Lay the same Celtic Cross on the table and make yourself a promise: you will read it three times, under three mutually incompatible world‑pictures. First as Jungian dream‑image. Then as a briefing from the dead. Then as a fixed‑fate pronouncement in the Valens sense: this will come to pass.
If all three passes feel coherent, the awkward question arrives: what, exactly, is the deck itself committed to?
The hardware problem: what survives a reboot?
Tarot did not start as an oracle at all. Fifteenth‑century Italians were playing a trick‑taking game with a set of trumps and pips; there is no evidence anybody was mapping Hebrew letters onto the Fool or reading the 7 of Cups as dissociation. The “Book of Thoth” story is De Gébelin’s eighteenth‑century fiction; the Golden Dawn’s Tree of Life attributions are nineteenth‑century. The metaphysical scaffolding most of us treat as “built in” is historically late.
Yet tarot has repeatedly survived being dropped into radically different cosmologies.
De Gébelin, writing in 1781, insists the same pack of cards is both a parlour game and an Egyptian hieroglyphic scripture encoded under Providence, a fragment of *prisca theologia* hiding in plain sight. Etteilla, within a decade, rewires it again: now it is literally the cosmogonic Book of Thoth, a machine for spelling out your destiny in concrete events—illness, catastrophe, betrayal—bordering on hard fatalism. Lévi, mid‑nineteenth century, overlays Kabbalah and ceremonial magic: the trumps become a synthetic key to occult philosophy, “a veritable machine for generating thought,” as he calls it. (The specific Hebrew letter‑to‑trump assignments most practitioners now take for granted are not Lévi’s but the Golden Dawn’s; Lévi associates the trumps with the Hebrew alphabet as a class, but the precise path attributions—Fool to Aleph, Magician to Beth, and so on—are a nineteenth‑century GD synthesis, not a received tradition.) Waite, a Christian mystical sacramentalist who distrusts vulgar fortune‑telling, still prints mundane “divinatory meanings” alongside veiled Christological exegesis. Regardie, in collating Golden Dawn material, sits divinatory instruction—timing love affairs and lawsuits—next to high‑theurgic pathworking on the Tree.
Then, in the twentieth century, Jung writes his foreword to Wilhelm’s *I Ching*, and the game is re‑framed again. No longer, primarily, a fate‑machine; instead, an acausal mirror of archetypal situation. A later literature—most influentially Sallie Nichols’s *Jung and Tarot* (1980), though the lineage extends through Rachel Pollack and beyond—takes that move and runs with it: tarot as projective test, as dream‑language of the unconscious.
The same paper rectangles keep carrying all this.
What is so structurally robust in the cards that you can run De Gébelin’s providential perennialism, Etteilla’s cosmogonic determinism, Lévi’s astral‑light Hermeticism, Waite’s sacramentalism, and post‑Jungian psychology without the whole thing collapsing?
If we borrow a metaphor from computing, the answer looks like this:
– the physical deck and spread topology are the hardware,
– the way positions, suit families, number, and visual relations behave is a kind of instruction set,
– the cosmology—psychological, animist, hard‑fatalist, ancestor‑field—is the operating system.
You do not change the hardware between readings. You change the OS. The question is how far that OS is really free to vary, and where the underlying instruction set quietly pushes back.
The “ISA” of a spread
In computer science, an Instruction Set Architecture (ISA) is the minimal grammar of operations a piece of hardware can perform: add, load, branch, jump. Different operating systems, written in different languages, all have to compile down into that instruction set.
Reading practice has something analogous. Before you say “this card is your father” or “this card is your shadow,” you are relying on much more basic moves: adjacency, contrast, reinforcement, opposition, temporal direction, rank. In a Celtic Cross, for instance, the “hardware‑level” relations include:
– a central *state* (card 1) and a *crossing* or tense relation (card 2),
– a *beneath / root* position (card 3) contrasted with *above / conscious aim* (card 5),
– a *behind / past* and *before / near future* (cards 4 and 6),
– an explicit distinction between *self* (7) and *environment* (8),
– a trajectory from hopes/fears (9) to outcome (10).
None of that, on its face, tells you whether the “environment” is other people, spirits, probability fields, or your own projections. But it does assume that:
– time can be segmented (past / present / future),
– there is some kind of layered structure (surface / depth),
– there is a meaningful difference between self and not‑self,
– something like a trajectory or *telos* exists.
That is already not cosmologically neutral. Structurally, it is much easier to plug this spread into Ptolemaic soft‑fate, Confucian moral process, Jungian depth psychology, Kardecian Spiritism, or a Golden Dawn initiatory path than into, say, a strict eternalist block‑universe with no meaningful “advice” or a radically non‑dual metaphysic where self/other is illusion. You *can* try to run those, but you will immediately feel where the grammar strains.
