Loading field map
Why do people seek, use, and sometimes ignore information? The question sounds simple, but the answers have shifted dramatically over the past seventy years. Early researchers assumed that information was a neutral resource and that behavior was largely a matter of system design. Later frameworks insisted that the user's mind, social world, or political context mattered more. The history of information behavior is a story of successive attempts to locate the real driver of human engagement with information—and each attempt exposed something the previous one had overlooked.
In the 1950s and 1960s, the dominant view treated information as a physical commodity that could be stored, moved, and delivered. The Physical Paradigm assumed that information existed independently of any particular person or situation. A document contained information; a user needed that information; the task of an information system was to match the two as efficiently as possible. Behavior was reduced to a simple act of retrieval: the user formulated a query, the system returned relevant documents, and the transaction was complete. This framework made perfect sense for the era of mainframe computers and large bibliographic databases, where the pressing problem was mechanical access rather than human understanding. Yet it left no room for the possibility that different users might interpret the same document differently, or that the same person might want different things at different times.
The Cranfield Information Retrieval Paradigm, which took shape in the 1960s and dominated through the 1980s, did not challenge the Physical Paradigm's core assumption that information was an objective substance. Instead, it narrowed the focus to a single methodological problem: how to measure system performance. The Cranfield experiments introduced controlled test collections, relevance judgments, and the now-familiar metrics of precision and recall. This was a powerful engineering achievement—it gave system designers a replicable way to compare retrieval algorithms. But the framework's very strength was also its limitation. By treating relevance as a stable property that could be judged in advance by a small set of evaluators, the Cranfield Paradigm pushed the user even further into the background. The user's actual situation, prior knowledge, or evolving goals had no place in the evaluation laboratory. Information behavior became whatever could be captured by a query and a relevance score.
By the 1980s, a growing number of researchers found the system-centered view deeply unsatisfying. The Cognitive Paradigm turned the subfield inside out. Instead of asking how well a system retrieved documents, it asked how people construct meaning from information. The unit of analysis shifted from the document to the individual's mental model. Researchers such as Brenda Dervin and Carol Kuhlthau developed models that described information seeking as a sense-making process or a constructive journey through stages of uncertainty. The Cognitive Paradigm directly rejected the Physical Paradigm's assumption that information was a self-contained object. Information, from this perspective, was whatever resolved a gap in a person's understanding—and that depended on the person, not just the document. The framework brought user studies, interviews, and qualitative methods into the mainstream. Yet its focus on the individual mind also created a blind spot. If every person's information behavior was unique, how could researchers generalize across groups? And what about the social and cultural forces that shape what any individual considers worth knowing?
The Domain-Analytic Approach, which emerged in the 1990s and remains active today, offered a direct answer to the Cognitive Paradigm's limitations. Birger Hjørland and other proponents argued that information behavior cannot be understood by studying isolated individuals. Instead, researchers should examine the knowledge domains—scientific disciplines, professional fields, or cultural communities—that shape what counts as relevant information in the first place. A physicist and a historian do not just search differently; they operate within different epistemological frameworks that determine which sources are authoritative, which questions are worth asking, and which methods are trustworthy. The Domain-Analytic Approach absorbed the Cognitive Paradigm's interest in meaning-making but relocated it from the individual mind to the collective practices of a community. This framework coexists with the Cognitive Paradigm rather than fully replacing it: cognitive models remain useful for studying novice-expert differences or individual learning trajectories, while domain analysis is better suited for understanding how entire fields organize and prioritize knowledge.
Social Informatics, also emerging in the 1990s and still influential, shares the Domain-Analytic Approach's conviction that context matters, but it defines context differently. Where domain analysis focuses on epistemological communities, Social Informatics examines how information and communication technologies are actually used within organizations, workplaces, and everyday life. Researchers in this tradition study how the introduction of a new system changes workflows, how informal information sharing complements formal databases, and how social norms shape technology adoption. Social Informatics narrows the Domain-Analytic Approach's broad cultural lens to a more concrete sociotechnical one: the unit of analysis is the organization or the networked group rather than the entire discipline. The two frameworks complement each other in practice—domain analysis explains why a community values certain kinds of information, while Social Informatics explains how that information flows through real institutional channels.
The most recent framework, Critical Information Studies, emerged around 2000 and continues to gain momentum. It does not reject the insights of the Domain-Analytic Approach or Social Informatics, but it adds a dimension those frameworks largely left unexamined: power. Critical Information Studies asks how information behavior is shaped by race, class, gender, colonialism, and other structures of inequality. Who gets to define what counts as authoritative information? Whose information needs are systematically ignored by mainstream systems? This framework draws on critical theory, feminist epistemology, and postcolonial studies to argue that information behavior is never politically neutral. The Cognitive Paradigm's sense-making, the Domain-Analytic Approach's knowledge communities, and Social Informatics' organizational contexts all operate within systems of privilege and marginalization that must be made visible. Critical Information Studies thus transforms the subfield's questions: instead of asking how people seek information, it asks how information systems can perpetuate or challenge existing power relations.
Today, the three active frameworks—Domain-Analytic Approach, Social Informatics, and Critical Information Studies—coexist in a state of productive disagreement. They agree on one fundamental point: information behavior cannot be reduced to system performance or individual cognition alone. Social context matters. Where they differ is on what kind of context is most important. Domain analysts look to disciplinary knowledge structures; social informaticians look to organizational and technological settings; critical scholars look to power and inequality. Each framework has developed its own methods—domain analysis often uses bibliometric mapping and discourse analysis, Social Informatics relies on ethnographic fieldwork and case studies, and Critical Information Studies employs critical discourse analysis and participatory action research. The frameworks overlap in practice: a study of how scientists share data might draw on domain analysis to understand disciplinary norms, Social Informatics to trace the institutional infrastructure, and Critical Information Studies to examine whose data is valued and whose is ignored. This pluralism is not a sign of fragmentation but of maturity. The subfield now recognizes that information behavior is too complex to be captured by any single lens, and the best research often combines frameworks to see the full picture.