The 72 Spirits of Goetia

The seventy-two are usually presented as a horror catalogue, and the presentation distorts the source badly enough to be worth correcting before any practice is discussed. In the Ars Goetia they are organised into three ranks, nine Presidents, thirty-six Dukes and thirty Kings, and grouped into families of two, three or four. Each entry names a visible form, an office and specific powers. Read as a set, that is a bureaucratic structure with procedures and jurisdictions, not a list of torturers. The medieval and early modern material they descend from is likewise administrative, and the modern brand layered on top of it is a much newer invention than most users assume.

Three ranks and thirty-nine families

The catalogue divides cleanly. Nine Presidents act as the governing tier, thirty-six Dukes hold substantial administrative authority, and twenty-seven Kings hold territory and command subordinates. These are the classic numbers, which is worth stating because popular versions frequently miscount the Kings as thirty and the total as eighty, and a text that cannot count its own spirits has already told you something about its reliability.

The families matter more than the ranks for practical purposes, because a family shares an office area and often a method. Spirits that appear together in a single entry under one name are a related set, and the entries give their number of members, their day and night governing times, and their planetary attributions. Learning one family properly teaches you more about the material than memorising all seventy-two names, and the catalogue is explicitly organised for that kind of partial study.

Where the catalogue was assembled

The material descends from Jewish and Christian demonological sources, particularly the Pseudepigrapha Book of Enoch, arriving in Europe through Arabic and Jewish intermediary material, the seventeenth century Conjuratio Ben-Maimon, and the Lemuria. That is a long and traceable transmission, and it explains the catalogue's structure: it is a list of administrators drawn from several earlier lists, assembled for a specific practical purpose by people who were not writing literature.

The practical purpose was invocation with a bounded request. The classic service is a spirit called to perform one task under one office, with the caller responsible for the terms. The offices are strikingly domestic by modern standards, involving washing, cooking, guarding rooms, fetching lost things and answering questions about the past. The image of the spirit as a monster in a cell comes almost entirely from nineteenth and twentieth century fiction and popular religion rather than from the source.

The layer that arrived recently

Modern Goetia as a brand, the video essays, the phone apps, the grimoire commentary channel, is a twenty-first century phenomenon sitting on top of older material. The classical text is concise, dry, and largely uninterested in what the spirits feel like. The modern layer supplies drama, personality, aesthetics, and a great deal of invented biography for spirits whose original entries contain nothing of the kind.

This matters for anyone who wants to work with these names, because the invented layer is where the risk is concentrated. A system that encourages you to build a relationship with a fictional persona, to feel its presence, and to attribute outcomes to its instructions has removed every brake the original design provided. The classical arrangement was restrictive on purpose, and the popular version has removed the restrictions while keeping the names.

From Solomon's legend to a usable entry

The Solomon framing in the source material is a literary device, not a historical claim, and it is worth setting aside so that the entries can be read as what they are: descriptions of powers and jurisdictions attached to names. Read that way, an entry becomes usable. A spirit given as a keeper of records is asked to disclose something, and a spirit governing a territory is asked to move something within it.

Select three or four families rather than one spirit. Read their entries completely, including the clauses about what they will not do, which are usually stated more precisely than the powers. Then write a single office sentence in your own words for each, and check that each sentence contains a verb with an object. If a sentence has no verb, you have written a mood rather than a request, and the working will not have a shape.

Reading an office without inflating it

The single most common error is inflation: reading a narrow entry as a general licence. An entry that grants knowledge of past events is not thereby a spirit that improves your finances, and the practical consequence of treating it as one is that you charge a working, get nothing, and conclude that the spirit is weak rather than that you asked for the wrong thing.

The defence is to write the boundary before the power. State in one sentence what you are not asking for, and put that sentence in the working itself as an explicit withholding. This costs nothing and it does real work, because the commitment then has a shape you can later assess, and because a working with a boundary is much harder to talk yourself into recharging indefinitely.

A working, from call to closure

The classical sequence is short. Preparation, a defined place, the office stated, the spirit named and called for that single purpose, the request made plainly, and then a dismissal with thanks. Everything else in the genre is later ornamentation, and the shortness is not a simplification introduced by modern laziness but the form the material actually takes.

Where the ceremony sits is a question of approach. If you invoke, the working happens in your own act and the only honest sign is the intended effect. If you evocate and expect something visible, you have set yourself a test you will probably fail, because a strong expectation of a presence is a reliable way to produce a subjective impression of one. Most serious practitioners in this tradition use invocation and score themselves on the outcome.

Where it breaks, and why that is informative

A Goetic working fails in predictable ways. The office is too broad, the request has no deadline, the timing claims are treated as more important than the wording, or the working is charged repeatedly because nothing happened and nobody wrote down a prediction in advance. Every one of these is a design error, and noticing which one occurred is worth more than the experience of the charge.

