THE SCAM AS AN ANALYTICAL CATEGORY
Documentary war images, platform logic, and the practice of pointing out
TALKS
SYNOPSIS
An image circulates on TikTok: a before-and-after montage of one man, set to music, shared and liked millions of times. I call it “the Ukrainian soldier” — a name of my own, not the platform’s. Whether it shows what it seems to show cannot be established. The account is anonymous, re-encoding has stripped the file of its metadata, and the chain back to a source cannot be reconstructed.
What happens to the authenticity of a documentary war image once it can neither be verified nor disproved? I read images like this through a single concept: the scam. Not an individual deception — a lie that someone could step forward and expose — but a setting, one in which undecidability itself has become the structural condition. The doubt is not new. Frank Hurley built single war scenes from several negatives in 1917; Robert Capa’s “Falling Soldier” of 1936 may have been staged. The two cases are not the same — one a composite, the other a possible mise-en-scène — but both unsettle the photograph’s claim to have simply recorded what happened. What is new is that the index — the physical trace that once tied a photograph to the world it came from — is, under platform conditions, structurally gone. The platform erases the very conditions under which authenticity could be tested.
How does one work with an image like that? Not by papering over the gap, and not by pretending to close it. I argue for a precautionary stance: what cannot be verified or falsified is not treated as settled, and the gap becomes the starting point of the inquiry rather than something to hide. Collaging offers one way to practise this — a pointing-out that shifts attention from the content of a montage to the setting in which it travels, and to the access that setting quietly denies.
This draws on my chapter “Der Scam als analytische Kategorie,” forthcoming in Zukunft des Vertrauens: Trust by Design (Innsbruck University Press, 2026, open access).