Devil's Advocate: The book of the dead

Multiple, overlapping and inconsistent databases are a problem. But there is one more, says Martin Brampton, all of us could really benefit from.

The government is now being quizzed over its plan for a deaths database, in the context of arguments over rationalisation. It's a pity, since it seems to me that a database of dead people will make a quite wonderful project.

One of the headaches faced by developers of databases of personal information is that bereaved families can get very upset at receiving mail for their departed relative. Once or twice is bad enough but organisations that have many databases get into difficulties through repeatedly writing to deceased contacts. Government is obviously in the position of having many databases.

In the case of a database of dead people, though, that problem does not arise at all. If letters are ever written using the database, presumably they would have to be addressed to 'Mr John Smith, deceased', or perhaps to 'the Executors of Mr John Smith'. Nothing there to cause offence and in fact there would rarely be any requirement at all to prune the database. By and large, dead people retain that status indefinitely.

The opposite error could be made, so that a still living person found their way onto the database of dead people. But that is much less of a problem, because most people are more amused than offended by receipt of a letter that presumes their decease. The rumours of my death, as Mark Twain said, are greatly exaggerated. Moreover, there is little need to correct such errors, since they will naturally correct themselves given sufficient time.

Local authorities could find the database valuable, especially if things are as described by Will Self in his brilliant short story 'The North London Book of the Dead'. In Self's world, people do not actually die, they merely move to another part of London.

Once embarked on projects of this kind, all kinds of new possibilities open up. We could do with a comprehensive database of all fictional people, along with an extensive bunch of attributes. Philosophers have sometimes puzzled over whether a statement about a fictitious person can be true or not. If a person does not exist, how can a statement about them be true? But surely that need not prevent us building databases of them.

Truth being allegedly stranger than fiction, could we devise a database of people yet to be born? It seems to present some quite serious challenges but as a research project it has enormous potential. There is no doubt it would be useful. There are many situations, especially in government, where there is a need to project the future as well as monitor the present. Changes would be needed in the procedure for registering births but they could surely be managed.

Another useful database would be all the people who do not live here. It would be very convenient if it identified people as belonging to well known types such as spies, terrorists or social security scroungers. This would save a great deal of time and trouble at points of entry into the country. Recent legal battles suggest that perhaps the Home Secretary already has something of the kind.

The ultimate database is the Library of Babel. It consists not merely of every book ever written but of every book that could have been written. As a result it contains numerous true accounts of the whole of your life, accompanied by arbitrarily many false accounts. It provides references to back up any hypothesis you care to think of, as well as accounts to undermine them all. But despite the best efforts of Intel, Microsoft et al it has so far proved beyond the power of current technology.