Undelete-files enable us to snatch back from Orcus that which was once destroyed, and even if we staunchly insist on our intention to destroy in spite of all the questions, there are still possibilities: in an 'unerase' feature, Norton Disk Utilities allow us to restore files which have formerly been destroyed. The resurrected files are then re-released into the system as information zombies with details on the extent of their disability and they then live on as living corpses.
Destroying computer data appears to be an atrocity. While the theory of programming languages is taking increasingly more sophisticated pains to store and retrieve quantitatively exploding amounts of data, there are hardly any strategies for meaningfully destroying data.
One of the most interesting programs on data destruction is provided by the Diskwasher program for computers from the VAX series. The Diskwasher is a program which the system administrator can use on a multi-user system if the users are threatening to overload the hard disk by storing large amounts of data. The program usually runs fully-automatically at set times to determine whether a memory load beyond that of a dangerous margin has been reached, and if so to look out for destructible data. At the EMS computer centre in Stockholm, the Diskwasher became activated independently late at night in order to be able to present the despairing users with emptied disk sectors the following morning. The difficulty in programming such a program is having to determine which data is most likely important and which in all probability can be deleted. For this the Diskwasher uses a series of criteria which will subsequently be introduced. (The attempt shall later be made to transfer the logic of the Diskwasher programs to the field of museology - indeed to the entire cultural sphere.)
1. The larger the object the more useless it is.
2. The older the object the more useless it is.
3. Should there be two objects with the same name, one of them is useless.
4. The higher the privilege-status of the object-creator, the more undeletable is the object.
If we consider the 4 criteria of the destructibility (or at least the dispensability) of data in the view of a computer specialist, they appear to be beyond any doubt: why should a document with an older date be kept to the disadvantage of one with a more recent date? It is obvious that every new document is superior to one of older origin. Why else should the more recent document have been created? In the logic of the technical sciences, a version with a higher running number has been created because the older version was worthy of improvement. If that were not the case, no new version would be created. In the field of marketing, a program with a higher version number proves its superiority and maturity over those of older version numbers. If this logic is transferred to the field of artistic production we soon run into value patterns which stand in the way of the immediately plausible technical logic: why has a first work such a high value on the art market? No technician would consider the 'alpha' version, indeed not even the 'beta' version, as worthy of documentation. The art market, on the other hand, considers the 'earliest Dürer sketch' to be less shameful - in fact, it is worthy of conservation.
If the criteria canon of the Diskwasher program is transferred to museology, the following imperatives result and these should be discussed within the context of art-historical conservation:
1. The larger the object the more useless it is.
Undoubtedly a useful imperative. The collections of the large museums are going to ruin due to their lack of space. Costly deep stores - such as the one belonging to the Museum of Applied Art in Vienna - devour vast sums of money which could be used for the purchasing of small, manageable works. Rubens' paintings, with their high demand in square metres, could be exchanged for the space-saving miniatures of a certain Gerwald Rockenschaub.
2. The older the object the more useless it is.
Not even a museum of modern art, which should really be committed to the criteria of the new, completely understands what is meant by this criterion: a Kandinsky from 1910 is lacking the new element which a Julian Opie from 1987 can provide.
3. Should there be two objects of the same name, one of them is useless.
Redundant panels with the description 'still-life' or 'reclining nude' are to be destroyed in good time. It completely suffices to retain one of these isonomic works.
4. The higher the privilege-status of the object-creator, the more undeletable is the object.
Finally a criterion which is realised in full in the practice of museums: the privilege-status of the painter is decisive for the importance placed on retaining a work, and not the quality of the work. No museum, no gallery and no private collection would part with a third-class Baselitz, with an unsuccessful Turner or with a Van Gogh produced in a frenzy. 'The name' is important - or to put it another way: the privilege-status of the creator.
translation Ann Thursfield