This is the first thing a serious practitioner gains by thinking in “hardware” terms: you can begin to distinguish what your deck/spread *already* presupposes from what you have layered on top of it.
Historically, we can see this grammar survive ontological reboot. The same sixteen geomantic figures move from late‑medieval fatalistic “judgements of fate” to Agrippa’s angelic and planetary correspondences without changing shape. Islamic *‘ilm al‑raml* couches them in Qur’anic monotheism; Renaissance Christians move them into a names‑of‑God framework. The rule‑set for generating figures and houses stays put. What changes is who, or what, is taken to be speaking through them.
Likewise, hexagrams in the *Yijing* retain their six‑line structure and correlative network as they go from Zhou‑period divinatory practice through Confucian moral cosmology to Wilhelm’s early twentieth‑century spiritual classic—and then into Jung’s synchronicity‑psychological reading. The Ten Wings give you one OS. Jung’s foreword explicitly boots another.
Tarot’s “ISA” is neither as tight as geomancy nor as ancient as the *Yijing*, but the pattern holds. De Gébelin moralises trumps as stages of the soul. Nineteenth‑century chapbooks treat them as blunt event‑tokens (“a dark man brings misfortune”). Lévi maps the majors to Kabbalistic paths. Waite splits sacramental major‑trump exegesis from banal fortune‑telling minors. Post‑Jungians read majors as individuation archetypes, minors as ego process. In each case, positional and relational grammar does most of the heavy lifting; cosmology tells you what the pieces stand for.
So the “ISA” here is not metaphysically empty. It is a relatively stable set of affordances—a game board with a certain kind of grid—and each cosmology is a rule‑set for what those positions and relations *mean in the world*.
Four emulations, one layout
Better to make this concrete. Take the same Celtic Cross, and force it through four OS’s, as strictly as you can.
Use the example sketched in the brief: querent asking why they feel blocked about leaving a job for freelance work. Cards fall as:
1. Present: 8 of Swords
2. Crossing: Devil
3. Root: 4 of Pentacles
4. Past: 10 of Wands
5. Conscious aim: Star
6. Near future: 2 of Pentacles
7. Self: 9 of Swords
8. Environment: 3 of Pentacles
9. Hopes/Fears: 5 of Cups
10. Outcome: Chariot
Now declare your operating systems up front.
Psychological OS
Assumption: the spread images intrapsychic structure. Synchronicity, not causality; cards show archetypal situation, not external decree.
– 8 of Swords / Devil as present + crossing: the querent is trapped in a self‑generated narrative of helplessness, compounded by compulsive patterns (perhaps addiction to security, shame, or people‑pleasing).
– 4 of Pentacles root: early imprinted scarcity complex; “hold on or you die.”
– 10 of Wands past: chronic over‑responsibility—family expectations, taking on too much.
– 9 of Swords self: clinical‑level anxiety, insomnia, rumination.
– Star aim: a conscious longing for healing and a more meaningful vocation.
– 2 of Pentacles near future: tentative juggling—testing freelance projects whilst staying at the job.
– 3 of Pentacles environment: actual collaborative opportunity is available; colleagues or a small team could support the transition.
– 5 of Cups hopes/fears: grief over imagined failures; fear of regretting either staying or leaving.
– Chariot outcome: integrated will. If the internal conflict is worked through, there is capacity to take the reins.
Advice is obvious: therapy or shadow work around scarcity and over‑responsibility; nervous‑system work for anxiety; graded exposure to risk.
Animist OS
Assumption: cards map a field of non‑human persons (spirits, land, egregores) affecting the situation. The querent is not the only agent.
– 8 of Swords present: the immediate felt trap.
– Devil crossing: a binding relationship with the company’s egregore, or a parasitic entity attached to this pattern of work—think “spirit of corporate overwork” rather than merely “bad boss.”
– 4 of Pentacles root: a land‑spirit or house‑spirit that has stabilised around a steady paycheque and dislikes the instability freelance life would bring.
– 10 of Wands past: ancestor or guardian spirits watching a history of overburdened slog; they may be “used to” you carrying too much and default to preserving that state.
– 3 of Pentacles environment: workplace spirits / *genius loci* that actually favour your craft and collaboration there.
– 2 of Pentacles near future: a period where you are asked to consciously negotiate between these beings—make offerings, declare intent, redistribute loyalties.
– Chariot outcome: if you renegotiate the spirit‑contracts (explicitly releasing, redirecting offerings, making a pact with a tutelary spirit of your freelance work), those forces pull rather than block; you move with a team, not against headwinds.
