Thesis: A representation can be almost visually unrelated to the function it serves. Money makes this obvious. The lesson matters for Voynich because the manuscript may preserve a representation while the system that made the representation useful has disappeared.
This article belongs to the explanatory Tangential Voynich series. Start with The Representation Trap, then use this article as a proof-of-principle: appearance and systemic function can be separated almost completely.
A Fifty-Dollar Note Is Not Fifty Dollars
Put a banknote on a table.
It has colour. Texture. Numbers. Portraits. Signatures. Security devices. Serial numbers. It can be measured, photographed, chemically analysed and catalogued.
But none of those visible properties is the main thing money does.
Money can function as a transferable claim, a unit of account, a medium of exchange, a store of purchasing capability, a settlement mechanism and a coordination layer supported by institutional trust. The physical note is a portable representation inside that wider system.
Money looks like paper. Its function lives in a network of expectations.
A person can understand every image printed on the note and still fail to understand money.
The Archaeologist Who Finds the Note but Loses the Economy
Imagine that civilisation disappears and, centuries later, archaeologists recover one banknote.
The future scholar might identify the portrait, recognise a national emblem, decode the date, distinguish the inks, map the decorative geometry and even reconstruct the printing technology.
All of that would be real knowledge.
Yet the scholar could still miss the dominant operational fact: the object once allowed people to transfer claims within a trusted accounting system.
The representation survives. The receiver architecture does not.
That thought experiment is valuable because it resembles the epistemic position of Voynich research. We possess the object. We possess its marks. We possess much of its material history. But we may not possess the cultural procedure that told a fifteenth-century user what to do with those marks.
Meaning Can Be Outside the Object
A banknote does not carry its entire monetary system printed on its surface. Its use depends on institutions, law, convention, confidence, accounting and the expectation that other people will accept it.
VISIBLE NOTE + INVISIBLE SOCIAL SYSTEM = USABLE MONEY
The same is true of many representations.
A map requires conventions. A score requires musical literacy and an instrument. A mathematical expression requires rules of notation. A circuit diagram requires knowledge of electrical components. A legal symbol requires a jurisdiction. A computer protocol requires an implementation.
The marks do not act alone.
A Map Is Useful Because It Is Wrong
Consider the familiar metro map.
A good transport map may distort distance, angle and geography. Two stations that are close on the page can be far apart on the ground. Rivers may bend unnaturally. Lines may be forced into clean angles. Dense central areas are expanded while peripheral distances are compressed.
This is not a defect.
The representation throws away geographic truth so that network truth becomes easier to use.
REAL CITY → SELECT ROUTING RELATIONSHIPS → DISCARD MOST PHYSICAL DETAIL → TRANSIT MAP
If a future archaeologist mistakes the map for a scale drawing of the city, better measurement can produce worse interpretation.
That is the critical Voynich lesson.
A Musical Score Does Not Resemble Sound
Black marks on five horizontal lines do not sound like an orchestra.
The score extracts selected relationships: pitch, duration, timing, dynamics, articulation and organisation. A trained receiver transforms those marks into coordinated action.
Now imagine recovering a score in a civilisation that no longer remembers music.
A researcher could measure symbol frequencies, spacing, repeated motifs and page layout without ever guessing that the marks were instructions for producing structured sound.
Even worse, the researcher might reasonably classify repeated dark shapes as decorative marks or count staff lines as a table.
The problem is not stupidity. It is the loss of the decoder-plus-practice.
Algebra Does Not Look Like Motion
An equation can describe a falling body without looking like a falling body.
A differential equation can represent population growth without resembling organisms. A matrix can represent a transformation without resembling movement. A probability distribution can represent uncertainty without resembling doubt.
Mathematics works partly because representation is liberated from appearance.
This is why visual resemblance is such a weak universal principle. Advanced symbolic systems often become powerful precisely when they stop looking like their referents.
Computer Code Does Not Look Like the Program
Source code can generate a three-dimensional game, bank transaction, medical device interface or weather simulation while visually resembling none of them.
