it does not reproduce the pagination of a real grimoire. You have three witnesses, A, B and C. All contain a sentence referring to a diagram. You have checked that the visible pagination belongs to each copy rather than to the numbering in a digital viewer.

In A, the reference works. In B, it retains a number that no longer leads to the diagram. In C, the number matches its layout. B’s anomaly deserves attention because it appears to preserve an element from a different arrangement of the book.

The first step is to describe that anomaly without drawing an arrow yet. The second is to propose explanations: B may have received the text from A, from a copy resembling A, or from an earlier model in which the diagram was also on page 12. The third is to look for evidence that distinguishes these possibilities.

Your initial conclusion might be that B preserves a reference consistent with an arrangement like A’s. You have not yet identified the physical object used to produce B. Nor have you established that C descends from B: C may have corrected it, depended on another source, or belonged to a more complex history.

Adding a second clue

Now suppose A and B share an unlikely omission in a nearby passage. Together, the evidence becomes stronger, but you must examine whether the two observations are independent. If the omission can be explained by the same reorganisation that moved the diagram, you may have two effects of one change rather than two separate pieces of evidence.

Investigation means improving the explanation, not accumulating marks in its favour. Record which new observation strengthens the proposal and which would weaken it. A hypothesis that admits no grounds for revision is difficult to assess.

8. From textual relationships to a transmission hypothesis

Three conclusions of different strengths

‘These witnesses share material’, ‘they belong to a related branch’ and ‘B was copied directly from copy A’ do not mean the same thing. Each statement makes additional commitments. The last identifies a particular relationship between objects and requires much more than general similarity.

Chronology can rule out some proposals. An object produced later cannot physically have served as the model for another already completed before it. But arranging dates does not establish the relationship that remains possible: lost intermediaries may stand between surviving witnesses.

Furthermore, a late edition may preserve an early reading, and an early copy may contain an innovation. The object’s age does not uniformly determine the age of every textual element it transmits.

Combining models

A copy may use more than one model. Stemmatological terminology uses the concept of reticulation for relationships that cannot adequately be represented by a simple branching tree, including those associated with contamination between copies. ^[9]

In practice, this means comparing parts separately. A witness may agree with A in a list and with B in a figure. You need not invent a mixture whenever a difference appears, but neither should you force the whole document to descend from a single model when the evidence does not fit.

Imagine a modern notebook copying a prayer from one edition and a drawing from another. Text and figure have different antecedents within the same object. This example does not prove that the same happened in a historical manuscript; it shows why a single line of descent may be insufficient.

A diagram should show its uncertainty

If you draw relationships, explain what each line means. A line may represent proposed dependence, shared features or simply chronological order. Using the same line for all three creates confusion.

Identify hypothetical intermediaries as hypothetical. Do not assign them an invented library or exact date. A model makes a provisional explanation and its implications visible; it does not add documents to the archive.

9. Writing a conclusion that can be checked

From evidence to wording

A useful conclusion contains an object, an observation, an inference and a limit. You do not always need four sentences, but you do need all four functions. ‘The tradition is ancient’ identifies almost nothing. ‘The catalogue dates this manuscript to around 1709’ shows which document and which authority support the claim.

Choose your verb after identifying the evidence. ‘States’ attributes a claim to a source; ‘records’ describes what an entry documents; ‘suggests’ signals an inference; ‘demonstrates’ requires an explanation of why the relevant alternatives can be excluded.

Do not use ‘perhaps’ to shelter any statement. If there is no reason to propose it, the adverb does not turn it into a scholarly hypothesis. Caution should accompany evidence or a defined problem.

A model report

The following paragraph brings the work together and identifies how the source was consulted:

‘The textual reproduction of the Grimorium Verum title page in Joseph H. Peterson’s digital edition was consulted. The title page states 1517 and attributes a translation to Plaingière, described using a Jesuit designation. The institutional reference postdating 1517 prevents uncritical acceptance of that date for the title-page state examined. On another page, Peterson transcribes a bibliographical description adding 1817 in square brackets. This is recorded as bibliographical dating transmitted by the editor. The argument presented does not, by itself, determine the exact year of the printed book or the composition date of all its materials.’

The report does not attempt to exhaust the book’s history. It answers a particular question, sets out its basis and retains its limits. A longer report should add examination of the copy, precise page references and discussion of the proposed dates.

Referencing without implying a consultation you have not made

If you know a detail through a modern edition, cite that edition. You may mention the document it describes, but avoid wording that implies you examined it directly. The chain of consultation should be visible.

A bibliography does not improve simply by accumulating titles. Explain what each resource contributes: the text consulted, a description of the object, historical context or a methodological tool. A work you have not read should not appear as though it supported an argument developed in the lesson.

10. Recognising a grimoire’s structure

Understanding a grimoire takes more than identifying its title and date. You also need to recognise how it organises knowledge. A list of names, a prayer and a recipe require different readings, even when they share a page. Treating every passage as the same kind of statement obscures the functions of its parts and the relationships they construct.

Your first reading of a volume should produce a map. Do not try to resolve every unfamiliar word as you go. Identify section openings, changes of voice, titles, numbered series, figures and internal references. This map will let you return to a difficulty with an understanding of its place in the whole. An isolated sentence may seem enigmatic because its function has been lost, not because it conceals a secret.

Six forms of writing to distinguish

A preface introduces the book, defines its reader or claims authority. An exposition explains what the world to which the text refers would be like. A catalogue organises entities and attributes. A prescription states what someone should do. A narrative recounts an event. An annotation adds information at a time that may differ from that of the main writing.

These forms can overlap. A recipe may begin with a prestigious attribution, continue with instructions and end with a promise. To describe it, separate those functions within the passage. The attribution concerns who guarantees the knowledge; the instruction, how the action is presented; the promise, the result the text announces.

Recognising the form changes the analysis. For a promise, ask what effectiveness the document claims and under which conditions. For a narrative, ask who is telling it, from what position and with what access to the event. For a list, ask about its ordering principle and possible discontinuities. None of these questions can be replaced by ‘is the whole book true?’

