Tangential Voynich object-world experiment 07. This article does not claim that Beinecke MS 408 is a machine, automaton or mechanical book. The mature European court-automaton culture used here as a comparator is later than the manuscript’s early-fifteenth-century parchment range. That chronological distance is deliberate. We borrow the interaction logic of a marvel object to ask whether a manuscript can have a resting state, an activated state, an operator, an audience and a designed sequence of revelation.
Previous: Reliquary or Vessel · Tangential Voynich | Systemic Parallax · Voynich Research Library
What If Looking at the Closed Object Is Not the Same as Experiencing It?
Put a mechanical marvel on a table.
Before anyone touches it, it is already an object.
It has material, surface, weight, ornament, rarity and presence.
Then somebody activates it.
A lid opens.
A figure moves.
A clock reveals another mechanism.
A hidden sequence unfolds.
The object that was sitting on the table becomes an event.
The Metropolitan Museum’s Making Marvels project examined European court objects from roughly 1550–1750—including automata, clocks, scientific instruments, precious vessels and other works in which artistry, technical knowledge, prestige and wonder met. Some were activated in social settings precisely because movement, surprise and transformation were part of their effect. The Met: Making Marvels — Exhibition Guide.
That world is too late to be used as direct Voynich provenance.
Perfect.
We do not want an ancestor.
We want a question-generator.
What if a manuscript has a difference between its resting state and its activated state?
A Codex Is Already an Interactive Object
A manuscript does not move by itself.
But neither is it static in the way a wall painting is static.
To experience a codex, a receiver performs operations.
- pick it up;
- orient it;
- open it;
- turn leaves;
- hold a spread;
- extend a foldout;
- rotate or reposition a large diagram;
- compare separated pages;
- close and reopen it at another location.
The physical object therefore has states.
S₀ = closed
S₁ = ordinary opening
S₂ = spread visible
S₃ = foldout partially extended
S₄ = maximum reveal state
This does not make the codex a machine.
It makes handling part of the evidence.
The First Test: Activation Topology
Imagine drawing every legal physical state of the manuscript as a node.
Connect two states when a reader can move directly from one to the other.
We now have an activation graph.
For an ordinary single leaf, the graph is simple: turn to it, look, turn away.
For a complex foldout, the graph may contain several intermediate states. Particular image relationships may exist only at full extension. Other components may remain hidden until a flap or folded region moves.
The Tangential test asks:
- Which manuscript information is visible in every state?
- Which appears only after extra physical operations?
- Which visual relations emerge only at maximum extension?
- Are high-complexity diagrams disproportionately associated with high-complexity handling?
- Does the physical path to information appear designed rather than incidental?
The neutral residue is activation topology.
The Second Test: Operator Dependence
An automaton needs someone—or something—to activate it.
A manuscript also requires an operator, but researchers rarely model that person explicitly.
Who knew how to handle the object?
Who knew which leaves should be compared?
Who knew that a diagram should be rotated, extended or read in a particular direction?
Who knew what to say while the page was being shown?
The automaton tangent introduces operator dependence:
How much of the object’s successful use depends on knowledge that is located in the human operator rather than on the visible page?
A low-operator-dependence manual teaches the user through the object itself.
A high-operator-dependence object may be perfectly functional inside a small trained community and baffling after that community disappears.
This is one way an artefact can become “mysterious” without ever having been designed as a mystery.
The Third Test: The Performance State
A marvel object can be understood differently in a private inspection and in a public demonstration.
Now give the Voynich Manuscript two hypothetical contexts.
Context A: one trained user consults it alone.
Context B: one trained operator shows it to several observers.
Some design choices might function differently in the two contexts.
- A foldout may aid calculation in A but create spectacle in B.
- A dense circular diagram may serve reference in A but create authority in B.
- An unfamiliar script may be efficient shorthand in A but signal exclusivity in B.
- A visually unusual page may be mnemonic in A but memorable in B.
One physical feature can support several functions.
The neutral test is contextual function multiplicity: can the same design element plausibly solve both a private-use problem and a public-performance problem without changing its physical form?
The Fourth Test: Suspense Architecture
A marvel is often structured around timing.
Not merely what happens, but when the receiver discovers it.
Before the mechanism moves, the observer forms an expectation.
The activation violates or exceeds that expectation.
Can a codex have suspense architecture?
Possibly, in a much weaker sense.
- Does page singularity increase before exceptional foldouts?
