The 72 Spirits of Goetia

Sixty or seventy-two rows of names, ranks, and offices fill a wall in most serious collections of grimoire material, and beginners treat the wall as a phone directory rather than a filing system. The distinction matters. A phone directory tells you who exists and how to reach them. A filing system tells you what shape of problem each entry is designed to address, which is the only part of the data that does work when you are stuck at two in the morning with a problem and no time to read three hundred pages. What follows is a way of reading that catalogue as a reference rather than a bestiary. It covers where the ranks came from, why the grouping into families is not the hierarchy people assume, what the original offices were actually for, and how to build a personal row and a working from a single entry without inflating what the entry claims.

What the catalogue is a catalogue of

The material presents a fixed number of named spirits, each assigned a rank, a visible form, an office, and a set of powers. That format is the reason the catalogue has survived four centuries of shifting fashion: it is a table. You can sort it, index it, and search it. Most other entity material in the Western tradition is narrative, which cannot be sorted and therefore cannot be used under pressure.

Treating it as a table changes the question from who is this to what is this row for. The names are the least informative column, because they are arbitrary labels chosen to be memorable rather than descriptive. The form column is descriptive of appearance only. The office column is the one that describes a function. Practically everything useful in the catalogue lives in that single column, which is why experienced users build a working from the office and ignore the rest.

The three ranks and what each one could do

The catalogue is organised into three tiers, and the tiers indicate administrative standing rather than destructive power. A President sits at the top of a house, a Duke manages a group of lesser spirits, and a King commands a family. The names read as frightening and the job descriptions read as bureaucratic, and the second fact is the important one.

Beginning users routinely misread the ranks as a threat hierarchy, which produces working with the wrong spirit selected entirely. Someone who wants help moving a decision forward has picked a name because the reputation was intimidating, and the office column turns out to have nothing to do with decisions. Reading the office first prevents this, and the office is the only column that will reliably tell you whether an entry suits the problem in front of you.

Families, and why the grouping is not a hierarchy

Entries appear in clusters of two to four, and the clustering is a grouping rather than a chain of command. The members of a cluster share a thematic area, and in some cases appear to be variants on a common description. A cluster is better read as a set of related jobs than as a family tree, and the person they answer to is not the person they resemble.

This matters for a practical reason. If you treat a cluster as a hierarchy you will try to work with the most senior member, which adds a step for no benefit. If you treat it as a set of related functions you will pick the member whose office matches your problem, and you will often find that a lower-ranked entry is a better fit than the one everyone names first. Nothing in the material obliges you to start at the top.

The layers the material was assembled from

The catalogue is a composite, and the seams are visible if you know where to look. Its oldest layer comes from Jewish and Christian demonological material, much of it traceable to a pseudepigraphic apocalyptic book that circulated widely in antiquity and was later dropped from most canonical collections. A seventeenth-century compilation of protective conjurations contributes the structured procedural frame. Medieval and early-modern cleansing literature contributes the ritual vocabulary.

None of these layers was written as a reference table. They were written as warning texts, protective instructions, and theological argument. The table format is a later editorial development that organised existing material into something searchable. That is a genuine convenience, but it also means the table's apparent authority is partly an artefact of formatting, and a row that looks authoritative because it sits in a neat column may rest on a source that was never making the claim the table implies.

Officials, not torturers: the context that was lost

In the original scheme these beings held administrative or disciplinary positions. They are described as having authority over a domain, powers within a limit, and a duty to be summoned. The lurid modern framing of sadistic tormentors comes from a later cultural layer that had little to do with the source material and everything to do with nineteenth and twentieth century entertainment.

Correcting this is not a matter of sentiment. If you read the office column as a job description, you get coherent requests. If you read it as a description of torture, you get incoherent ones, because a torturer's office has no boundary and no procedure, whereas an administrator's office has both. The source material is far more usable once you have put the entertainment layer back where it belongs.

The modern brand and what it added

There is now a substantial commercial layer built on the catalogue, consisting of apps, video essay series, and a commentary culture that treats the seventy-two as a fixed set of personalities. Some of it is informative and well researched. A good deal of it presents a specific row's description as its interpretation rather than its record, and the distinction is easy to lose because the presentation style is identical.

The commercial layer has one genuine contribution: it has made the material accessible to people who would never have opened a four-hundred-year-old compilation. It has also produced a great deal of confident detail with no source. If you use these sources, use them for orientation and go back to the older material for the actual row. A secondary source that does not say where it got something is decoration.