Even the same visible code may behave differently depending on the runtime environment, libraries, permissions, data and machine state.
Meaning is distributed across representation and execution context.
SOURCE + RUNTIME + DEPENDENCIES + STATE = BEHAVIOUR
This does not imply that Voynich is code. The analogy does a different job. It demonstrates that a representation may not reveal its function by resemblance and may remain underdetermined without its operating environment.
A Recipe Is More Than Ingredient Names
Even a familiar historical genre demonstrates the problem.
A recipe can omit actions considered obvious to its intended practitioner. It may name a material whose preparation method is assumed. It may use local measures. It may encode timing through custom rather than explicit notation. It may rely on tools the reader already knows.
Two people can possess the same written recipe while only one possesses the practice required to reproduce the result.
A historical manuscript may therefore be less self-contained than modern readers assume.
The Voynich Error We Are Trying to Prevent
The dangerous inference is:
THIS LOOKS LIKE X → THEREFORE IT REPRESENTS X → THEREFORE ITS FUNCTION IS X
Tangential Voynich separates those arrows.
APPEARANCE ? REPRESENTATION ? FUNCTION
Each bridge requires evidence.
A plant-like image can be classified morphologically without deciding that it represents a biological species. A circular diagram can be measured geometrically without deciding that it is astronomy. A repeated string can be treated as a recurring graphical unit without deciding that it is a word.
What If the Representation Was Designed for Action?
Representations often exist not to depict but to operate.
A checklist directs action. A map supports navigation. A ledger enables reconciliation. A score coordinates performance. A table supports lookup. A diagram can support calculation. A legal form changes institutional state. A tally supports accounting. A mnemonic cue reactivates memory.
If Voynich contains operational representations, asking “what does this picture depict?” may miss the more productive question:
What could this representation allow a trained receiver to do?
That is not permission to invent functions. It is permission to test functional hypotheses separately from visual identity.
Representation Loss Has Two Directions
First, the historical maker compressed reality into representation.
WORLD → SELECTION → TOKEN → REPRESENTATION
Second, history compressed the representation further.
REPRESENTATION → COPYING → DAMAGE → REBINDING → LOST CONTEXT → DIGITISATION → TRANSCRIPTION
We are therefore not reversing one transformation. We are trying to reverse a chain of transformations, several of which may be lossy.
This is why high confidence should be difficult.
The Receiver Reconstruction Problem
Instead of asking only what each form “means,” we can ask what knowledge the receiver must have possessed for the page to be useful.
- Did the receiver need a spoken language?
- Did the receiver need a professional procedure?
- Did the receiver need a memorised key?
- Did the receiver need an external instrument?
- Did the receiver need other books?
- Did the receiver need to know a sequence that the page only cued?
- Did the receiver use the page for lookup rather than continuous reading?
- Were some visual elements indexes rather than depictions?
- Did different page families require different receiver skills?
These questions turn “meaning” into an architecture problem.
Why Tangential Systems Help
Money gives us claims without resemblance. Maps give us network topology without scale. Music gives us structured transformation without lexical semantics. Code gives us stateful behaviour without visual similarity to output. Law gives us symbols whose force comes from institutions. Mathematics gives us abstraction whose power increases as appearance recedes.
These systems teach one shared lesson:
Representation is a designed interface between reality and a receiver, not a miniature copy of reality.
Once that sentence is accepted, Tangential Voynich stops looking eccentric.
It becomes a control against the assumption that a lost representation must resemble the thing it once helped somebody understand or do.
World Return
Money does not look like money.
A map does not look like a city.
A score does not look like sound.
An equation does not look like a physical process.
A program does not look like its execution.
Therefore the Voynich Manuscript does not owe us visual resemblance between representation and function.
That does not mean resemblance is useless. It means resemblance must be demoted from conclusion to candidate signal.
The deeper research question is not merely:
What does this look like?
It is:
What relationship between world, extraction, token, representation and receiver would make this object useful?
That question is harder.
It is also much closer to the problem we actually have.
Research routes: The Representation Trap · Information & Representation Hub · The Tangential Lens · Voynich Research Library