Occult Culture and Modern Practice

Somebody once asked me how to find a real teacher, and I gave an answer about lineage that I now think was wrong in a way worth correcting. The reason it was wrong is not that the answer was inaccurate. It was accurate about the tradition and useless about the situation, and the gap between those two is where this entire subject lives. What has happened to occult practice in the last thirty years is a change of medium, and almost every question people bring to it is a question the previous medium answered well and the current one answers badly. Search has replaced the ladder. That is a genuine gain in access and a real loss in something specific, and the specific thing that was lost is not knowledge. It was transmission: the local, unremarkable, undocumented practice that ran through families and neighbourhoods and was never written down because writing it down was not what it was for. Everything that survives online is the part that was already documented somewhere. The rest was practice, and practice does not travel through search results. What follows is an attempt to be precise about that, about what the medium genuinely changed, and about what to carry across the transition if you are one of the people trying to.

A scene, described as a distribution system

The useful way to think about contemporary occult practice is as a distribution problem rather than a content problem. A content problem asks what is true. A distribution problem asks how a thing reaches a person, who decides what is worth putting in front of them, and what gets lost in the distribution. Almost every disagreement in this area is really a disagreement about distribution being mistaken for a disagreement about content.

The audience is overwhelmingly online now, and the model of discovery is search and forum. That has three consequences worth naming. There is no gate, so there is no filtering by the person who knows. There is a ranking system, so there is a permanent incentive to make the material legible to a search engine. And there is an archive, so there is a version of the record that survives the people who made it. All three are improvements on the previous arrangement and all three are losses, and the losses are the reason this subject feels different now than it did in 1975.

Search replaced the ladder

The older route was a ladder. You found a person, they initialled you, they passed you something specific that you would not have found otherwise, and the sequence took years and had a definite end. The ladder did three things well. It sequenced the material, so you did not meet advanced things before their prerequisites. It filtered it, so most of what you were shown had survived somebody's scrutiny. And it gave you a person, which is the only form in which feedback about your specific case is possible.

Search does the opposite in each case, and the opposite is not worse in itself. You can find the advanced thing first. You are shown the material nobody filtered. And you get an enormous quantity of people who cannot tell you anything about you in particular. A person who has read ten times what you have read is not a teacher. A search engine that surfaces ten thousand documents is not an initiatory sequence. Both are true, and a lot of confusion in this field comes from treating them as substitutes for one another.

The genealogy that never passed through reconstruction

The witchcraft most English-language beginners encounter is a nineteenth- and twentieth-century construction. It was assembled, at a specific historical moment, from folklore fragments, a novel, a speculative anthropology, and the practical needs of people who wanted a nature religion they could practise outdoors. It is a real tradition now, with real practitioners and real internal disputes, and it deserves to be taken seriously on its own terms.

It is not, however, the continuation of the practices of the people in Peru, Mexico, Brazil, or the Caribbean. Those traditions are different in kind. The work is often local and specific rather than derived from a European frame. It runs through families rather than through publications. In several regions it includes plant knowledge that was never written down, held by people who had no reason to write it down and no contact with the reconstruction. The reconstruction and those traditions share a continent sometimes and little else, and treating them as branches of the same tree is an error that damages both.

What was lost when the road opened

Not the plant knowledge in the sense of it being gone. It is in places still held, and where it has been lost, the loss is from violence, dispossession, and the collapse of local economies rather than from the internet. What was lost is the transmission mechanism. Nobody documented that knowledge because documentation was not the point, and the person holding it was not looking for a movement.

The consequence is that a beginner online in 2026 cannot find it by searching, and a search engine has no way to surface what was never indexed. The practical version of this is that there are regions with living practice, and the person who is in a hurry to get started is the person who is most likely to be told about the reconstruction instead. The version of this that affects you is narrower and more manageable: when you read about witchcraft, find out whether you are reading a reconstruction or a description of something continuous, because the two support completely different conclusions about what you can expect to do.