This classification is a TMF reading tool. It does not seek to impose a universal theory of genres on every document. If a passage combines functions, describe them. If a category is unhelpful, explain its limitation. The map’s usefulness lies in the questions it enables you to ask about the particular document.

The structure announced and the structure preserved

The opening of the Grimorium Verum announces three parts: arrangements of characters and spirits, natural and supernatural secrets, and a key to the work. The historical text can be consulted in Peterson’s transcription, from its opening to ‘Commence la Clef de l’œuvre’. The announced structure is evidence you can compare with what the edition consulted actually preserves. ^[1]

Do not assume that a contents list or opening programme was realised exactly as stated. A volume may announce a missing part, combine sections of different origins, or preserve cross-references that no longer work. First record the proposed organisation; then examine how it is fulfilled. If there is a difference, describe its location and form before deciding its cause.

This procedure also prevents circular reading. If you believe the book must have three parts, you might force any change of subject to produce three divisions. It is better to record the actual markers and explain the division you propose from them. A modern editor may have added headings to aid reading; these must be distinguished from historical headings.

As a short exercise, create two columns headed ‘what it announces’ and ‘what I can locate’. In the first, record only explicit divisions. In the second, identify the beginning and end of each section found. An empty cell does not yet mean loss: there may be different terminology or an organisation you have not recognised.

A contrast: organisation by aphorisms

The Arbatel offers another way of organising knowledge. The Latin text and historical English translation brought together by Peterson contain numbered aphorisms and groups of seven. You can locate a unit by its number and study how it relates to those before it. Brevity does not make it a modern summary: each aphorism belongs to a sequence of prescriptions and statements. ^[10]

Comparing this organisation with a list of entities teaches you to adapt your reading. In a list, repeated fields and exceptions matter; in a series of aphorisms, conceptual progression and internal references matter. You need not decide which work is ‘more authentic’ to recognise that they construct different reading experiences.

A structural map should include a field for relationships: which section prepares another, which figure accompanies a passage, and which list is developed later. Describing order alone produces an inventory. Describing these connections begins to explain how the book guides its reader. The distinction will help when comparing an abridged edition with a longer one.

From a map to a research question

Imagine a fictional volume with a preface promising to teach ten subjects, eight surviving chapters, and a conclusion mentioning all ten. Several explanations are possible: loss, an unfinished project, irregular numbering or editorial selection. Your map does not decide which occurred, but it locates the discrepancy. The next step is to examine material completeness and related witnesses.

If the object is complete and other copies contain the same eight chapters, the hypothesis of loss from this particular copy becomes weaker. If leaves are missing and another copy preserves the remaining chapters, you have different evidence. The structural map thus becomes a research tool rather than a decorative contents list.

11. Names, voices and authority

A text can name Solomon, speak in the first person and address a ‘you’ without those three positions corresponding to identified historical individuals. Before attributing a statement, ask who speaks within the document and who produced the object. Textual voice, claimed author, copying hand and translator may differ.

The Grimorium Verum title page brings together several signs of authority: a Solomonic connection, a stated source language and a figure presented as translator. You have examined its chronological problem. Now ask what that presentation does within the book. By associating the contents with prestigious names and languages, it proposes a way for readers to receive them. ^[1]

That observation does not require knowledge of the private intentions of whoever designed the title page. You can describe the visible devices and their relationships. Claiming that a printer intended to deceive a particular audience would require investigation of production circumstances, intended readers and circulation. Distinguish the textual function you analyse from the psychological motivation you cannot directly observe.

Attribution and pseudepigraphy

An attribution associates a work with an author. Pseudepigraphy means presenting a writing under an authorial name that does not belong to its historical author. Applying the second term requires arguments establishing that discrepancy. While assembling them, ‘work attributed to’ accurately describes the documentary position without anticipating the whole investigation.

Do not confuse attribution with material forgery. A book genuinely produced in one period may attribute its contents to a figure from another. Its material form remains historical. Nor do all attributions work in the same way: they may appear in titles, prefaces, headings, later catalogues or readers’ comments. Record who makes the attribution and where.

An author’s name can organise a textual tradition broader than an individual work. Two books associated with Solomon may share a claim to authority while transmitting different contents. ‘It is Solomonic’ therefore does not replace a title, witness or passage. The category helps formulate a comparison; it does not establish identity.

The reader imagined by the book

When an instruction addresses a ‘you’, the text constructs an intended reader. That reader may be assigned assumed knowledge, obligations and abilities. Studying this figure lets you ask what reader the document expects. Do not automatically identify that figure with all historical buyers or owners. The intended reader and the actual reader are two different objects of investigation.

In a fictional example, a preface requires command of several languages, while annotations in one copy translate basic words. This could show a gap between the text’s demands and a reader’s needs. The hypothesis depends on the authorship and date of the notes. The example teaches you to connect a textual expectation with evidence of reception.

You can also observe how access to knowledge is distributed. Some sentences promise clarity; others withhold an explanation or refer to later instructions. Before treating that withholding as evidence of a secret oral lineage, check whether it functions as a textual device, an internal reference or a consequence of incomplete transmission. Silence permits several explanations.

Authority within a hierarchy

Lists of spirits may assign titles, subordinate relationships, territories or functions. In a historical reading, you study how the text organises that authority. You are not compiling a census of verifiable beings. The wording ‘the text presents X as superior to Y’ preserves this distinction and allows changes between witnesses to be compared.

The opening of the Grimorium Verum lists Lucifer, Béelzébuth and Astaroth as three powers. Later come subordinate relationships and differentiated attributes. The contents offer a textual system you can describe, but you should not complete it with details from other books without identifying the addition. A figure with the same name may occupy a different position in another source. ^[1]

The temptation to unify is strong when preparing encyclopaedic entries. You want one answer about rank, function and image. Yet a rigorous reference library can record different answers with their references. ‘In this witness’ does not diminish the entry: it allows the history of differences to be reconstructed instead of erased by a uniform synthesis.

