The Angelical Keys and Liber Logaeth are related parts of Dee’s project, but they are not the same text in two formats. Their manuscripts, dates, structures, and translation evidence lead to different kinds of analysis.
Two manuscripts, two textual situations
Liber Logaeth is associated with Sloane MS 3189. The manuscript contains prose material and an extraordinary series of large letter tables: ninety-five grids of forty-nine by forty-nine cells, together with two empty tables and four still larger arrangements in the scholarly description. The physical dominance of the grids makes letters, positions, and extraction procedures central to any interpretation.
The Angelical Keys are preserved in Sloane MS 3191 as a section commonly described as forty-eight Keys. Eighteen are fixed invocations. A further form is applied successively to the thirty Aethyrs, producing the familiar later count of nineteen Calls. The manuscript also preserves interlineated English, which makes this section the main basis for proposed vocabulary and grammar.
These differences are not minor cataloguing details. A translated invocation invites comparison between clauses. A letter grid invites questions about direction, cell position, and method of extraction. Joining them in one undifferentiated “Enochian dictionary” removes the information that tells us what each string is and how it reached the page.
The earlier Logaeth material
The Logaeth sessions belong to the earlier phase of the project in 1583. The prose passages often have a different audible character from the later Keys: rhyme, alliteration, and pronounceable runs are more noticeable in several transcriptions. The tables, however, resist ordinary continuous reading. Their scale and density place pressure on every modern extraction method, because a single change of direction can produce a different sequence.
Instructions associated with the material speak of writing or reading right to left and from the back toward the front. The surviving manuscript itself is laid out in the ordinary direction. That mismatch does not license any one modern reversal automatically. It shows that “direction” may refer to a stage in reception, copying, or interpretation rather than the final page alone.
The tables also complicate claims about words. A string selected along a row, column, diagonal, or rule-based path is not equivalent to a prose form recorded as spoken language. Before an extracted group enters a lexicon, the method, table, coordinates, and manuscript reading must be stated. Otherwise a reader cannot distinguish a source form from a later construction.
The later Keys and their translations
The Keys were recorded in 1584 and copied into a ceremonial book whose sections also include Liber Scientiae and Heptarchia Mystica. Their language is often harder to pronounce by ordinary English habits: consonant clusters and long vowel sequences appear where the earlier prose can sound more fluid. That difference supports treating the two corpora separately before asking whether they share roots.
The interlineated English is indispensable and difficult. It is not consistently word for word. One Angelical form may sit beside several English words, and the English can expand a compact clause into ornate theological prose. The first two Calls show this repeatedly. They also supply genuine recurrence: commands such as zodacare and zodameranu can be compared in more than one location.
The count of forty-eight does not mean forty-eight independent modern “keys” of equal length. Eighteen fixed texts plus one form addressed to thirty Aethyrs explains why both forty-eight Keys and nineteen Calls appear in accurate descriptions. The count is structural evidence about the manuscript, not proof that every later ritual arrangement reproduces Dee’s book.
A worked comparison: what counts as translation evidence
Suppose a form occurs in the first Call beside an English clause and a similar form occurs in the second beside a related clause. That repetition supports a tentative lexical family. If a visually similar string is then extracted from a Logaeth grid, the resemblance is worth recording, but it does not carry the translated meaning automatically. The grid string may arise from a different segmentation or reading direction.
The same distinction applies to letter frequency. The Calls can be counted as continuous transcribed prose, although spelling variants affect totals. The tables encode a fixed rectangular space in which repetition may result from construction rules. Combining both into one frequency chart would compare unlike objects unless table coordinates and prose boundaries remain separate.
A strong comparative edition would therefore maintain at least four fields: manuscript and folio, date or session, textual form such as prose or grid, and the presence or absence of an accompanying translation. Our concordance deliberately begins with the first two translated Calls. The alphabet tool begins with the twenty-one-letter register. Neither tool claims to decode the Logaeth tables.
What the separation clarifies
Separating the corpora makes connections easier to evaluate. A proposed shared root can be tested against two dated settings. A spelling difference can be assigned to a witness rather than flattened into error. A modern extraction can be recognized as an operation performed on a table rather than confused with a phrase Dee recorded in sequence.
It also clarifies the larger archive. Sloane MS 3188 preserves early conference records and the 1583 alphabet moments. Sloane MS 3189 preserves Liber Logaeth. Sloane MS 3191 preserves the Keys and other distinct works. The three manuscripts document a developing project across several years, not one finished textbook.
Readers can continue with the Liber Logaeth manuscript guide, the Angelical Keys guide, or the forty-eight Keys and nineteen Calls analysis. Each route keeps the source object visible while connecting it to the broader sequence.