Appearance
To navigate the ethics of digital reconstruction, your guiding rule is intellectual transparency: never let a render imply more certainty than the evidence supports. In practice that means grading every element by how well it is attested, encoding that uncertainty visually, and publishing the reasoning behind each decision as paradata. The London Charter and the Seville Principles give you the framework; this guide turns them into a step-by-step workflow with sensible defaults and the pitfalls that trip people up.
What is the core ethical problem, and which frameworks address it?
The core problem is simple to state and easy to violate: a polished 3D model of a vanished temple looks equally convincing whether a wall is documented by excavation or invented to fill a gap. Photorealism launders conjecture into apparent fact. Two frameworks exist precisely to counter this:
- The London Charter (2009) — principles for computer-based visualisation of cultural heritage: intellectual transparency, documentation, sustainability, and access.
- The Seville Principles (2011) — applies the Charter to virtual archaeology, adding authenticity, scientific transparency, and historical rigour.
Treat them as a checklist, not decoration. If you cannot show how you complied, you probably did not.
How do I grade evidence and assign certainty?
Start by classifying every reconstructed element before you model it. A simple three-band scheme works for most projects and maps cleanly onto a visual convention later.
text
Tier A attested - survives, excavated, or in a measured drawing
Tier B inferred - reasoned from comparanda, typology, or a written source
Tier C conjectural - plausible but unevidenced; needed to complete the formRecord the tier as a property on each object in your model so the data, not just your memory, carries the certainty grade. This single discipline prevents the most common failure: a beautiful model whose author can no longer say which parts were real.
How do I show uncertainty visually?
Pick conventions before you texture anything, and keep them consistent across the whole model. Sensible defaults:
| Certainty tier | Default treatment | Rationale |
|---|---|---|
| Attested (A) | Full colour and texture | Documented, so render faithfully |
| Inferred (B) | Desaturated or lower-opacity texture | Signals reasoned but unproven |
| Conjectural (C) | Greyscale, wireframe, or 40% transparency | Clearly reads as hypothesis |
Offer a viewer toggle that switches between the photoreal view and the uncertainty-coded view. The photoreal version is fine for public outreach provided the coded version is one click away and the caption says so.
What documentation do I have to produce?
Paradata. This is the non-negotiable scholarly output, and skipping it is the deepest pitfall in the field. Your paradata should let another researcher rebuild the same model from the same evidence. A minimal structure:
yaml
# paradata.yaml
element: "north_colonnade"
certainty: "B-inferred"
sources:
- "Excavation report, Trench 4, 1998, pp.112-119"
- "Comparandum: temple at site X, column spacing"
assumptions:
- "Column height extrapolated from surviving base diameter (1:8 ratio)"
decided_by: "E. Reed"
date: "2025-05-02"Without paradata you have an artist's impression; with it you have a citable scholarly argument.
When is a reconstruction ethically inappropriate to make at all?
Technical possibility is not permission. Sacred sites, burial places, and contested heritage carry obligations that override your curiosity:
- Community consent — some sites may not be represented, or only under specific conditions, by people outside the associated community. Seek consent before, not after.
- Contested narratives — reconstructing a destroyed monument can take sides in a live political dispute. Acknowledge the contestation rather than presenting a single "true" past.
- Restricted knowledge — certain cultural knowledge is not meant for general viewing; a reconstruction can violate that even when sources are public.
The default here is to ask first and to be ready to not build the model at all.
How do I make the reconstruction reproducible and durable?
Version the model, its source data, and its paradata together, and deposit them in a repository with a persistent identifier so the work is citable and revisable. Use open or well-documented formats (glTF, OBJ, plus the original project files) and record software versions, because a reconstruction is a hypothesis that new evidence may overturn. A reconstruction you cannot revise or re-examine in ten years has failed the Charter's sustainability principle no matter how good it looked at launch.
Key Takeaways
- The central duty is intellectual transparency: never imply more certainty than the evidence allows.
- Use the London Charter and Seville Principles as a working checklist.
- Grade every element (attested, inferred, conjectural) before modelling and store the grade on the object.
- Encode uncertainty visually and offer a toggle between photoreal and coded views.
- Publish paradata documenting sources, assumptions, and decisions for each element.
- Seek community consent for sacred or contested heritage; technical ability is not permission.
- Version and archive model, data, and paradata so the reconstruction is reproducible and revisable.
Frequently Asked Questions
What is the main ethical risk in digital reconstruction?
Presenting conjecture as fact. A photoreal render of a lost building can make an educated guess look like documented certainty, so the central duty is to communicate uncertainty clearly to every viewer.
What are the London Charter and Seville Principles?
They are the two recognised frameworks for computer-based visualisation of heritage. The London Charter (2009) sets out principles like intellectual transparency and documentation; the Seville Principles (2011) apply them specifically to virtual archaeology.
How do I show uncertainty in a 3D model?
Use visual conventions such as transparency, greyscale, or colour-coding for conjectural elements, and publish a paradata document explaining the evidence and reasoning behind each decision.
What is paradata?
Paradata is the record of the interpretive process behind a reconstruction: the sources used, the assumptions made, and why each visual decision was taken. It is what separates a scholarly reconstruction from an artist's impression.
Should I reconstruct sacred or contested heritage?
Only with the consent and involvement of the associated community. Some sites carry restrictions on who may view or represent them, and a technically possible reconstruction can still be ethically inappropriate.
Do reconstructions need version control and archiving?
Yes. Models, source data, and paradata should be versioned and deposited in a repository so the reconstruction is reproducible, citable, and revisable as new evidence emerges.