Writing a record of attributed authority

A useful record includes the form of the name, witness, precise reference and recorded attribute. Add a column for variants and another for interpretative notes. If a translation regularises the name, retain both the form you read and the form it uses for identification. Do not combine a historical spelling with a rank taken from a different edition.

For example, a fictional record might read ‘Asteron, treatise A, folio 12r, called a messenger’. Another might read ‘Asteron, treatise B, page 30, called a ruler’. Comparison starts with that difference. Before explaining a promotion in rank, you need to establish that the two mentions are historically related rather than merely sharing a name.

This section’s final question is specific: who guarantees what, to which reader, and in which document? If you can answer it, you will have separated several voices often conflated in the phrase ‘the grimoire says’. That separation prepares you to read translations and commentary without attributing them to the same speaker.

12. Reading lists and vocabulary without inventing equivalences

A list seems transparent because its items are separated. Yet its organisation may contain information lost when you copy only the names. Order, categories, numbers and exceptions belong to the document. Record the list as a whole before extracting individual entries.

Imagine a fictional series assigning a name, office and number to each entry. One entry lacks an office and another has two numbers. There may be a copying error, an intentional difference or a layout problem. Normalising every row to fit a modern table conceals precisely what you need to investigate.

What translating a term means

Translation means choosing how to express a meaning in another language. Sometimes the correspondence is relatively stable; sometimes there are several meanings or historical uses. ‘Spirit’, ‘character’, ‘art’ and ‘secret’ should not automatically receive their everyday contemporary meanings. Establish the meaning from the context and the form of the text consulted.

In texts from this corpus, ‘character’ may designate a graphic sign rather than a psychological disposition. That contextual observation does not, however, justify translating every occurrence identically. Ask which verb accompanies it, which object it indicates, and whether a related figure exists. Local context may resolve an ambiguity left open by a dictionary equivalent.

Keep three records separate in your notes: original form, translation used and working explanation. If you do not know the language, attribute the translation to its author. You can analyse how that version organises meaning, but should not present an unsupported linguistic correction of your own. Acknowledging dependence on a translation makes the reading more precise.

The danger of appealing etymologies

Similarity of sound does not establish an etymological relationship or historical identity. Nor is it enough to divide a name into fragments resembling familiar words. Etymological proposals require documented forms, chronology and relevant linguistic rules. In an introduction, you can record an attributed interpretation without making it the definitive explanation of the name.

This care is particularly necessary when a name circulates in several languages. It may undergo errors, adaptations, partial translations or reinterpretations. Two very similar spellings may be related; two very different forms may also be linked by a documented chain. Appearance alone decides neither possibility.

A reference library should make variants findable without erasing their sources. Use a main entry as a navigational aid and preserve the attested forms beneath it with their precise references. The main entry is a modern editorial decision. It should not imply that every document uses that same form.

Numbers and the promise of a complete catalogue

When a text announces a number of entities, compare that announcement with the enumeration. There may be repetitions, omissions or grouped entries. Count by an explicit criterion: distinct names, numbered positions or descriptive units. Results may differ without either necessarily being wrong; they count different units.

Suppose a fictional catalogue announces twelve figures and presents twelve entries, one containing two names joined by ‘or’. You can count twelve positions and thirteen name forms. Do not conclude that the author made a mistake before deciding whether that entry offers two entities or two designations. Grammar and context should guide the count.

Order deserves attention too. A series may be organised by rank, time, space or simply an inherited sequence. Do not assign a symbolic principle to every number without support from the document or research. Sometimes the more relevant explanation lies in copying or page layout.

A table that preserves uncertainty

Build your table with fields that can remain open. Include ‘not stated’, ‘uncertain reading’ and ‘reference pending’ as different states. ‘Not stated’ describes an absence in the passage; ‘uncertain reading’ describes a reading problem; ‘pending’ describes the state of your work. Replacing all three with a dash deprives readers of that distinction.

To practise, take five entries from a list and transcribe only their structure: which fields appear, which repeat and which are absent. You need not copy practical instructions. Then explain what information would be lost by turning those entries into an alphabetical list of names. You will have begun to read the catalogue’s structure as well as its vocabulary.

13. Images, diagrams and the written page

A figure is not a dispensable addition to the text. It may organise relationships, identify an entry or serve a function within a prescription. Its meaning depends on the page and the witness. Extracting it, cleaning it up and placing it in a modern record changes its context, even if its shape seems preserved.

Begin with a description someone else can check. State its location, relative size, relationship to the text and visible elements. Distinguish letters, lines, figures and annotations. You need not interpret every stroke to record its presence. The description should leave room to revise a later identification.

Describe before identifying

For a circular figure, ‘a circle with three groups of letters and a central stroke’ may be an initial observation. ‘The original seal of a particular entity’ adds identification, function and historical priority. You may eventually substantiate some of those claims, but each requires a different argument. Visual familiarity does not replace a precise documentary reference.

A reader may recognise an image from a modern edition. That edition may have reconstructed it, rotated it or combined it with another. Reverse the direction of your search: start with the object consulted and document its features, then compare. Do not force the witness to match the version you remember.

The relationship between figure and text may also be ambiguous. An image near a name does not necessarily belong to that name if the layout has changed. Look for headings, cross-references, continuity in the series and other copies. Spatial proximity is evidence to read alongside the document’s organisation.

What redrawing does

Redrawing turns observed strokes into new graphic decisions. It straightens lines, regularises curves and resolves joins that may have been uncertain. This can help explain a structure, but should not silently replace a reproduction of the witness. Identify it as a diagram or redrawing and explain what has been simplified.

In a fictional example, two strokes barely touch in the reproduction. Joining them in a clean drawing may create a letter or figure that the original does not clearly show. Separating them makes another decision. Using design tools is not the problem; concealing that the new clarity results from interpretation is.