Advice: ritual, offerings, explicit diplomatic work. That is not a metaphor in this OS.
Ancestor‑field OS
Assumption: cards show lineage patterning and the pressure of the dead.
– 4 of Pentacles root: grandparents who knew real hunger or precarity, passing down a vow: “never risk a good job.”
– 10 of Wands past: parents or earlier ancestors who worked themselves into illness maintaining stability; you have internalised this as love.
– Devil crossing: entangled loyalty—if you break the pattern, some part of you fears you betray their sacrifices.
– 8 of Swords present: you are bound in invisible filial piety.
– 5 of Cups hopes/fears: grief at potentially “disappointing” or “leaving behind” the line.
– 3 of Pentacles environment: living family members echoing the script: “Be sensible.”
– 2 of Pentacles near future: a liminal period where you are half in the old line, half in your own path.
– Chariot outcome: you becoming the ancestor who moves the pattern forward, but only if you explicitly acknowledge the dead, mourn with them, and ask for their blessing or at least their release.
Advice: altar work, naming the ancestor stories, perhaps ritual acts that symbolically carry them into your freelance future so you are not “abandoning” them.
Hard‑fatalist OS
Assumption: events and main choices are fixed on a script; cards show what is written. This is Valens, not Ptolemy.
– 8 of Swords / Devil: the script includes a period of inescapable bind and bondage to necessity. You will not, in fact, leave the job yet.
– 4 of Pentacles root: the economic conditions are such that holding on is non‑optional, regardless of psyche or spirits.
– 2 of Pentacles near future: a fated time of juggling both job and side work; your attempts at a clean break will be thwarted.
– 3 of Pentacles environment: the universe has written in useful colleagues and modest success in the current workplace; this is not a trap of error but a stage.
– Chariot outcome: at a specific (not yet revealed) point, the line turns; the move happens whether or not you “work on yourself.” You will leave; not yet.
Advice position, under pure fatalism, has to be re‑coded: not “here is how you change the plot” but “here is the posture in which you endure it with least suffering.”
Notice what just happened. The ISA‑level pattern—the sense of bind, scarcity root, overwork history, potential movement after a juggling period—persisted across all four passes. What changed was:
– who is construed as the blocker (internal complex, spirits, ancestors, impersonal Necessity),
– what counts as a legal move (therapy, ritual, lineage repair, attitude shift),
– how you handle agency and blame.
Structurally, the spread compiled quite cleanly to the first three OS’s. Under hard fatalism, you could feel the friction immediately: a spread with explicit advice and outcome positions is architecturally pro‑intervention. To make it fit strict determinism, you had to twist “advice” into “attitude.”
That is where you start to see what the deck’s grammar itself will and will not sustain.
Is this just projection in a funny hat?
A sceptical colleague will interject here: you are not comparing ontologies, you are just displaying interpretive agility. Any half‑competent reader can retrofit any layout to any story. Where is the test?
Two honest moves in response.
First, drop the pretence of doing science. Borrowing the language of ISA, model pluralism, and causal graphs is useful because it forces conceptual discipline—not because tarot is secretly a Bayesian thermometer. We are not computing posteriors. We are borrowing the shape of “multiple models can fit the same data differently” to think more clearly about our own practice.
Second, insist on some non‑trivial standards. If a “reboot” is going to be more than rhetoric, each cosmology you run on the spread needs to:
– generate substantively different recommended actions or stances, not just different mythic wrappers for the same advice;
– be operationally coherent with the grammar of the spread (an OS that turns the advice position into permanent “nothing can be done” in every reading is a bad pairing with that board);
– produce, over time, some distinguishable track record for particular classes of question.
This is not controlled experimentation. But you can, over years, notice patterns such as:
– timing/legal/immigration questions tend to track best when you allow a strong fate/probability frame and read “environment / near future / outcome” positions in that mode;
– entrenched relational patterns often respond better—both in felt sense and in eventual behavioural shift—when you treat certain positions as ancestral voice rather than exclusively as inner child;
– spirit/animist toggles around land, house, sickness, and business egregores sometimes light up in ways a pure psychology frame struggles to explain.
If, on the other hand, your “animist” and “psychological” passes always lead to the same prescriptions, just with different names attached, you have not meaningfully rebooted the OS. You have reskinned one metaphysic.
The value of the exercise is precisely here: it lets you catch yourself in that act.
