There is a particular kind of email that arrives every few months from a well-meaning developer, usually a graduate student or an early-career studio, asking whether the photogrammetric mesh from a destroyed monument could be turned into a "virtual reconstruction" — a Unity walk-through, a VR experience, a digital twin. The implied subject of the email is always "I have the technical skill to do this." The unspoken object is "I want permission."

I am not going to dismiss those emails. I have answered some of them at length. But I want to set down, in one place, the framework that the answer needs to pass through, because the question itself is harder than the engineering, and the engineering is not easy.

The minaret at the Umayyad Mosque

In April 2013, during the battle for the old city of Aleppo, the eleventh-century Seljuk minaret of the Great Mosque collapsed. Each side blamed the other. The minaret had stood for nine hundred years.

There is, fortunately and unfortunately, a comparatively complete photogrammetric record. The German Archaeological Institute had run an extensive photogrammetric campaign on Aleppo's old city between 2007 and 2011, the CyArk archive holds a partial laser scan, and there is enough community photography in Wikimedia Commons to push a SfM reconstruction to roughly 4 cm fidelity on the upper shaft. With the right pipeline you could rebuild the minaret as a virtual object that is, geometrically, indistinguishable from photographs of the original at almost any focal length.

You could. Should you?

The framework I keep coming back to

I have been writing about this since 2019 and the framework hasn't simplified, only sharpened. It is essentially four questions, asked in order, and the project does not begin until each of them has been answered by someone who actually has authority to answer.

1. Whose heritage is it?

This is the first question and it is the one most often skipped. The minaret was, formally, the property of the Syrian Ministry of Awqaf, but its meaning was held by the people of Aleppo, the broader Syrian Sunni Muslim tradition, the Seljuk historiographic record, and arguably the entire community of medieval Islamic architectural history. None of those constituencies maps cleanly onto an institutional review board in Berlin or Sydney.

The answer is rarely a single body. The closest current model is the consultation framework that the Syrian Heritage Archive Project uses, which puts Syrian heritage professionals at the centre of decisions about Syrian material. That is not perfect — many of those professionals are themselves in exile, and the framework has nothing useful to say about objects whose meaning is contested between communities — but it is the floor, not the ceiling, of what a serious project needs.

2. What kind of object is the reconstruction?

A photogrammetric mesh, a CAD model, a polygonal walk-through, an immersive VR experience and a generative-AI "completion" of missing geometry are five different ontological objects, even if the source data is the same. Each makes a different epistemic claim on the viewer.

A photogrammetric mesh of what survives says: "here is what we measured."

A CAD reconstruction labelled with confidence intervals at every junction says: "here is what we measured, and here is our best architectural inference about what is no longer there, and here is how sure we are at each point."

A polished VR walk-through says, whether the project intends it or not: "here is the building." It is a claim about the original, dressed up as a measurement, and it is the form that gets used in newspaper coverage and tourism marketing and the long memory of the public record.

The first two are publishable in almost any circumstance. The third should require a much higher bar of consultation, because once it is out it cannot be recalled. Palmyra's Triumphal Arch replica — installed in Trafalgar Square in 2016 — is the case study I keep going back to. The replica is technically superb. It is also, as several Syrian scholars argued at the time, a kind of cultural ventriloquism: the original speaks through a copy commissioned by institutions in the country that drew the Sykes–Picot line through Syria a century before.

3. What is the reconstruction for?

There are four legitimate purposes I can think of for a digital reconstruction of a destroyed building, and the ranking of authority required is different for each.

Purpose	Authority needed	Notes
Scholarly reference (e.g. a 3D figure in a published paper)	Author + peer review	Lowest bar; the artefact is contextualised in a scholarly argument.
Conservation planning (where would the rebuilt walls go?)	Local heritage authority + community consultation	This is the actual job CyArk was set up to support.
Public memorial (a virtual space where the building's loss can be witnessed)	Affected community + religious/civic authority	Highest bar. The form carries meaning beyond the geometry.
Tourism / entertainment	I am genuinely not sure this is ever the right project.	Discuss before you build.

The first two are routine archaeological work and ought to proceed. The third is fragile and important and almost always needs partners I cannot bring to the project alone. The fourth needs a hard conversation before any geometry is touched.

4. Who pays for it, and who profits?

This is the question that makes a lot of well-intentioned projects unravel when asked. A reconstruction funded by a Gulf state's tourism authority is doing different work than the same geometry funded by a UNESCO emergency-response grant or by the community-driven Project Anqa initiative. The technical output may be identical. The political and economic implications are not.

If you cannot answer "who profits if this is successful, and who is exposed if it goes wrong?" you are not yet ready to commission the model.

What I do, in practice

I do publish photogrammetric records of damaged or destroyed sites, with confidence annotations, when the source data is in my hands and the relevant community-facing partners have signed off on the publication. The Knossos and Pompeii field records on this site are examples; both have institutional sign-off from the local Soprintendenza or Ephorate.

I do not currently work on reconstructions of monuments destroyed in active conflicts in regions whose heritage authorities are themselves displaced or coerced. This is not a universal rule for everyone — I know respected colleagues who have made the opposite call and whose work I trust. It is the rule I have arrived at for myself, and it gets re-examined every year.

I will be moderating a roundtable on this in July with the Syrian Heritage Archive Project at the Museum für Islamische Kunst. If you have something to bring to it, the registration link is on the Museum's site and I am happy to be put in touch.

— Elara