The same applies to colour. A monochrome document may receive modern colours to distinguish layers or relationships. This aids learning if a key explains the intervention. Without one, readers may attribute the colours to the source. A good visualisation improves understanding without implying that more has been observed.

A figure supplied from another source

The digital edition of the Grimorium Verum identifies places where figures are missing from the printed book used and presents material from other witnesses. Its notes allow this editorial intervention to be studied. When citing an inserted image, retain the stated provenance rather than attributing everything visible on the digital page to the same printed copy. ^[1]

This case shows why a web page is sometimes a space for comparison. It contains historical text, corrections, images from different sources and editorial commentary. Appearing on one screen does not make it a facsimile of a single document. A unified interface may bring together several documentary units.

The edition of the Lemegeton likewise provides an appendix with drawings identified by their different manuscripts. Making these origins visible is precisely what makes it useful. If you gather those images into a record, retain each attribution and avoid presenting the selection as a historical page that never existed. ^[11]

Orientation, scale and cropping

Before comparing two images, check their orientation. A reversal may result from reproduction, printing or editing; do not assume a symbolic transformation. Scale can also mislead: a small stroke enlarged may appear more emphatic than another shown at its original size. Compare features and relationships, not just visual impressions.

Cropping removes contextual information. An isolated figure may lose the heading identifying it, a note questioning its reading, or the edge showing its position. Keep a reference to the complete page and describe the crop used. In an educational library, readers should be able to connect a detail with its context without guessing.

Resolution limits what you can claim. A small image may show the general arrangement but cannot determine whether a stroke is ink or a flaw in the writing material. Record that limitation rather than treating a blurred enlargement as evidence. Increasing pixel dimensions does not necessarily produce new information about the object.

An image reference record

Your record should include the source, precise reference, description, textual relationship and editorial intervention. Add reuse conditions if you intend to publish the image. Being able to view it online does not mean you have all the rights to reproduce it. If you do not include it, a stable link and a description can support the exercise without fabricating a facsimile.

An accessible description should convey the information relevant to the task. ‘Image of a seal’ is usually insufficient. ‘Modern diagram with three columns distinguishing document, edition and reading’ explains its function. Do not make alternative text a more certain historical interpretation than the visible explanation.

To check your learning, choose a figure from the resource you are studying and write two paragraphs. The first should describe what is visible without identifying its meaning. The second should propose an interpretation and cite its supporting evidence. If you cannot separate them, review which words introduced conclusions into the description.

14. Historical passages with commentary

The following commentaries show how to move from a textual unit to a bounded interpretation. They focus on functions, classifications and devices of authority. They do not reproduce instructions for performing practices. The references let you return to the historical text and check the starting point for each reading.

First reading: a promise of organisation

In the opening of the Grimorium Verum, the book divides its contents into parts and announces knowledge the reader will encounter later. The passage combines an introduction to the volume with a promise of access. You can read it as a textual contract: it explains what is offered and creates expectations of the whole’s usefulness. ‘Textual contract’ here is an analytical tool, not a legal document. ^[1]

To comment on it, perform three operations. Identify the subjects announced, examine the verbs connecting readers with those subjects, and then locate their development. The first two describe the programme; the third checks its realisation. A reading that merely repeats the topics does not explain how the book guides the experience of consulting it.

The promise of knowledge may be broader than the surviving contents. That difference does not justify inventing lost chapters. It lets you formulate a question about composition and transmission. To answer it, relate the programme to the copy’s completeness and to other versions. Interpretation of the preface should engage with the material description.

A brief, defensible commentary might read: ‘The opening presents the volume as an organised whole and anticipates the subjects of its parts. This announcement guides reading, but does not replace examination of the structure preserved.’ You have interpreted a function without claiming who believed the promise or how many readers followed it.

Second reading: categories of the secret

Aphorism 22 of the Arbatel defines the secret within a religious account of knowledge. Aphorism 24 organises secrets into groups and assigns them different scopes. You can consult both numbers in the Latin text and Robert Turner’s historical translation. The question here is how the text classifies knowledge, rather than whether its promises should be accepted as facts. ^[10]

Start by asking which conditions make something ‘secret’ within the text. Do not impose the everyday definition of information deliberately hidden. Then ask who can reveal it and which uses are considered legitimate. These relationships show a theory of access to knowledge, even if you do not share its premises.

The classification brings together subjects that a contemporary reader might place in different categories. Instead of correcting the document to fit a modern library, describe its own organisation. You can then compare that organisation with yours, explicitly recognising two classification systems. Anachronism arises when you make today’s system the only imaginable one.

A useful teaching explanation retains that distance: ‘The passage presents certain kinds of knowledge as dependent on revelation and orders them by importance. This classification tells us about the text’s criteria.’ It does not claim that this knowledge was attainable or that every reader understood the classification in the same way.

Third reading: the good of the reader and their neighbour

The Arbatel’s introduction relates knowledge to divine glory and to benefit for oneself and one’s neighbour. Aphorism 23 places the purposes of knowledge within that framework again. These formulations belong to the historical text and allow us to examine how it legitimises its project. ^[10]

Do not automatically dismiss these expressions as decoration. Ask what limits they set, what reader they construct and how they relate to the contents that follow. Nor should you assume they demonstrate benevolent conduct by every reader. A norm expresses what the document proposes; historical behaviour requires its own investigation.

Comparison with another work should focus on a particular point. You can ask how each text justifies seeking knowledge, or what role it gives to benefiting other people. You need not oppose imaginary blocks of ‘good magic’ and ‘bad magic’. Such labels may belong to historical controversies and should be attributed when they occur.

A good reference-library question would be: ‘How does this passage define the legitimate use of knowledge?’ A less useful one would be: ‘Is this book positive or negative?’ The first leads to locatable expressions and bounded comparisons. The second tends to replace analysis with an overall judgement whose scale you have not explained.

