ARISTEIA

This is the least flattering piece in the set, written on purpose. Entity work is the corner of this material where the claims are weakest, the inherited language is most confident, and the least gets said out loud. The name in this piece has no catalogue entry, which makes it a convenient place to start asking what any of the names are for.

The Problem With Every Version of This

Here is the objection that survives every escape route people have found. You cannot distinguish an external figure from your own construction of one, because the evidence available to you is generated by you. Whatever you feel when you work with a name is produced by the system doing the feeling, and the system is the one that also produced the expectation.

This is not a new problem and it is not specific to magic. It is the standard difficulty in any practice where the observer is also the patient. People have tried to solve it in three main ways, and each has a cost worth naming.

The first approach is to add external verification: independent witnesses, other practitioners, physical artefacts that could not have been produced unconsciously. This raises the evidentiary bar substantially and, in practice, almost nobody clears it. The catalogue material is the only part of the field with a documentary record, and the documentary record is about medieval compilers rather than about what happens in a room in the present.

The second is to redefine the target as something internal. This is the most common move in modern practice and it is the one most often presented as a translation rather than a retreat. It does make the practice safe and it does make it honest, provided you are willing to say plainly that this is what you are doing, because a quietly redefined target is how people end up sincerely believing they were chasing something outside themselves.

The third is to accept the ambiguity as permanent and work anyway. That is defensible as a stance about a practice rather than as a claim about the world, and it is where a lot of good practitioners sit without fussing about it. What it does not license is reporting the result as though it settled something.

Offices Are Written to Be Unfalsifiable

Look at how the office statements are actually phrased and a pattern appears. They grant capabilities that are large, continuous, and stated as tendencies rather than events. A figure is said to grant wealth, to teach, to move favourably, to reveal. None of these has a natural failure condition.

Compare that with how you would state a claim if you wanted it to be wrong sometimes. You would name a domain, a time limit and an observable outcome. The catalogue almost never does this, which is a property of the genre rather than a fact about the figures.

This is not an accusation of fraud. These texts were written to be inherited and to serve a purpose, and a statement that is hard to disprove travels further than a precise one. But the consequence for modern practitioners is real: anyone who works with these descriptions has been handed claims that cannot be checked, and then asked to notice whether something happened.

The countermeasure is dull and effective. Restate the office in your own words as something that could be false, with a date attached, before you begin. If you cannot, you are not testing anything, and whatever happens later will be a story rather than a finding.

What the Catalogue Actually Is

The seventy-two figures that dominate this corner of the material are a literary and administrative artefact. They descend from Jewish and Christian demonological writing of the first millennium, pass through the apocryphal material, and arrive in their present shape in a seventeenth-century compilation that organised them for practical reference.

Read the surviving descriptions closely and the recurring picture is bureaucratic. These are functionaries with jurisdictions, permissions and areas of competence, the sort of figures a large institution generates to handle categories of problem. They are not torturers of innocents, and the more lurid modern readings are largely projections onto a medieval taxonomy that never made those claims.

The consequence for a practitioner is that the inherited material is far more useful as a catalogue of problem types than as a set of personalities. The figure associated with lending, discovery or the movement of obstacles is a well-organised description of a human problem, arrived at centuries ago, and that is genuinely valuable.

It is not a description of a being who will arrive, and the modern insistence on agency is a layer added on top of a document that never claimed it. That layer is where most of the difficulty in this field lives.

Why the Self-Reference Problem Does Not Resolve

There is a well-known move in the literature that appears to dissolve the difficulty: treat the figures as aspects of the practitioner's own mind, separate fragments with their own characters and purposes. It is elegant, it neutralises the risk, and it leaves the evidence problem completely untouched.

The reason is straightforward. If the figure is a fragment of your own mind, then your feeling of a distinct encounter is exactly what you would predict from your own mind, and there is nothing left to corroborate it. You have removed the danger and the information simultaneously, which is why it is so popular and why it satisfies nobody who is trying to know whether anything is on the other end.

