Many scientists over the past 10 - 15 years have discussed the social process of generating new knowledge. There are lots of models, most tending toward the general conclusion that traditional disciplinary investigation (within psychology or within biology, etc) limits our understanding. Gibbons et al have suggested several modes of knowledge production (Gibbons et al. The production of new knowledge: the dynamics of science and research in contemporary societies. London: Sage, 1994).
The traditional academic approach (Mode 1) is organized on disciplinary lines, is homogeneous, hierarchical and focuses on problems largely set by the scientific community. Researchers identify themselves with their discipline ("I am a sociologist") and undertake a programme of research that builds the agenda of their discipline. Cognitive and social norms determined what should be considered "significant problems" worthy of investigation, and some topics clearly would not be fundable, or would not count as thesis topics in that discipline. Often, knowledge was actively valued if it was "pure," "disinterested" and unrelated to any immediate practical goal.
Over time, social pressure gradually forced universities to address socially relevant issues and, recognizing that nature is not exclusively structured along the lines of university faculties, we learned to acknowledge the existence of folks in other Departments and (ahem) Faculties. Over time we moved toward interdisciplinary collaboration, then multidisciplinary teams. Under the Mode 2 model, research is applied and seeks to address issues important to society; the research is heterogeneous and organizational forms are transient (I may collaborate with completely different groups of people for different projects).
The resulting knowledge was seen as combining the inputs from distinct academic traditions, and we still used phrases indicating the disciplinary origins such as "From an academic viewpoint,..." There were sporting diversions such as the jousting between qualitative and quantitative researchers, positivists and post-modernists, and so forth; each person retained his or her disciplinary identity in order to retain respect in their "own" discipline.
The goal of transdisciplinary thinking is to create knowledge that extends beyond the conceptual approaches of traditional disciplines, and that is qualitatively different from the sum of the disciplinary approaches that produce it. It may develop new theory, but that theory will not be readily attributable to an existing discipline. For example, the work on income inequalities and health searches for theoretical explanations that include but transcend the economic aspects of income, blend the social environment with the personal emotional reactions to living in a given environment, and accept the notion that all these influence immune or endocrine biological response in the individual.