Fourth reading: classifying figures

Return to the Grimorium Verum’s list of powers and subordinate relationships. Describe which relationships appear: superiority, dependence, distribution of functions or association with places. You do not need to draw a complete hierarchy to observe these devices. A small group of entries can show how a textual organisation is constructed. ^[1]

Then distinguish name from position. If another witness preserves the name but changes its place, you have two variables. Comparison may reveal changes in classification, but does not explain them by itself. They could result from reorganisation, error, selection or interpretation. Choosing an explanation requires more than a neatly ordered final table.

Your commentary should mention exceptions too. An entry lacking the attribute repeated by the others may contain an omission or a significant difference. Do not complete it from a modern record before studying the problem. Exceptions belong to the evidence and may be more informative than regularity.

Writing your own commentary on a passage

Use four steps. Locate the unit within the work, describe a particular feature, explain its function and delimit the conclusion. Avoid beginning with a summary of the entire history of esotericism. A commentary gains depth by showing how a passage works, not by introducing many names unrelated to the question.

A 300-word commentary may require considerable work. It should select evidence and reason from it, rather than compress a catalogue of claims. If context is needed, include only what changes the interpretation. A date or variant is relevant when it helps explain the problem under examination.

Check that your verbs attribute the ideas: ‘defines’, ‘presents’, ‘organises’, ‘promises’, ‘classifies’. Make the shift visible when moving to your own conclusion. Separating content from interpretation need not make the writing cumbersome. Readers simply need to recognise who maintains each point.

15. From the book to its readers and society

A grimoire exists within social relationships: someone produces, acquires, preserves, changes or describes it. Investigating those relationships broadens the reading of its contents. But you cannot infer an entire community from a title page or an annotation. Social history requires documents to be connected with people and circumstances through specific evidence.

Begin by separating production, circulation, possession and use. A printer participates in production; a seller, in possible circulation; an inscription may indicate possession; a note may provide information about reading. One person may fulfil several roles, but you should not combine them without evidence. ‘Owner’ does not automatically mean ‘author’ or ‘practitioner’.

Manuscript and print as forms of circulation

The relationship between manuscript and print is not a simple linear replacement. You can investigate handwritten copies of printed material and editions reorganising manuscript material. Davies’s study listed in the bibliography provides context for these interactions. Do not treat medium as an automatic scale of antiquity, authenticity or proximity to practice. ^[7]

A manuscript may be a study copy, a personal selection or a commissioned object. A printed book may preserve important variants or receive unique annotations. The useful question is which operation produced that witness and what evidence supports its description. A handmade appearance does not guarantee privileged access to a lost origin.

The availability of materials affects what can be copied, although the particular influence must be investigated. A compiler with several books could select among them; someone who knew only a translation would have a different repertoire. These are possibilities to guide research, not biographies you can assign to an unknown copyist.

Reading marks and their limits

An underlined word establishes an intervention on a page, provided the image allows it to be recognised. Interpreting it as a reader’s emphasis requires consideration of its relationship to the text. It might result from editing, later classification or bibliographical review. The mark alone does not reveal its maker’s emotions or beliefs.

Several marks may form a pattern. If a reader systematically corrects dates or translates terms, you can describe those activities. Even then, avoid turning a small sample into a complete portrait of a person. The history of reading relies on documented actions and recognises dimensions that remain beyond reach.

A fictional example helps distinguish levels of inference. Ten marginal translations allow you to state that someone added linguistic equivalents. If the hand is identified, you can attribute them more precisely. They do not allow you to conclude that the person performed the rituals or belonged to an order. Evidence of a skill does not automatically establish an affiliation.

Price, demand and actual reading

A sale advertisement may document how a book was offered. It does not establish how many copies sold. A purchase record documents a transaction; it does not necessarily indicate how the volume was read. A print-run figure describes production, whereas estimating readership requires further assumptions. Preserve these distinctions when interpreting commercial information.

Advertising is also historical content. Its promises can be studied as presentation strategies without accepting them literally. An advertisement describing a book as ‘exceedingly rare’ documents that description, not a count of surviving copies. Assessing rarity requires a defined corpus and relevant bibliographical information.

Do not make a high price automatic evidence of ritual use. Collecting, age, physical condition and other circumstances may play a part. A social explanation must compare alternatives and rest on evidence. Even documented intensity of demand does not, by itself, reveal every buyer’s motivations.

Accusation, prohibition and description

If you find a document condemning a book or practice, identify its genre and authority. A condemnation provides information about a judgement and an act of classification. Reconstructing what it condemns requires examining how it obtained information, what it selected and what purpose it served. Hostility does not make the witness useless, but it shapes how it must be read.

Avoid the opposite leap too: condemnation does not establish that the object possessed the powers attributed to it. It may document fears, conflicts, institutional decisions or forms of control. Historical explanation should ask questions those sources can answer. Claims of effectiveness still require their own kind of evidence.

In a fictional case file, an authority prohibits possession of ‘books of secrets’ without listing them. You cannot automatically apply that prohibition to every title surviving in the region. You need to establish how the category was understood and which objects it included. The administrative category may not match your library’s category.

Comparing without universalising

This lesson’s corpus focuses on particular European texts and the forms through which they reach us. Its findings do not, by themselves, describe all magical or religious traditions. Comparison with other contexts requires appropriate sources, languages and specialists. A library’s breadth grows by adding documentation, not by extending a local explanation to the whole world.

A formal similarity, such as the use of signs or sacred names, can begin a comparison. It does not establish historical contact or a shared universal tradition. First state which feature you are comparing and its function in each context. Only then investigate transmission relationships where the evidence allows.

This discipline also protects living communities from being reduced to categories borrowed from other archives. Their own terms and explanations form part of the necessary documentation. In an introduction, a responsible limit means acknowledging which corpus you study and what additional knowledge a broader comparison would require.

16. Reference library: concepts and reading paths

This section lets you return to a particular difficulty after completing your first reading. Concepts are explained through their function in documentary work and refer back to the sections that develop them. They do not replace the cases: a definition helps you recognise a problem, but you learn the reasoning by solving it.