- Do visually ordinary sequences terminate in unusually complex interfaces?
- Are some large diagrams hidden behind ordinary-sized closed states?
- Does the object contain reveal events whose visual impact cannot be inferred from the preceding page?
If so, we can measure expectation discontinuity: how abruptly visual or structural complexity changes across an interaction boundary.
The Fifth Test: Replayability
A good marvel often survives repetition.
The second viewing is not identical to the first because the receiver now knows where to look.
This is deeply relevant to manuscript design.
Some pages may reward repeated consultation because they contain layered relationships. The first pass produces gross recognition. Later passes reveal recurrence, correspondence or detail.
Call this revisit yield.
For a given receiver:
RY(n) = new recoverable structure on encounter n
A trivial decorative object may have high first-glance impact and rapidly falling revisit yield.
A densely structured object may remain productive over many encounters.
This distinction will become important when we finally ask whether fascination itself could be a design function.
The Sixth Test: Hidden Mechanism and Visible Behaviour
An automaton has an internal mechanism and an external performance.
The audience may see the bird sing without seeing the gears.
This creates a useful formal separation for Voynich:
latent rule system → visible manuscript behaviour
We already observe visible behaviours: positional constraints, recurring forms, boundary effects, varying page populations, repeated visual grammars.
We do not know the latent mechanism generating them.
That mechanism could involve language, notation, copying, cipher, formulaic generation, trained scribal convention or some combination.
The automaton metaphor encourages one discipline:
Do not infer the hidden mechanism merely from the beauty of its visible performance.
Many different mechanisms can produce similar outputs.
That is an identifiability problem, not a mystery-romance problem.
The Seventh Test: Audience Coupling
A marvel has an audience because reaction is part of its social return.
Applause, surprise, conversation, prestige and memory happen in people, not inside the mechanism.
This gives us a receiver-centred variable:
audience coupling = how strongly the object’s value depends on the response it produces in observers
A multiplication table has low audience coupling. It works even if nobody is impressed.
A court marvel can have high audience coupling. Astonishment is part of the result.
Where does a manuscript sit?
We cannot answer historically yet. But we can distinguish design features whose value is largely receiver-independent from those whose value rises when shown socially.
The Eighth Test: Transformation Without Machinery
The most useful return from the automaton world may be this:
An object can transform without containing gears.
A closed codex becomes an open spread.
A compact foldout becomes a large field.
A sequence becomes a network when distant pages are compared.
An unintelligible script can become increasingly predictable as a trained reader learns its conventions.
The transformation can occur jointly in object and receiver.
experienced state = physical state × receiver state
The same page is not cognitively the same page before and after the receiver has learned its local grammar.
This is a powerful idea for Tangential Voynich because it stops treating meaning as a property simply waiting on the parchment. Use may require the right object state and the right receiver state at the same time.
Where the Automaton Model Fails
The literal analogy fails quickly.
There are no demonstrated gears.
No spring-driven mechanism has been identified.
No evidence shows that the codex was a mechanical performance object.
The court-marvel tradition used as our clean comparator largely post-dates the manuscript.
So the automaton itself leaves the room.
What remains is interaction.
Hostile Controls
To test whether the new variables actually discriminate, compare Voynich with known objects:
- ordinary codices with no foldouts;
- technical manuscripts with functional foldouts;
- maps and astronomical diagrams requiring extension;
- presentation books;
- later mechanical marvels with known activation sequences;
- books designed for repeated teaching demonstrations.
If activation topology and reveal structure are normal consequences of large diagrams, the marvel explanation loses force. If unusually staged interaction survives comparison, the neutral finding becomes interesting.
Metaphor Removal
Delete automaton.
Delete machine.
Delete court marvel.
What survives?
- activation topology;
- operator dependence;
- contextual function multiplicity;
- expectation discontinuity;
- revisit yield;
- latent mechanism versus visible behaviour;
- audience coupling;
- physical-state × receiver-state interaction.
World Return
The imaginary gears stop.
The manuscript is again only a manuscript.
But “only” now means something richer.
A codex is an object that changes state when a human interacts with it.
Its foldouts, spreads, sequences and comparisons can reveal different structures at different moments.
And its historical function may have depended on a trained operator, not merely a passive reader.
That is enough to change the next experiment.
Next: Tangential Voynich | Treat the Manuscript as a Nested Object — Russian-doll logic, recursive reveal and the possibility that the same grammar repeats at several scales.