Objectivity, in the form of the application of external scrutiny according to standards agreed within a research community, is an essential characteristic of information science research whether pursued from positivist, interpretative, or action research perspectives. Subjectivity may represent both a legitimate focus of research (e.g., people's perceptions and attitudes) and a legitimate component of methodology (e.g., enabling researchers to enter, experience, and share the perceived worlds of their subjects). However, subjectivity that both (a) is not open to external scrutiny and (b) gives rise to contingent dependencies is problematic for research. The issue of problematic types of subjectivity is considered, and the contributions to the debate concerning possible solutions of two key thinkers-the cybernetician Gordon Pask and the methodological philosopher Brenda Dervin-are discussed. The need identified by Dervin for researchers to be able to escape (expose and test) their own assumptions is explored in terms of a dynamic interplay between relatively subjective and objective forces, each requiring the liberating and constraining energies of the other. The extent to which meta-methodological awareness-a prerequisite for any such escape-can be fostered, for example, by the quality of research environments, is explored along with implications for those responsible for managing and funding research.
IntroductionThe long and strong positivist research tradition in information retrieval and system development aspects of information science, intimately entailing human factors particularly in terms of performance, has failed to develop robust and coherent models of human factors capable of usefully informing system design and development. The more recent interpretative tradition of information behavior research since the 1980s has fared little better. There is also a marked lack of integration between the two traditions-indeed, systemscentered researchers in information science seem in the past to have paid scant attention to the results of human-centered information science research (Ford, 2000;Saracevic, 1996Saracevic, , 1997Spink & Greisdorf, 1997). Arguably central to this problem-and attempts to address it-are the notions of objectivity and subjectivity entailed in different research approaches, discussed below. These notions are also arguably useful in thinking about the constraining and empowering characteristics of different research environments and styles of research management. These issues will be discussed in The Role of Research Environments.An essential defining criterion of research is that it should produce evidence that is open to, and bears, scrutiny. Implicit in the concepts evidence and scrutiny is the notion of objectivity. Objectivity is defined here in the sense of judgment and acceptance according to standards agreed by scrutinizers other than the person conducting the research-typically standards agreed within a research community.A researcher's actions and interpretations must be justifi...