Object, work and transmission

Work. An intellectual or textual unit recognised across different manifestations. Its boundaries may themselves require investigation when compilations and substantially different versions exist. Use the term to frame the continuity you are studying, without assuming that every book with a similar title transmits exactly the same contents. Return to section 2.

Witness. A document preserving a form of the text and available for comparison. The term emphasises its relationship to a textual tradition. It does not mean it is an eyewitness to the events narrated. A late witness may preserve important readings, so date does not automatically determine comparative value. Return to sections 5 and 8.

Copy. An individual object that can be identified and located. Two copies of an edition may differ in completeness, annotations or ownership history. When your claim depends on a particular mark, cite the copy. Naming only the edition may lead readers to attribute that mark to every copy. Return to sections 2 and 6.

Compilation. A collection bringing together material from different units or sources. Recognising a compilation does not mean you already know its sources. Examine openings, titles, changes of subject and relationships between parts. A binding may bring together texts that do not share an author or composition date. Return to sections 4 and 10.

Variant. A difference between compared forms of a text or graphic element. It may concern words, order, omissions or layout. Its importance depends on the question: a small difference may be decisive for a reading but unhelpful in establishing relationships. Describe the difference before assigning a cause. Return to section 7.

Hypothetical intermediary. A witness proposed to explain relationships between surviving documents. It is not a discovered object. Its usefulness depends on making the evidence intelligible and remaining open to revision. A diagram must identify it as hypothetical, without an invented library or shelfmark. Return to section 8.

Production and consultation

Provenance. An object’s documented history of ownership or custody. It can be reconstructed from internal marks and external documents. It is not synonymous with textual origin, authorship or place of composition. A library may hold a manuscript produced far away and acquired long after it was written. Return to sections 6 and 15.

Shelfmark. An institution’s identifier for locating an object within its collection. It forms part of the reference and should be copied carefully. Do not confuse it with a page number or a digital image identifier. If the institution records changes of shelfmark, retain that information when connecting older references. Return to section 2.

Facsimile. A reproduction intended to represent a document visually. Establish its scope: it may involve cropping, changes of scale or omitted pages. It is not a transcription, and it does not make digital consultation a complete physical examination. Check which elements it reproduces before using it to describe the object. Return to sections 6 and 13.

Transcription. A written reading of a document’s signs. Its principles may preserve or normalise spelling, expand abbreviations and mark uncertainty. A transcription facilitates searching but incorporates decisions. Before assessing a variant, consult those principles and check whether differences might originate in the edition. Return to sections 2 and 12.

Precise reference. Information locating a particular passage or element: page, folio, section or aphorism number. A link to a lengthy work is not always enough. Distinguish the viewer’s displayed number from the document’s numbering. Where they differ, record both and explain their correspondence. Return to sections 6 and 14.

Interpretation and writing

Attribution. A relationship a source establishes between content and a person, authority or origin. State who makes the attribution and where. It may be historical or modern, secure or disputed. Recording an attribution does not mean accepting it as established authorship. This distinction is essential when a book invokes prestigious figures. Return to section 11.

Prescription. A statement indicating what should be done. It documents a proposed model of conduct, not automatically its performance. Studying actual actions requires other evidence or contextualised indications. Nor does an instruction’s stated aim establish that it was achieved. Return to sections 1, 10 and 15.

Paratext. Elements introducing a work and guiding access to it, such as the title page, preface or certain headings. This is an analytical category, not an unimportant area. It may contain attributions, announcements and editorial decisions shaping reading. Identify its origin before assuming it belongs to the same period as the textual body. Return to sections 3 and 10.

Reception. Ways in which a work, figure or idea is read and used subsequently. An edition, annotation or reinterpretation may document it. Continuity of a name does not imply continuity of all its functions. To study changes, follow each element through specific references. Return to sections 11 and 15.

Reconstruction. A proposal assembling evidence to represent a state not observed in its entirety. It may be textual, graphic or historical. It should explain what it adds, where the addition comes from and what uncertainty remains. A rigorous reconstruction is useful precisely because it allows its components to be distinguished from direct observation. Return to sections 6 and 13.

Reading paths for your question

If you have just found a PDF, start with sections 2 and 6: identify the resource and check its completeness. Then use the map in section 10. This order prevents hours spent interpreting a modern selection as though it were a complete reproduction of a historical copy.

If you are investigating a figure or name, combine sections 11 and 12. Record attributes by witness before creating a general entry. Add section 13 if images are involved. The resulting record should distinguish the name’s history, textual classification and iconography rather than merging them into an identity without dates.

If your problem concerns a date, return to the cases in sections 3 and 4. Then review your conclusion using section 9. Finding an alternative year does not finish the task: state where it comes from and which part of the object or text it is intended to date.

If you are preparing a comparison, use sections 5, 7 and 8. First formulate a focused question and record observations separately from interpretations. When the diagram looks too simple, check whether you have confused chronological order with dependence or dismissed intermediaries without argument.

17. Integrated case file: from discovery to library entry

This case file is fictional and designed to combine what you have learnt. It does not correspond to a discovered book or a real collection. You have a reproduction, a catalogue and an editorial note. Prepare an entry for an educational library without turning uncertainty into an invented history.

The working materials

The reproduction contains 42 images. Its title page reads ‘Secrets of the Ancient Sage’ and states the year 1600. The viewer numbers images from 1 to 42; printed pages run from 1 to 36, alongside covers and unnumbered leaves. Page 9 announces a figure ‘at the end of the book’. You cannot find it among the available images.

The catalogue of the Biblioteca del Puerto, an institution invented for this exercise, identifies the object as BP 214. It proposes a date between 1820 and 1840 and records 38 printed pages. A note states that the title attributes the contents to an ancient sage without identifying an earlier source. This extract does not explain the basis of the catalogue’s dating.

A modern edition bearing the same title includes a drawing at the end. Its editor states that it was taken from another copy, BP 319. No comparison of the two copies is provided here. The edition regularises names and alphabetically reorders a list that is numbered in the reproduction.

