Practical and Philosophical Considerations for Defining Information as Well-formed, Meaningful Data in the Information Sciences

Abstract

This paper demonstrates the practical and philosophical strengths of adopting Luciano Floridi's "general definition of information" (GDI) for use in the information sciences (IS). Many definitions of information have been proposed, but little work has been done to determine which definitions are most coherent or useful. Consequently, doubts have been cast on the necessity and possibility of finding a definition. In response to these doubts, the paper shows how items and events central to IS are adequately described by Floridi's conception of information, and demonstrates how it helps clarify the muddy theoretical framework resulting from the many previous definitions. To this end, it analyzes definitions, popular in IS, that conceive of information as energy, processes, knowledge, and physical objects. The paper finds that each of these definitions produces problematic or counterintuitive implications that the GDI suitably accounts for. It discusses the role of truth in IS, notes why the GDI is preferable to its truth-requiring variant, and ends with comments about the import of such a theory for IS research and practice.

Curating the infosphere

“…Dinneen and Brauner (2015) consider the relevance of PI to LIS by examining its treatment of materials and processes of relevance to LIS, and conclude that the definition of information within PI is more appropriate than that in numerous other theories of information; see also Pleshkevich (2016) for a perspective based in the Russian literature.. Dinnen and Brauner raise one problem which had caused problems for the acceptance of PI within LIS; its insistence on the truthfulness of information. While this may be readily avoided by suggesting, as Floridi does, that what LIS deals with is meaningful data / semantic content, this has posed an apparently psychological problem to some LIS specialists, who dislike the idea that what they deal with is not information.…”

Information-not-thing: further problems with and alternatives to the belief that information is physical

“…Despite qualifying that information-as-thing is a representation of knowledge that cannot be confused with the knowledge itself, Buckland also argues at length that information systems deal directly with physical information, including bits, books, and objects (p.352). In Dinneen and Brauner (2015) we noted two problems with this view: (1) that identical objects can afford non-identical information, and vice versa that (2) nonidentical objects can afford identical information. These problems suggest that physical things cannot be information, and information therefore cannot be a physical thing; here we examine an additional, related problem.…”

“…In Dinneen and Brauner (2015) we argued that all three senses of information discussed in Buckland (1991), information-as-thing, -as-knowledge, and -as-process, which represent many previously given definitions of information, are practically and theoretically inadequate for use in IS. Here we pursue this thesis further by noting another problem entailed by viewing information as physical, and showing how using the type/token distinction avoids this problem while better accounting for examples of information.…”

Information theory for information science: Antecedents, philosophy, and applications

“…For example, Losee's Information from Processes [1] discusses how information is provided by the output characteristics from a process, providing a model of information that is completely consistent with Shannon's measures of information. Other recent philosophical discussions are provided by Floridi [13] and Dinneen and Brauner [14] who describe more semantically oriented views of information.…”