Documentation As Meta-Level Activity

Abstract

Today we live in an information society, which is to a large extent also a document society (Buckland, 2018). In our daily lives, we must cope with an ever-growing flood of different documents, keep track of them, pick out relevant content, and produce documents in suitable forms. In addition, tools based on artificial intelligence (AI) technology, such as ChatGPT[1], are making their way into the document world, challenging us with new affordances, and questions about how to deal with them and what changes this will bring. Existing documentation models such as the model of complementarity (N. W. Lund, 2004), the ontology of human expression (Olsen et al., 2012), and the document phenomenology (Gorichanaz and Latham, 2016) contain valuable indications from which we can learn a lot about the process of document production. However, current models are missing out on the relations between the variables of the greater system in which they are embedded such as documents as instruments, the community, rules, and the division of labor, as well as the subject's actions, motives, and greater goals in the documentation activity. This conceptual paper presents the model of documentation activity, which on the one hand can unite and expand existing models in the field of document theory, and on the other hand provide a common crystallization point for practitioners and scholars from related fields to analyze the documentation activity within a dynamic socio-technical system. The model of documentation activity is developed along existing document theory concepts using activity theory (AT) (Engeström, 1987) as a framework. [1] ChatGPT is a system for natural language processing (NLP) based on deep learning technology which was developed by the company OpenAI. ChatGPT can generate human-like conversations by understanding the context of a conversation and generating appropriate responses. (Deng and Lin, 2023).