The image of the endpaper shows a note: ‘De Mateo, 1891’ (literally, ‘of/from Mateo, 1891’). The handwriting has not been identified. An annotation beside page 12 reads ‘see the other book’, without a title or reference. There are no purchase records, correspondence or further details about Mateo.

Your task and the decisions required

Prepare a 250–350-word entry identifying the object and the resource consulted. Retain both dates and their functions, explain the discrepancy in extent, and distinguish the figure supplied by the modern edition. Add a note on the annotations without turning them into a biography of the owner.

Then write three research questions in priority order. The first should resolve a difficulty directly affecting your description. The second should improve a textual comparison. The third may broaden the history of reception. Explain why this order enables progress without unnecessary assumptions.

Before writing, decide what ‘the same title’ means. It does not suffice to establish that BP 214 and BP 319 are copies of the same edition. They may be different versions. Your entry should describe the known relationship and leave the unverified relationship open.

18. Research workshop

Complete the tasks before moving to the answers. Keep an uncorrected first response: comparing it with your revised version will show which part of the reasoning you have learnt to control.

Task 1. Correct an overconfident record

A student’s record reads: ‘Grimorium Verum. Written by Solomon and translated directly from Hebrew by Plaingière. Published in Memphis in 1517. Since the Jesuits came later, we know it was printed in exactly 1817.’

Use the information in sections 2 and 3. Divide the record into claims. For each, identify its possible source and decide whether it is documented, attributed or insufficiently supported. Then write a new record of 120–180 words.

Your answer must retain the stated date, identify the origin of the alternative year, and avoid presenting authorship, source language or printing location as established. Replacing every verb with ‘might’ is not enough.

Task 2. Compare without inventing a chain

Work with the following fictional case file. The years are exercise data, not dates assigned to historical objects. The ‘phrase omission’ means a shared omission of a defined passage; it is not assumed to be a rare error.

Write two explanations consistent with the whole set. In one, C may depend on B; the other must include an intermediary or common source. Explain which observations both hypotheses account for, and propose two checks that would help distinguish them.

Then assess this statement: ‘C was physically copied from B because they are identical.’ Distinguish similarity, textual relationship and use of a particular copy. The task is not to draw the most appealing tree, but to justify which parts of it remain unknown.

Task 3. An absence in the reproduction

You receive a PDF containing twenty images. The text in image 8 refers to a figure you cannot find. Printed numbering jumps from 10 to 13, while the viewer moves from image 8 to image 9 without interruption. A catalogue note describes twenty-two pages.

Write a six-step checking plan. Decide what to compare before claiming that the physical copy has lost a leaf. Also explain why inserting a drawing from another book with a similar title does not solve the documentary problem.

This exercise permits several answers. Assessment concerns your distinction between digital sequence, book pagination and bibliographical description. Inventing a precise cause for the jump does not earn credit.

Task 4. Apply the method to a new source

Choose an institutional record for a manuscript or early book. The object need not concern magic. Prepare a record giving the institution, reference, title, dating, resource consulted and one interpretative difficulty. Include the stable link and access date.

Write three sentences: an observation or detail from the record, a bounded inference, and a question still open. State which additional resource you would need to answer it. Applying the method to a different object checks that you have learnt a procedure rather than simply an answer about 1517.

19. Answers and explanations

20. Review your work and retain what you have learnt

Review criteria

Do not use this table to calculate a mechanical score. Use it to identify the next revision. If your identification is inadequate, improve the record first: no amount of elegant reasoning compensates for a source that cannot be located.

Recall without looking

The next day, answer from memory: why is a title-page date insufficient? How does a copy differ from an edition? What could an inherited page reference establish? Why is an older manuscript not automatically the original?

Then open the lesson and correct your answers. Mark the distinction you had forgotten. A week later, repeat the task with a new catalogue record. Changing the object requires you to retrieve the procedure rather than remember what an example looked like.

The central lesson

Reading a grimoire historically requires you to identify the document before treating it as an authority. Dates, names and images are evidence to be placed within a history of production, transmission and consultation. A rigorous conclusion shows how it moves from observation to interpretation and where it must stop.

21. Sources and further reading

Numbers in square brackets refer to this list. Sources were consulted on 5 October 2026. Shelfmarks are documentary identifiers needed to locate objects; they are not internal production codes.

Documents and editions consulted

[1] Joseph H. Peterson, editor. Grimorium Verum, French digital edition, 1999; stated update: 17 July 2025. This lesson uses the textual reproduction of the title page and the presentation of witnesses. The page distinguishes the text’s edition from its modern introduction. The link allows the passage to be checked without confusing a digital edition with the physical copy.

https://www.esotericarchives.com/solomon/gv.htm

[2] Joseph H. Peterson, editor and translator. Grimorium Verum, bilingual page, 1999. This lesson uses the title page and the bibliographical description transcribed at the end, corresponding to 8632.aa.5. The notation 1517 [1817] is cited through this transcription, not through direct inspection of the current British Library catalogue.

https://www.esotericarchives.com/solomon/gv_bi.htm

[3] Formula of the Institute (1540). Jesuit Sources Digital, Institute for Advanced Jesuit Studies, Boston College. Digital edition of the document associated with Regimini militantis Ecclesiae, dated 27 September 1540. It is used as an institutional chronological reference, not to date the Grimorium Verum on its own.

https://jesuitsourcesdigital.bc.edu/3df7cmk/

[4] Wellcome Collection. L’art Magique, MS.983, catalogue record; proposed date: c.1709. This lesson uses the identification, content description and note on MCCCIX. The record identifies S. A. J. Moorat’s catalogue as the source of its description. No image of the manuscript is reproduced here.

https://wellcomecollection.org/works/kc3yhd2a

[5] Joseph H. Peterson, editor and transcriber. Clavicules du Roi Salomon, Livre Troisieme, based on British Library, Lansdowne 1202. Digital edition, 2003, updated 26 November 2023. It is used to identify a related witness and the editorial mediation through which it is consulted.

