The term ``information" is used differently by individuals in different walks of life, from specialists working in information based professions, such as communication media and information management, to those in the computing and cognitive sciences, as well as by people involved in less scholarly pursuits. For example, communication scholars express concern about information overload due to the plethora of mass media sources. Electrical engineers strive to improve modem technology to increase the amount and speed of information transmitted, i.e., the number of bits per second transmitted over coaxial or fiber optic circuits. An interest in understanding information as well as the ability to measure it allows these specialists and others to work on their chosen problems, make reasonable decisions, and to communicate effectively among themselves.
Many have developed interesting and useful definitions of information for specific disciplines and classes of problems. Some definitions or measures may be consistent with ideas held by several disciplines. These interdisciplinary definitions are superior in some senses to discipline specific definitions useful in only one domain when the more general definition encompasses all the phenomena of interest to the field that is covered by the field-specific definition and is consistent with the field-specific definition. A more general definition allows frameworks, theories, and results to be transferred across disciplinary boundaries, and provides for dialogue across these boundaries, while at the same time allowing individual disciplines to focus on the specific information phenomena of their discipline. Unfortunately, people in different fields and professions differ on what information is or how to evaluate the different definitions that are assumed explicitly or implicitly by different fields or social groups.
We suggest that there are phenomenon common to what most definitions of ``information" refer and that this phenomenon is information and that most definitions of information refer only to the subset of information as studied in that particular discipline. Information may be understood in a domain-independent way as the values within the outcome of any process [Los90]. By ``value" we refer to a variable's attribute or characteristic, and not to economic value unless economics is explicitly mentioned [HR92]. Following Russell [Rus37], we view a variable as a component in a system whose ``value" may be replaced by another value with the system remaining the same type of system as before.
This definition of information has the fortunate consequence of allowing different kinds of processes to be information carrying or information producing phenomenon. These processes may be viewed as ranging from the simple, mathematical function, to processes of such complexity, often involving humans, that some may believe they can't be studied or accurately characterized. The ability or inability to study the nature of a process will not affect our study of the value in the output of the process, and our work may be applied to both processes that are formally defined and to those that are viewed as non-formal in nature.
Viewing the processes together, in an organized fashion, facilitates the engineer, the linguist, and the television producer speaking in the same terms about how what they produce (the output of a process) ``carries" information or is informative. A common language allows for theories and data to be transferred from discipline to discipline without confusion; term and concept migration are common as academic fields change and grow [LW86,Los95,Mul74,NCB72]. Considering how information and informativeness can be viewed consistently across different environments is the focus of this work. We provide examples of ways that different disciplines might wish to describe some of their information-phenomena of interest in terms of processes, making them compatible with our approach to information; these are suggestions, and other models also may be useful in integrating other discipline based information-phenomena into the proposed general definition of information. This general definition has the capability to model information as understood by physicists and those studying human behavior. It also has the advantage of being able to formally capture the subjective and individual processing and culture-specific nature of human processes and knowledge that affect information.
We propose that data from a variety of disciplines suggest and is consistent with this field-independent conceptual framework for information. Our definition thus is a generalization of the work of discipline-specific concepts of information provided by scholars such as Buckland [Buc91], Dretske [Dre81], and Shannon [SW49]. While we believe that our definition is consistent with the problems and the data found in a wide range of fields in both the humanities, social sciences, and ``hard" sciences, our definition will conflict with other definitions that limit the domain of a field's study of information; this conflict arises between any general definition and more specific and limiting definitions. We are not saying that these field-specific definitions are wrong for the field in which they are defined; instead, we suggest a more general definition that can be applied to a broad range of fields and can facilitate communication about information-related phenomena.
Note that we do not consider ``errors" or ``misinformation" in our discussion of information until the penultimate section. The reader should assume processes are error-free until that point, unless otherwise stated.