The proposition that a painting should be seen in the room it was made for is a defensible one that has been asked to carry more than it can. In its modest formulation it maintains that scale, viewing distance, and the direction and colour of the illumination were components of what a painter worked with, and that a picture displaced from its setting has lost information. That is manifestly true of an altarpiece designed to be read from twelve metres away by candlelight, and it explains why the identical object can appear coarse in a gallery and exact in a chapel.
In its immodest formulation the proposition becomes an assertion about authenticity: that the original setting is the only place the work can be properly seen, and that a museum installation is at best a compromise and at worst a falsification. Presented that way it encounters a difficulty, which is that the original setting is not available either. The chapel has been reglazed, the candles have disappeared, the wall behind has been repainted twice, and the congregation that comprehended what the picture was about has been superseded by visitors who do not. What is on offer is not the original setting but a subsequent condition of the same room, and there is no principled justification for preferring the room's subsequent condition to the museum's.
There is an additional difficulty concerning which room counts. A great many pictures were relocated within a decade of being completed, sold, truncated, extended, and rehung in circumstances that the painter neither anticipated nor would have sanctioned. For such a work the conception of the setting identifies no particular place, and asking where it belongs is a question with several equally satisfactory answers and no procedure for choosing between them.
The modest formulation withstands both objections, and it is worth retaining, because it tells a curator something instructive: reconstruct what can be reconstructed -- the height, the angle, the quality of the illumination -- and be explicit about what cannot. A picture suspended at its original height under illumination of the original colour constitutes a better piece of evidence than the identical picture at eye level under a spotlight, and no claim about authenticity is necessary to say so. The argument for reconstruction is that it retrieves information, which is a claim that can be tested against the object, and not that it restores a vanished authenticity, which is a claim that cannot. Furthermore, it allocates the burden appropriately. A curator who reports that the picture hung nine metres up and is exhibited here at three has communicated something verifiable to the visitor and left the inference to her. A curator who maintains that this is how the picture was intended to be seen has made a claim about a state of affairs nobody can now scrutinise, and has inscribed it on a wall label, where it cannot be contested.