https://esotericarchives.com/solomon/ksol3.htm

Research and methodological tools

[6] Folger Shakespeare Library. Provenance, Folgerpedia. Institutional resource on evidence of ownership and the description of provenance. It is used to distinguish an object’s history from relationships between texts.

https://folgerpedia.folger.edu/Provenance

[7] Owen Davies. ‘Print Grimoires and the Democratization of Learned Magic in the Later Early Modern Period: Bricolage Tradition and the Cross-Cultural Transmission of Knowledge’. Entangled Religions, 14.3, 2023. DOI: 10.46586/er.14.2023.10440. Research article available in the University of Hertfordshire repository. Further reading on circulation between manuscript, print and new cultural uses.

https://doi.org/10.46586/er.14.2023.10440

[8] Parvum lexicon stemmatologicum. Digital academic resource hosted by the University of Helsinki. An introduction to the vocabulary of textual transmission. Consulting it helps clarify terms; it does not replace the reasoning required in each case.

https://wiki.helsinki.fi/xwiki/bin/view/stemmatology/Parvum%20lexicon%20stemmatologicum/

[9] ‘Reticulation’, in Parvum lexicon stemmatologicum, University of Helsinki. An entry on relationships that a simple tree does not represent adequately. It is used to introduce the possibility of copying from multiple models.

https://wiki.helsinki.fi/xwiki/bin/view/stemmatology/Reticulation/

Other texts used to study content

[10] Arbatel de magia veterum, Basel, 1575; historical English translation by Robert Turner, London, 1655. Joseph H. Peterson’s digital edition includes the Latin text and historical translation. This lesson uses the opening material, the sevenfold organisation and aphorisms 22–24. Its commentary analyses the historical public-domain text; it does not reproduce the modern introduction or offer a new complete translation. The page explains its editorial normalisations.

https://www.esotericarchives.com/solomon/arbatel.htm

[11] Joseph H. Peterson, editor. Lemegeton Clavicula Salomonis, digital edition. This lesson uses the appendix of drawings and their attributions to different manuscripts to examine the distinction between an editorial page and a documentary unit. The images are not reproduced. This reference supports the description of that editorial arrangement, not direct inspection of all the manuscripts cited.

https://www.esotericarchives.com/solomon/lemegeton.htm

Scope of this lesson

Examples A, B, C and X, the cases without historical identification, the workshop’s fictional PDF and the Biblioteca del Puerto case file are teaching examples created by The Method Files. They do not describe newly discovered manuscripts or reproduce findings from original research. The visual diagrams are explanatory, not facsimiles. This lesson develops documentary and historical reading skills; it does not teach the performance of ritual practices.

Sources for this reading

1. Joseph H. Peterson, editor. Grimorium Verum, French digital edition, 1999; stated update: 17 July 2025. This lesson uses the textual reproduction of the title page and the presentation of witnesses. The page distinguishes the text’s edition from its modern introduction. The link allows the passage to be checked without confusing a digital edition with the physical copy. Consult the source
2. Joseph H. Peterson, editor and translator. Grimorium Verum, bilingual page, 1999. This lesson uses the title page and the bibliographical description transcribed at the end, corresponding to 8632.aa.5. The notation 1517 [1817] is cited through this transcription, not through direct inspection of the current British Library catalogue. Consult the source
3. Formula of the Institute (1540). Jesuit Sources Digital, Institute for Advanced Jesuit Studies, Boston College. Digital edition of the document associated with Regimini militantis Ecclesiae, dated 27 September 1540. It is used as an institutional chronological reference, not to date the Grimorium Verum on its own. Consult the source
4. Wellcome Collection. L’art Magique, MS.983, catalogue record; proposed date: c.1709. This lesson uses the identification, content description and note on MCCCIX. The record identifies S. A. J. Moorat’s catalogue as the source of its description. No image of the manuscript is reproduced here. Consult the source
5. Joseph H. Peterson, editor and transcriber. Clavicules du Roi Salomon, Livre Troisieme, based on British Library, Lansdowne 1202. Digital edition, 2003, updated 26 November 2023. It is used to identify a related witness and the editorial mediation through which it is consulted. Consult the source
6. Folger Shakespeare Library. Provenance, Folgerpedia. Institutional resource on evidence of ownership and the description of provenance. It is used to distinguish an object’s history from relationships between texts. Consult the source
7. Owen Davies. ‘Print Grimoires and the Democratization of Learned Magic in the Later Early Modern Period: Bricolage Tradition and the Cross-Cultural Transmission of Knowledge’. Entangled Religions, 14.3, 2023. DOI: 10.46586/er.14.2023.10440. Research article available in the University of Hertfordshire repository. Further reading on circulation between manuscript, print and new cultural uses. Consult the source
8. Parvum lexicon stemmatologicum. Digital academic resource hosted by the University of Helsinki. An introduction to the vocabulary of textual transmission. Consulting it helps clarify terms; it does not replace the reasoning required in each case. Consult the source
9. ‘Reticulation’, in Parvum lexicon stemmatologicum, University of Helsinki. An entry on relationships that a simple tree does not represent adequately. It is used to introduce the possibility of copying from multiple models. Consult the source
10. Arbatel de magia veterum, Basel, 1575; historical English translation by Robert Turner, London, 1655. Joseph H. Peterson’s digital edition includes the Latin text and historical translation. This lesson uses the opening material, the sevenfold organisation and aphorisms 22–24. Its commentary analyses the historical public-domain text; it does not reproduce the modern introduction or offer a new complete translation. The page explains its editorial normalisations. Consult the source
11. Joseph H. Peterson, editor. Lemegeton Clavicula Salomonis, digital edition. This lesson uses the appendix of drawings and their attributions to different manuscripts to examine the distinction between an editorial page and a documentary unit. The images are not reproduced. This reference supports the description of that editorial arrangement, not direct inspection of all the manuscripts cited. Consult the source