Choosing an entry by office rather than by reputation

The selection procedure is short. Write the problem as a single sentence using a verb that names a function rather than a wish. Then scan the office column for verbs. Ignore the name, ignore the powers, ignore the reputation. You are looking for the row whose office contains the verb you wrote, and if no row does, that is information about the aim rather than a reason to pick the closest famous name.

When a match exists, narrow the office to the smallest version of the function that still covers the problem. A general entry covering several domains is harder to work with than a narrow one, partly because the office is harder to hold in mind and partly because you will be tempted to keep widening it later. The narrow office that you can state in five words is the better working.

A personal sigil built from a Goetic row

Once you have a row, the working is short. Take the office, reduce it to a clause, cut that clause down to letters, and treat the resulting shape as your mark. You are not drawing the spirit and you are not reproducing anything from a plate. You are compressing a job description into a token you can hold attention on for ninety seconds, which is all the charge has ever been asked to do.

The reason to derive it rather than copy an existing design is that copying imports somebody else's office. A mark you cut from your own words is tied to your own phrasing, and when you need to remember what it was for you can read it back in one glance. Existing designs are fine for reference. Deriving your own is faster.

The reference sheet and how to build one

Build a single page, not a database. On it, for each entry you have actually looked at seriously, put four things: the name, the rank, the form in three words, and the office in one sentence. Stop there. Resist adding the powers, because they are the column most likely to be paraphrased beyond recognition in secondary sources and the column you will never use.

Ten rows is a usable sheet. Seventy is worse than ten because you will not consult it. The purpose of the sheet is to be readable in the ninety seconds before you start, and anything that takes longer to scan than the working itself is not earning its space.

Where the numbers get quoted wrong

The count itself is not stable across versions. You will encounter a number that reflects a different editorial decision about which entries to include, and you will encounter people who treat that variation as evidence of fraud. It is not. It is evidence that the table was assembled and revised by people over a long period, which is what any compilation of this kind looks like.

The number is also the least interesting thing about the catalogue. What matters is that the rows are structured enough to be used. If your working depends on the set being exactly some particular total, you have made the total load-bearing, and it was never carrying anything.

Reading an office without inflating it

The office descriptions use dramatic language. The discipline is to read them as scope statements rather than power claims. An office that describes authority over a domain means the entry is the appropriate one to work with for problems in that domain. It does not mean the entry can do anything within the domain, and it certainly does not mean the entry can do anything at all.

The practical test is whether you can state the office in your own words without adding anything. If your paraphrase contains a word that is not in the source, you have inflated it, and you should cut the word before you charge. Inflated offices produce working that tries to do too much, which produces disappointing results, which gets read as a charge problem rather than a drafting problem.

The row you will actually use

Practitioners who keep the material on a single page tend to use the same two or three rows repeatedly, and the rest becomes reference. This is not a failure of ambition. A row you have read closely, whose office you can state without embellishment, is worth more than twenty rows you have skimmed, because the value is in the quality of your relationship with the description rather than the breadth of the collection.

Pick the row whose office describes the thing you actually need this month. Work it properly for a month. If it worked, you have something better than a catalogue. If it did not, you have a specific and useful piece of information, which is what the reference was for.

Frequently Asked Questions

Is the Arcanum trustworthy enough to work with?

As a reference structure, yes. Each of the seventy-two entries assigns a rank, a visible form, an office, and a set of powers, which makes it the most structured entity catalogue in Western grimoire literature and therefore the most usable one. As a historical document it is a composite of several sources, and the confident tone of a formatted table should not be mistaken for the reliability of a single author.

Are these spirits real?

That question has several answers depending on what you want the catalogue for. The tradition holds that they are independent intelligences, and modern practice mostly treats them as pattern fragments useful for interrupting repetitive thought. The most useful position is that the entries describe functions rather than persons, and that using the described function is legitimate whichever view you hold.

Should I pick a famous spirit instead of a matching one?

No. Reputation in this material comes from later entertainment culture rather than from the source text, so a famous name is a poor predictor of a matching office. The row whose office contains the verb you need will serve you better, and you will find that narrow matches produce more coherent results than broad famous ones.

How many of these should I work with at once?

Fewer than you think. Each active row is a small ongoing commitment, and attention is the limited resource. One row worked properly for a month teaches you more than five rows opened and abandoned, because the skill being built is the drafting and holding of the office rather than the collection of names.

Do I need to memorise the lore?

No. The office is the column that does the work, and you can read it off a sheet at the moment of the working. Memorising the lore produces a feeling of mastery that is unrelated to function, and it is a common way to spend months of effort on the least useful part of the material.