The Encyclopedia Gallica of Events - or Why Geographic Information Science Is Not Like Physics

Abstract

GIScience 2016 Short Paper Proceedings The Encyclopedia Gallica of Events (or Why Geographic Information Science is Not Like Physics) H. Couclelis University of California, Santa Barbara Santa Barbara, CA 93106-4060, USA Email: cook@geog.ucsb.edu Abstract As a contribution towards the further development of user-centered information systems, I present an argument for a contextual and subjective view of events and related concepts in information science that is distinct from the factual view prevalent in empirical science and everyday life. The central notion is that of ‘R-Event’, where ‘R’ stands for ‘Relevant’. Drawing on representations of process such as dynamic models or AI approaches to problem-solving and planning, R-Events would improve the precision and value of search results by foregrounding and ranking information about events of high relevance to the user. There is no suggestion here that this proposal is easily implementable. It is offered at this stage as a potentially fruitful thought experiment. 1. Introduction The title is one that Borges would have loved 1 . Had he written the story of the Encyclopedia Gallica of Events, this would have also included processes, perdurants, occurrences, and non-events. And the Encyclopedia in question would have had to be infinite in length, and self-contradictory. I will argue that in an informational (as opposed to an empirical) context, (a), the above concepts do not refer to intersubjectively definable individuals and (b), unlike most other abstract concepts, the things these refer to are not free-standing notions but binary relations connecting information source and observer, and depending on the specific interest the latter has in certain detectable state(s) or change(s) in the world. Thus the news of my birth signaled a momentous event for my family (and a very real painful process for my mother) – but for the rest of humanity the fact of my birth was a non-event, barely registering as a blip of plus-one added to some country’s population. Similarly, major flooding in south-eastern Tajikistan may be an empirical fact, but it is an event if you are in the business of importing cotton from that area, a process if you are in charge of evacuating local populations, an occurrence if you are tabulating floods in Central Asia, and noise if south-eastern Tajikistan is not among the things you care about. Similar kinds of context-dependency also hold for the endurant-perdurant distinction, and as has been argued repeatedly over the years, also in the case of fields and objects. Information systems sensitive to user-centered semantics should be able to make these determinations. 2. On Science and Metascience It may be useful to distinguish clearly between the empirical sciences that directly measure and represent phenomena in the world, and the information sciences (which are meta-sciences) that process and present information about these phenomena in ways that meet and support the interests and purposes of information users. These are really two different epistemic layers, with different See Borges (1939) The Total Library; Borges (1943) The Library of Babel; and the “certain Chinese Encyclopedia” in Borges (1952) The Analytical Language of John Wilkins.