The harder case is a working that produces a clear result on a matter you did not want to hear about. Some of the offices in this material concern loss, exposure and information you would rather not have, and the tradition's insistence on narrow offices is partly a recognition of that. Treat any unasked-for result as a signal to narrow the next office, not as a reason to expand the current one.

Materials, hours, and the price of a bad office

A sheet, a pen, a sigil or a plain written call, and a defined place with no interruptions. Two twenty-minute sittings for a single request is ample, and the tradition's own advice about the length of a Goetic operation is short enough that most modern elaborations contradict it. The apparatus has grown enormously while the source text stays the same, and the growth has not improved reliability.

The expensive mistakes are not material. They are charging a general request, charging repeatedly, and building a personal mythology around a spirit in order to make a result feel significant. All three are invisible at the time and all three make the practice less testable, which is the only thing it ever had going for it over simply talking to a person about the problem.

A record you would show a sceptic

For each working, four fields: the date, the office sentence, the observable outcome with a deadline, and what occurred. No impressions column, or if you keep one, keep it on a separate page that the scored record never references. The scored record is the one that would be evidence if you were arguing with someone, and it should be readable by someone who thinks the whole field is nonsense.

Review monthly and keep the misses. After a year you will have either a small number of results worth taking seriously or a stack of unfulfilled requests, and both of those are acceptable outcomes that most practitioners never reach because they never keep the record. The record is the only part of this work that improves with age.

Effects you can credit and effects you cannot

Separate three categories. External events that were predicted in advance and occurred on time, which are the only category that counts. Internal experience, including feelings of presence, pressure and reply, which is produced by concentrated attention in a quiet room and appears in control runs with the same intensity. And coincidences noticed because you were primed, which is a cognitive effect with a well-documented mechanism and no external content.

The first category will be small. That is not a failure of the practice but the actual yield, and a practitioner who has ten years of first-category results has something rare and can say exactly what it looks like. A practitioner with ten years of second and third category material has something much more common, which is a rich interior life organised around a set of names, and it is worth being clear which of the two you are being offered as a description of the work.

Restitution as the closing form

The classical texts close an operation with a request for restitution, which is a return or offering, sometimes a food, and always something given rather than received. The instruction is specific about the nature of the gift, which is another sign that the material was written by people thinking about procedure rather than atmosphere.

Take it seriously as a design element rather than as a formality, because it is the one part of the sequence that does something psychologically real. A working that ends with something given closes a loop, and the loop-closing is what keeps a request from feeling like an open transaction in your head. Whatever you give, choose something you can genuinely part with, and choose it before the working rather than selecting the leftovers afterwards.

The point at which to close the book

For most practitioners the useful stopping point is a working familiarity with three or four families and a written record of a handful of results. The rest of the catalogue is reference material, and treating all seventy-two as equally available produces a shelf of unmemorised names rather than any working competence. Depth on a few entries beats coverage of the list every time.

And the honest reason to be careful with this material in the first place is that the genre around it has made a great many claims on your behalf that you would not have made yourself. The names are old and the structure is documented. Everything added since, the personal relationships, the dark atmospheres, the sense of contact with a separate being, came later and came from people who were not transmitting anything. Reading the source is a way of taking the material back, and it is available to anyone with a library card and an afternoon.

Frequently Asked Questions

How many spirits are in the Ars Goetia?

Seventy-two, divided into three ranks: nine Presidents, thirty-six Dukes and twenty-seven Kings. Popular sources frequently give thirty Kings or claim eighty spirits, both of which are counting errors that have circulated for decades. The spirits are further grouped into families of two, three or four, and learning one family in depth is considerably more useful than holding all seventy-two names in memory.

Are the seventy-two demons really torturers of the innocent?

Not in the source material. The classical entries describe administrators and officials with defined jurisdictions, and the offices include washing, cooking, guarding rooms and answering questions about the past. The torturer image comes from later fiction and popular religion rather than from the texts that carry the names, which is part of why the media version of this material is so much harsher than the documents it claims to represent.

Which spirit should I work with for a specific problem?

Pick by office rather than by reputation. Read the entries for the families whose jurisdiction touches your problem, then choose the one whose powers are narrowest for what you want, because a narrow office gives you a testable outcome and a broad one does not. Write the office sentence before choosing the sigil, and if the sentence needs the word and more than once, you have picked too broadly.

Do I need a full Goetia apparatus to do this properly?

No. The classical operation is a defined place, a stated office, a call, a plain request and a dismissal with restitution, and it runs to about twenty minutes. Everything beyond that belongs to the later layers of commentary. The elaborate apparatus in modern practice adds variables you cannot isolate, and it does not make the request more likely to be answered.

How do I know a result is real and not just the power of suggestion?

Write the expected outcome and a deadline before the charge, keep no column for impressions, and treat feelings of presence as internal data rather than as confirmation. A real result is an external event you could not have predicted, dated in advance. If you find yourself able to say afterwards that you always expected it, you are looking at motivated reasoning rather than at a prediction, and the honest move is to record it as unclear.