For my master thesis in Archival Science, I have arranged and catalogued the archives of the Uppsala InternationalShort Film Festival. The materials consisted of approximately 160 binders and filing boxes and some banners.Additionally, there was a large amount of photos from screened films. Some of the material was in good order whilesome was not. The archive was catalogued in Visual Arkiv. After cataloguing the material, approximately 30 binderswith routine records such as reciepts, tax forms et cetera were removed and destroyed. The archive is completelyopen for the public with no limitations in accessUppsala International Short Film Festival started in 1982 and has arranged annual festivals eachautumn. It is usually approximately one week long with 130-300 participating films. Along with the screenings,lectures and debates with the participating directors are also arranged. A jury appoints the best films in severalcategories. The directors of the winning films are awarded a statuette.

Personal archives are a largely neglected subject in archival theory. Among the consequences of this is the absenceof any general established or formalized practices when it comes to arrangement and description. Thisstudy opposes the notion that an archival institution without formal systems of arrangement and description doesnot order and describe archives in accordance with a general conceptual framework of a correct order. Supportingitself on the anthropological theory of Mary Douglas on dirt and cleanliness, it studies the implicit notions oforder that can be found in the archival descriptions of the personal archives kept by the Swedish national archives.Several patterns where revealed by the study: among them the primacy of quantity stood out. The vaguesystem of categories based on Martin Grass description model appears only to be employed loyally where quantitiesof the particular categories are such that they constitute complete volumes, which seem to be the cardinaljustifier of categorical division within the archive. The model itself is rarely applied with orthodoxy, and itsroughness often means that the categories engage in “border clashes” over which documents belong in whichcategory. Though these are often caused and generally determined by quantities, they also reveal an internalhierarchy of relations between specific types of records and categories. Furthermore, certain categories appearsmore stable than others, and when Grass system collapses, it reveals how certain of them – based on a principleof pertinence or theme – habitually fall out of use whereas categories defined by document types appear to remainfar more stable even in smaller archives and archives with low degree of differentiation.