There is a subtler version worth naming because it is the honest one. Practising with a named figure is a structured way of running a fixed problem through your own attention. The name supplies a frame, the frame holds your attention steady, and a held attention produces more coherent thinking about that particular problem than unstructured worry would. Everything real in this is happening in the thinking, and the sense of company is part of what makes the thinking work.

That account is not a consolation prize. It predicts the actual benefits people report, it explains why the work feels different from simply thinking about the problem, and it does not require a claim about what is present in the room. What it does not do is tell you whether you are right, and anybody who says the internal reading settles that question is overclaiming.

The Three Failure Modes Nobody Writes Down

The first is identity drift. You begin working with a figure because you need something specific from it, and after a few months the figure has become a general companion who is consulted about everything. The original bounded use was replaced by a relationship, and relationships are much harder to set down than tasks.

The second is escalating scope. The small working does not produce a result, so the next one asks for more, and the next asks for something adjacent. There is usually no decision point; the widening is gradual and feels like commitment rather than like a change in ambition. The tell is that you can no longer state what a successful outcome would look like.

The third is using a figure to avoid a decision. This is the most common and the least discussed. A working with a name attached to it can be carried indefinitely because the work is ongoing, and an ongoing work is a respectable place to leave an unmade choice. If a figure has been consulted about the same question more than three times, the consultation is the avoidance.

None of these is a spiritual hazard and all of them are ordinary failure modes that show up in any practice involving commitment, narrative and repetition. Naming them in advance is the whole of the practical defence.

The Servitor Comparison Is Not Close

The distinction between a servitor and a figure from a catalogue is usually drawn along lines of cosmology, and the cosmological part is the least useful part of the comparison.

The practical difference is scope. A servitor is purpose-built. You define a narrow job, you construct it to that specification, and you can inspect it. Its description contains a function, a method of activation and a retirement condition, which means it can be audited, debugged and eventually dismantled.

A catalogue figure arrives with a wide mandate, a body of inherited description you did not write, and no retirement condition at all. Nothing in the material tells you how to undo the relationship, which is the practical definition of a commitment you cannot exit.

That asymmetry is why the servitor is the better default even for people who hold the two frameworks to be equally true. One is engineered and reversible. The other is inherited and not. If you are going to build something to a specification, building it to a specification you can read is a large part of the benefit, and the metaphysics is a smaller part than the literature suggests.

Where the Historical Reading Is Simply Better

Setting the metaphysics aside produces a more interesting account of the tradition, not a thinner one, and it is worth stating plainly where the older material was doing something well.

The older sources describe preparation, timing and a stated aim with real precision. They are obsessive about conditions, and the obsession is not decorative: it is a way of building a specific and unusual state on demand, and it is the same technique used in disciplined arts from cooking to instrument practice. Nothing about it requires a non-human interlocutor.

They also treat the operation as reversible in a way the modern version has largely lost. Closing properly, marking the end, disposing of the residue, restating the boundaries: all of that is in the sources, and it exists for reasons that are practical rather than devotional. A working that cannot be closed accumulates, and the accumulation is the actual risk in this field.

The strongest reading available is therefore the boring one. These are techniques for producing a particular kind of attention under controlled conditions, inherited through a documentation chain that also gave us a memorable and largely inaccurate ontology. Keep the techniques, downgrade the ontology, and the practice becomes both safer and more honest than the version most people are running.

What Survives If You Drop the Ontology

More than most practitioners expect, and the reason is that the working does not depend on the ontology. The structure is: fix an aim, reduce it to a glyph, create a state, commit for a defined period, close, record. Every one of those steps is executable without a claim about what is present.

What you lose is the sense of company, which is the part people miss and the part that carries no evidential weight. What you keep is considerably more: the attention, the commitment, the structured thinking, the record, and the observable changes in conduct that committed attention produces.