DIY, running parallel since the late seventies

The other thing that has been continuous is not institutional at all. Punk and industrial subcultures have been a consistent site of improvised magical practice since the late 1970s, running alongside formal orders the entire time and rarely connected to them. This is the most under-reported fact about the history of modern occultism, because it happened in rehearsal spaces, print zines, and record labels rather than in institutions, and none of it generated paperwork.

What the DIY line contributes is a specific and durable method: take the components, assemble your own system, test it, publish it, let other people take it apart. That is much closer to how this field actually advances than the transmission model is, and it explains a lot of the field's character, including its habit of producing its own critics. If you are looking for a line of descent that has been running continuously since 1977, the orders are the wrong place to look and the small publications are the right one.

Memes as shared symbols, taken seriously

Some of what circulates in this field online is trivial, and the reflexive dismissal of it as trivial has been unhelpful. A meme archetype circulating across thousands of unrelated accounts, with no central author and consistent features, is a shared symbolic object in the ordinary anthropological sense. People are using it to mark affiliation, to signal a shared frame of reference, and to make jokes that only make sense if you share the object.

That is what symbols are for, and the fact that the transmission is through image and repetition rather than through a text does not make it something else. Symbolic objects require shared reference. Shared reference is exactly what a distributed, gate-less medium produces most easily. The honest reading is that this is a new form of a very old thing, that it is thinner than the forms it parallels, and that it is doing real work in the lives of the people using it. A field that dismisses it has decided its own symbols are the only real ones, which is not a position anybody else is obliged to accept.

What the medium changes and what it does not

The medium changes four things. Access: nearly unlimited, including for people with no money and no local community. Sequencing: mostly lost, since search returns the most legible thing rather than the right thing. Filtering: mostly lost, replaced by a ranking system with different incentives. Memory: greatly improved, because the record persists and can be checked, which is the one unambiguous gain of the whole arrangement.

It does not change four other things. Whether a practice produces a result. Whether a claim is true. The concentration of attention in a person, which is the actual mechanism. And the social cost of being wrong, which is if anything lower online, because nobody knows what you were doing at two in the morning. The medium is better at storage and worse at sequencing and filtering. Everything else it does to you is downstream of those two words, and a person who has noticed that will be far less confused than one waiting for the medium to solve the wrong problem.

Reading a scene without romanticising it

There are two available narratives here and both are bad. One says the tradition is dying, has been dying since the industrial revolution, and the internet is the last thin light. The other says the internet democratised access and the old gatekeepers were always gatekeepers of nothing. Both are comfortable and neither helps anybody decide what to do on Tuesday.

The version that holds up is duller. There is a real body of practice, distributed unevenly across a very large number of people, with better documentation than it has ever had, almost no filtering, almost no sequencing, and a handful of serious injuries sustained by people who took an unfiltered sequence as a curriculum. That is a scene. It has people in it who know things and people who have read a lot and people who are currently making a mistake. That is normal. Every practice community ever documented looks like that, and the ones that look otherwise are the ones not being written down.

The regional traditions, on their own terms

If you have any interest in the material outside a European frame, the practical rule is that you are a visitor, and the practice is to learn from somebody inside it rather than about it. That can be awkward and it can be slow, and it is the only version that produces anything other than a slightly better-informed version of the reconstruction.

The barrier is real and worth naming rather than glossing. A great deal of the material is in Spanish, Portuguese, and indigenous languages; the interpretive framework is local, not theoretical; and the priorities are frequently different, with a much stronger emphasis on family and community context and a much weaker one on individual practice. The attempt to import individual aims into a practice organised around relationship produces the same category error that individual aims produce in a community-structured tradition generally. Read first. Ask second. Assume third.

Where online discovery actually goes wrong

Three specific failures, all of them predictable and all of them expensive. The first is taking the top result as a foundation, which in this field means taking a piece written for engagement and treating it as a curriculum. The second is mistaking volume of production for depth of practice, which is a search-logic error: a term with a large audience produces a large amount of writing and a large amount of writing is not a large amount of experience.

