knowledge creation—from inspiration to information, to analysis, synthesis and dissemination. In concluding, I will turn to John Lombardi to advise us on how this new framework and the other thoughts and ideas he has heard here can be best employed in serving principal needs of our universities. He will also advise us on how to present this new vision to senior university administrators and how they might envision supporting our transformation.1 The Nature of the Collection First, a quick review of what the collection was and is. I suggest that the viewpoint commonly expressed regarding the comprehensive research library collection of the recent past fails to incorporate extensive shortcomings. Working as an archivist and special collections administrator for nearly 20 years, I am aware of the extent to which our preserved record of the past is remarkably incomplete and that we have limited knowledge of how and why and to what extent it is incomplete. Sumerian archaic cuneiform script is generally considered the oldest known writing system, beginning in the early Bronze Age, ca. 3100 BC, but to what degree is this knowledge uniquely based on the survival of the clay tablets on which it was recorded? And since some tablets were reused rather than preserved and preservation was restricted to those tablets that were fired, either in kilns or perhaps sometimes when cities were burned by invading armies, what portion of that record do we hold? This paradigm of highly selective preservation can be applied often, and how have regional climates impacted the preservation of document forms? Many of our libraries hold medieval manuscripts produced in Latin by monastic scribes, but do we hold any of the 700,000 “Timbuktu manuscripts,” produced in Arabic script or Africanized versions of Arabic from as early as the 13th century? I know that the Library of Congress does and an increasing number are now available on the web. In the modern day, we are aware that in some places, access to printing presses was restricted, thus shaping the early printed record. We are fully aware of the difference between the record of the conquerors and the conquered, the haves and the have-nots, aboriginal and non-indigenous populations, women and men, gay and straight. And of course, we have almost no record of those numerous societies for whom their principal transfer of history, science, and literature was verbal. RLI 277 2 Rebalancing the Investment in Collections ( C O N T I N U E D ) DECEMBER 2011 RESEARCH LIBRARY ISSUES: A QUARTERLY REPORT FROM ARL, CNI, AND SPARC