A fourth objection comes not from the sceptic but from the committed practitioner: if you adopt an animist OS, you are not merely selecting an interpretive lens. You are entering into relationships with non-human persons, accepting obligations that do not dissolve when you decide to reboot. A house-spirit whose assistance you have invoked does not disappear because you have switched to a psychological frame for the next reading. Ritual and relational commitments accumulate across sessions; ontology-switching carries actual relational costs in the frameworks that take relations seriously. This objection has real bite. It is also, precisely, an OS-level claim—about whether spirits are genuine others with standing or psychic contents with numinosity. The emulator framework does not resolve that dispute. What it does is make the stakes of the choice legible: you are not just picking a metaphor, you are committing to a set of obligations, and different frameworks carry different ones.
One layout, incommensurable explanations
Philosophy of science has lived with a related problem for a long time. Ian Hacking showed that observation is always theory‑laden: what you see depends on the instrument and the interpretive framework you bring. Nancy Cartwright argued that scientific models do not represent the world directly but only relative to a context of application—fit is always local, never global. Ron Giere pressed the point: the relation between model and world is mediated by a purpose, a user, and a set of background commitments that are never fully explicit.
Treating an oracle as an emulator of cosmologies is the same move at a ritual level.
– The spread is the measurement device: it generates structured output (pattern of cards).
– Your cosmology is the measurement model: a mapping from “pattern of cards” to “claims about the world.” “If this is ancestral, what sort of layout would I expect?” is a real question.
– Different cosmologies fit the same spread by emphasising different causal stories and different interventions.
Underdetermination hits divination as hard as it hits physics. Does that mean cosmology is just aesthetic? Not if you keep your eye on action.
Two OS’s that produce indistinguishable practical consequences for a class of questions are not genuinely different models for that domain; they are poetic variants. When hard fatalism and psychology both say, “You can do nothing; accept things as they are,” they are functionally aligned, despite different metaphors. When an animist read prescribes ritual re‑negotiation and a psychological read prescribes only reframing, you have a concrete divergence.
That is where, as a practitioner, you can begin to ask: which causal grammar seems to move this situation? When my clients actually do the ancestor rituals suggested by the ancestor‑frame readings, do external circumstances shift in ways that were not shifting under purely psychological work? When I predict timing under a Ptolemaic “tendencies modifiable by choice” reading versus a Valens “this will happen regardless” reading, which track more cleanly?
The answers will be messy, biased, path‑dependent. They will not settle metaphysics. But they will tell you something about how your deck, your spread grammar, and your psyche collaborate in practice.
The non‑neutrality of the hardware
Back to the first objection: is there really an ISA‑like layer that is independent of cosmology, or am I just smuggling my own metaphysics in at a lower level?
Tarot’s so‑called “grammar” is absolutely not primordial. Number symbolism, suit elementalism, major/minor hierarchies, Kabbalistic pathworkings—they are all late synthetic overlays. Early playing‑card cartomancy chapbooks happily assign blunt event meanings to pips without any Hermetic doctrine. Anti‑divinatory Christian authors moralise the trumps as an allegory of the soul’s journey whilst denouncing fortune‑telling. Wirth pushes an initiatory emblem‑book reading and sneers at mundane prediction; contemporaneous folk readers with the same deck ran entirely different software—event-tables, lucky and unlucky pip-counting, marriage and money predictions—with no knowledge of and no interest in Kabbalistic paths or theurgic self-transformation.
So the ISA is historically assembled, internally contested, and nowhere near neutral. What it is not is infinitely plastic. A spread grammar built around self/other, past/present/future, surface/depth, and outcome/advice already presupposes that time is asymmetrical, that a self distinct from environment exists, that the present moment is accessible to counsel. These are not arbitrary choices; they are the structural precipitate of centuries of actual divinatory use, tested against the kinds of questions people brought to the table. They tell you what this hardware was shaped to handle and where it strains.
That residue is not a problem. It is what makes the grammar stable enough to carry disparate cosmologies without collapsing. The practitioner who understands the layering—who can say “this is the hardware pushing back” rather than “this is my cosmology forbidding it”—gains something technically important: the ability to revise their interpretive framework without discarding accumulated reading skill. The Celtic Cross they have drilled for a decade still compiles. What changes is the OS. That change can be provisional, experimental, even temporary. The deck survives it.
The cards are not neutral. Four centuries of use have shaped them into tools that presuppose certain things about time, agency, and the self’s relation to a wider field. That is neither a flaw nor a metaphysical commitment to any particular cosmology. It is a design feature—the residue of historical testing, the convergent grammar that survived because it was useful enough to carry very different worlds. A practitioner working deliberately across multiple OS’s is not free-floating in relativism. They are running controlled experiments on historically stable hardware, with a grammar that has already earned its keep.