Some people find that removing the ontology removes the practice. That is a legitimate result and it says something real about what they were doing. Others find the internal version works better precisely because it is not waiting for confirmation, and that these are different activities rather than two settings for the same one.

The useful test is empirical and cheap. Run a small working under the internal framing for a month, with the same logging discipline you would use for anything, and compare against your own baseline. If nothing observable changes, the framing was not doing the work and you have learned something useful cheaply.

Working With a Name You Have Not Traced

The name in this piece is a good specimen for the general problem, because it appears in no standard collection. There is no office statement, no described form and no document behind it, which means there is also nothing that can be checked.

That situation has two possible explanations worth separating. The name may be recent, a coinage, or invented, in which case the working is a technique with an arbitrary label attached and that is fine as long as you know it. Or it may be a corruption of something older, in which case you are working with a piece of noise.

The general principle is that effort spent tracing a name returns more than effort spent refining a working with it. These catalogues are indexed, reproduced in dozens of editions, and searched in a few minutes. A name that resists that process is telling you something, and the something is not usually encouraging.

If you do work with an untraced name, then document the gap rather than papering over it. Record that the source is unknown, record what you assumed about its function, and be aware that you are supplying the office yourself and therefore testing your own assumptions rather than the material's.

A Verdict Worth Defending

The honest summary of this corner of the practice is that the case for it does not rest on the ontology, and the ontology is what almost everyone in the field is arguing about. That is a strange position for a literature this confident, and it is worth naming as strange rather than treating as normal.

The case for the practice rests on a narrower and more defensible foundation. A defined aim, encoded, charged under controlled conditions and closed properly, produces a period of unusually coherent attention on a specific problem. That is a real effect, it is available to anyone, and it does not require anyone to believe anything about the entities involved.

The case against the strong version is that it is untestable as stated, that the sources are administrative documents rather than encounter reports, and that the vocabulary of agency was added much later by people who found the older material too dry. Each of those is a reason to downgrade the claim, and none of them is a reason to stop practising.

So the position I would defend is this. Use the technique, drop the premise you cannot support, keep the discipline of logging, and treat any result that seems to demonstrate an external presence as a question about your own attention rather than about the room. That position is less exciting than the one most people hold, and it is the only version of the claim I can actually stand behind.

Frequently Asked Questions

What is the strongest objection to working with named spirits?

That you cannot separate the figure from your own construction of it, since the evidence is generated by the same system that generated the expectation. Adding independent verification raises the bar far above where any of this evidence sits, and redefining the target as something internal removes the risk but also removes anything that could corroborate a result. The honest position is that the ambiguity is permanent and the practice is worth doing on non-ontological grounds.

Do the goetia figures correspond to anything real?

The catalogue is a documentary artefact built from earlier Jewish and Christian demonological writing and organised for reference in the seventeenth century. What survives in it reads as administrative, describing jurisdictions rather than predators. Whether anything corresponds to it is a question no one has managed to answer with evidence, and the material was never a set of encounter reports to begin with.

How do I stop a working with a name if I need to?

Write the ending down before the beginning, because that is the only time you will reliably do it. Classical practice gives you the materials: close the operation explicitly, state the boundary, dispose of or archive whatever the method specifies, and stop using the name. The absence of any retirement condition in the modern literature is a genuine design flaw rather than a feature, so supply your own in advance.

Is it worth learning this if the ontology is doubtful?

Yes, on the strength of what the technique actually does. Fixing an aim, encoding it, holding a specific state of attention and closing cleanly will improve how you think about a chosen problem, and that does not depend on what is or is not present. Drop the claim about external presence and you lose the sense of company, which is the part that carries no evidential weight anyway.

How do I stop using a figure as a way of avoiding a decision?

Count the consultations about the same question. If a name has been brought in more than three times about one issue, the consulting has replaced the deciding, and the mechanism is the same as any other avoidance loop. Write down what the decision is, the date by which you will make it, and treat the next consultation as the signal that the deadline needs to move rather than the question needing another reading.