Libraries have long sought to develop comprehensive collections of the scholarly works within the disciplines taught at that university. In so doing, libraries typically rely on publishers and journal editors to provide the critical role of selecting the material worth collecting. For example, acquisitions librarians know well the publishers and journals that are well respected within each scholarly discipline.

Electronic publication disrupts nearly every aspect of this arrangement:

• The volume of material is too great for practical storage.
  
  
• There is no external authority to distinguish quality from junk.
  
  
• There is no clear distinction between ephemera and completed works. In the on-line culture, all web pages are "under construction".
  
  
• The ease of editing on-line material makes it impractical to assign a distinct edition to electronic publication. For example, if you return to this page at a later date, you may find that its content has substantially changed.
  
  
• On-line materials are often based on specific software resources that may not exist in future years, making archival storage a dangerous practice. Most users have collections of old disks whose contents are lost because the computers and software needed to read the disks no longer exist.

The notion of any finite institution maintaining a comprehensive collection of electronic publications is a myth that can never become reality.

Should libraries even try to catalogue everything?

Electronic texts may even challenge our notions of textual authority.