The third and worst is the unfiltered sequence, and I want to be specific about why it is worse here than in most hobbies. A lot of what circulates contains instructions. Instructions taken in the wrong order, or by somebody in the wrong state, or applied to a situation the instruction was not written for, do the kind of damage that a badly taught recipe does not. This material has always had this problem and it was always mitigated by the ladder. Without the ladder, the mitigation is the individual doing the sequencing, which means the sequencing is the actual skill now and not the practice.

What to carry across the medium change

Sequencing, filtering, and a person. Those are the three things the previous arrangement supplied and the current one does not, and they are the whole content of the useful advice available to somebody starting now. Sequencing means reading in an order you chose deliberately, from material that assumes the earlier material, rather than from whatever comes up. Filtering means finding a small number of sources and reading them thoroughly enough to know what they are not saying. A person means somebody whose judgement you have seen operate on a case you did not choose.

The three are available and the field contains all of them, which is the optimistic reading. The reason they are not widely advertised is that none of them scales, and the medium rewards scale. A field can be simultaneously much larger and much less able to get one person from here to a working practice, and that is what has happened. If you are the person asking the question, then those three are what you are actually looking for, and it is worth being slightly unfashionable about wanting them.

The question the scene raises

The scene is now large, well documented, unfiltered, and full of people who are confidently wrong from a single search. It has never been more open and it has never been more dangerous to take badly sequenced material seriously. Both of those are consequences of the same change, and neither is going to reverse.

So the question worth asking is not whether the scene is healthy, which is not a question with an answer, but whether you are going to be one of the people who handles it. That means sequencing your own reading, checking what a source is before trusting it, and finding one person to check your work against. It is a smaller ambition than a movement, and it is the only one that has ever been available to an individual, in any medium, in any period of this practice. The internet did not change that. It only made the failure mode cheaper.

Frequently Asked Questions

Is witchcraft in Latin America the same as Western witchcraft?

No, and the difference is structural rather than a matter of local flavour. The reconstruction most English-language beginners encounter is a nineteenth- and twentieth-century European construction. The traditions of Peru, Mexico, Brazil, and the Caribbean are continuous with local practice, often organised around families and communities rather than individuals, and in several places they preserve plant knowledge that was never written down. Treating them as branches of a shared tree is an error that damages both, mostly by making the non-European traditions sound like regional variants of something that came first.

What did the internet actually ruin?

Sequencing and filtering, specifically. Before it, the material reached you in an order and somebody with judgement had decided what you were ready for. Now the material arrives unordered and unfiltered, and a ranking system with incentives to be legible has replaced the person with judgement. Access improved enormously and documentation is far better. The two losses are the serious ones, because a lot of the material contains instructions that do damage when taken out of order, and the ladder was what prevented that.

Do memes and shared online symbols count as real symbolism?

Yes, in the ordinary anthropological sense, and the reflexive dismissal of them is a position rather than a finding. A symbol needs shared reference, and shared reference is what a distributed gate-less medium produces most easily. These are thinner than the forms they parallel and they are doing real work in the lives of the people using them. A field that dismisses them has decided that its own symbols are the only real ones, and nobody else is obliged to agree.

Can I learn from a book alone?

You can learn a great deal, and there is enough documented now that the books are not the bottleneck. What a book cannot supply is sequencing, which is the part you have to construct, and feedback on your specific case, which is the part that actually catches errors. The DIY tradition that has been running since the late 1970s is closer to what you want than any curriculum, because it assumes you will assemble your own system and test it. That is a viable path. It requires you to be the one doing the sequencing.

How do I tell a real source from a content farm?

Ask who wrote it and what they wanted it to do, then check whether the text is legible to a search engine. Both tests catch most of it. A practice document written by somebody with a practice in it will be specific about failures, and a document written for a ranking system will be general, confident, and organised around a promise rather than a procedure. It is worth remembering that a source can be sincere and still be bad, and that the first question does not settle anything. It only narrows the field enough to read more carefully.