The Production Path Hypothesis creates a necessary freedom.
It allows text ancestry, scribal training, physical production, assembly, binding and ownership to occupy different places.
That freedom solves one problem.
It creates another.
If every contradiction can be assigned to a new city, a new workshop, a travelling scribe, a lost exemplar or a later assembly stage, then the path can explain anything.
The Minimum Path Principle says: use the fewest geographic and production handoffs required by the evidence, and charge every additional handoff a complexity cost.
The Path Must Be Allowed—but Not for Free
Consider three models.
| Model | Path | Complexity |
|---|---|---|
| A | One coordinated workshop in Padua. | Low. |
| B | Paduan source tradition → transalpine-trained scribe working in Padua → later Venetian binding. | Medium. |
| C | Florentine image source → Paduan medical compiler → German scribe → Venetian illustrator → Viennese assembly → Prague ownership. | High. |
Model C is not impossible merely because it is complicated.
But it should not be preferred unless independent evidence actually requires those extra stages.
Parsimony Is Not “Always Choose One Place”
The simplest explanation is not automatically the explanation with the fewest city names.
If the manuscript contains genuinely incompatible production strata, a one-workshop model may become artificially complicated because it needs many exceptions.
A two-stage path may then be simpler in explanatory terms.
So the principle is not:
one city good, many cities bad.
It is:
every additional stage must explain evidence that a simpler path cannot explain without greater strain.
What Counts as a Handoff?
- a text moves from one intellectual centre to another;
- a scribe trained in one region works in another;
- a codicological unit is transported before assembly;
- images are added in a second workshop;
- independent booklets are joined in another place;
- a binding is added or replaced elsewhere;
- the completed manuscript changes owner or institution.
Not every handoff is equally important.
A later ownership transfer may tell us little about original production.
A second physical-production site changes the origin model much more strongly.
The Coordinate Stack Tells Us Which Handoffs Cost More
The Coordinate Stack separates knowledge ancestry, text origin, image ancestry, scribal training, physical production, assembly, binding, ownership and reception.
A new C1 knowledge-ancestry coordinate is relatively cheap historically because texts travel readily.
A new C5 physical-production coordinate is more expensive because it requires the physical manuscript or one of its units to move.
A new C6 assembly coordinate adds another physical event.
A new C8 ownership coordinate may be expected in a long-lived object and therefore carries a different cost.
The path should therefore distinguish informational movement from object movement.
Pal. lat. 1265 Shows Why the Principle Is Needed
Pal. lat. 1265 is catalogued to Padua and southern Germany.
That object tells us that multi-region production is historically real.
But it does not tell us to assign multiple regions to every difficult manuscript.
The comparator establishes permission.
The target manuscript still has to earn the complexity.
A Handoff Must Pay Rent
For every added stage, ask:
- What specific observation requires this handoff?
- What simpler model fails without it?
- What new prediction follows from the handoff?
- What physical or documentary trace should the handoff leave?
- Can the same evidence be explained by a travelling text rather than a travelling codex?
- Can the same evidence be explained by a travelling scribe rather than a second workshop?
If the handoff explains no unique evidence, remove it.
The Explanatory-Strain Test
Complexity is not only the number of stages.
A one-place model can also become complex if it accumulates rescue clauses.
For example:
- all five scribes were foreign;
- all astronomical parameters were imported;
- all image traditions came from elsewhere;
- all non-local codicology is unusual workshop variation;
- all missing local evidence was lost.
That may be geographically one place but explanatorily many exceptions.
A better model may use two coordinates with fewer special excuses.
Minimum path means minimum total explanatory cost, not minimum number of place names.
The Handoff Cost Ledger
| Added element | Typical cost | What can justify it |
|---|---|---|
| Imported text tradition | Low | Known transmission history, copied authorities, shared textual family. |
| Foreign-trained scribe | Medium | Palaeographical bundle plus historical mobility. |
| Separate illustration workshop | Medium–high | Different materials, sequence, style and production evidence. |
| Second physical-production region | High | Codicological unit-level divergence that regional controls explain better. |
| Later assembly elsewhere | High | Independent booklets, sewing, catchwords, offsets, trimming or binding evidence. |
| Later ownership transfer | Low–medium | Documentary provenance, marginalia, inventories. |
Minimum Path and Expected Evidence Work Together
The Expected Evidence Problem asks what each stage predicts.
The Minimum Path Principle asks whether a stage is worth keeping after those predictions are tested.
Suppose we propose a Viennese assembly after Paduan production.
Then we should ask what Viennese assembly would leave:
- binding features;
- quire reorganisation;
- later paratext;
- ownership evidence;
- material interventions.
If none appears, the handoff loses support.
Minimum Path and Evidence Independence Work Together Too
A complicated path should not be justified by five clues that are really one clue repeated five times.
The Evidence Independence Problem therefore runs before path expansion.
Only independent or clearly distinct evidence channels should be allowed to purchase extra path complexity.
A Practical Model Comparison
When comparing localisation models, record four things:
- Coverage: how much of the evidence does the model explain?
- Exceptions: how many special rescue clauses are needed?
- Handoffs: how many information or object transfers are introduced?
- Predictions: how many new observations can distinguish the model from rivals?
The preferred model should not merely fit.
It should fit with fewer unsupported commitments and better predictive consequences.
The eduKate Minimum-Path Test
- Start with the simplest one-coordinate physical-production model.
- List every observation it fails to explain.
- Attempt local variation, travelling texts and travelling scribes before adding a second physical-production region.
- Add one handoff at a time.
- Require each handoff to explain at least one independent unresolved observation.
- Write a prediction for every handoff.
- Remove handoffs whose predictions repeatedly fail.
- Penalise ad hoc rescue clauses as well as extra geography.
- Compare total explanatory cost across rival models.
- Retain the minimum path that survives the evidence.
What This Means for Padua
A single Paduan workshop remains a valid candidate if it can explain the material, scribal and visual evidence without large exception costs.
A Paduan contribution plus a later transalpine stage becomes preferable only if independent evidence demands that extra stage.
A sprawling Padua–Venice–Vienna–Prague chain should not be rewarded simply because every link is historically plausible.
Historical possibility is cheap.
Required history is expensive.
Research Links
- The Production Path Hypothesis
- The Coordinate Stack
- The Expected Evidence Problem
- The Evidence Independence Problem
- How Sensitivity Analysis Works
- Voynich Research Library
The Final Idea
A more complicated history is sometimes the true history.
But complexity must be purchased with evidence.
The best production path is not the one that can explain everything. It is the smallest path that must exist if the surviving evidence is true.