A LinkedIn discussion in the Higher Education Teaching and Learning Group is currently revisiting the perennial question of whether or not fields like psychology, engineering, and medicine ought to be allowed to call themselves “sciences”. This question seems to be of primary interest to those who have some vested interest in discrediting findings emerging from research in these fields, for reasons that are generally more political and/or economic than philosophical. But the question is worth considering.
My late father taught for many years in the Epidemiology Department in the School of Public Health at the University of Michigan – a school that embraced microscope-wielders as well as a variety of social and behavioral scientists. He noted on numerous occasions that the major divide within the faculty was approximately measured by who got to wear a while lab coat to faculty meetings and who didn’t. My father, although he did get to wear the white coat, never felt that this somehow gave added weight to his views, although many of his colleagues did, to the chagrin of the wearers of tweed and corduroy. My recommended solution was always to simply issue white lab coats to everyone, thereby rendering the distinction nugatory, but somehow this didn’t seem to have wide appeal. The symbolism of the coat took on an importance far beyond its function as a garment – just as does the categorization of an entire field of inquiry.
My position in this discussion has been that the question is fundamentally wrong. It’s not whether or not some domain of inquiry is in its essence “scientific”, but whether inquiry within that domain is governed by the norms of scientific behavior. In short, the question isn’t so much whether you are or are not “a scientist”, as whether or not you behave like a scientist in the course of your inquiry.
Science is distinguished from other modes of knowing by the idea of systematic inquiry based on falsifiability. (Note that falsifiability does not imply verifiability – propositions are never “verified”, only “supported”). Objections to Karl Popper aside, this approach at least provides a basis for distinguishing science from other modes of knowing such as faith or heuristics, both of which we probably rely on day-to-day far more than we rely on science, except indirectly. Science implicitly embraces the idea that not-knowing is the norm, and that our aim is to try to chip pieces off that mass of not-knowing as we can, in search of useful results. Science formulates models and theories and tries to assess the degree to which they can be used to predict useful things. “Utility” here includes not just things that are immediately practical in daily life, but also things that predict other things; thus something like quantum mechanics is scientifically “useful”, even though its applications in daily life are minimal.
(It’s worth noting as an aside that although the theory of relativity doesn’t really factor into ordinary calculations of satellite orbits, it’s absolutely critical to creating and maintaining GPS systems. Here’s something that’s supposed to be almost the textbook definition of pure science, and it turns out to be essential to resolving an engineering problem. One never knows when science will be called on to critically support daily life.)
By this definition, psychology along with medicine, sociology, and engineering is in fact scientific. There are numerous scientific inquiry paradigms; research doesn’t have to be experimental to be both valid and useful. My own sub-field of organizational psychology has produced some enormously helpful and replicable models and approaches based on many different kinds of studies. The research that Tora Bikson, other colleagues, and I conducted starting in the early 1980s on the effects of information technology on managerial work was a combination of quasi-experimental and observational studies, both quantitative and qualitative, largely based on the principles of socio-technical design. Nonetheless, it laid useful groundwork for many of today’s IT applications. It isn’t necessary to have a mathematically complete formulation of a phenomenon for it to be understood “scientifically”.
Ironically, some circles of the behavioral sciences community experience “physics envy” – the tendency to exclude from “scientific psychology” any research or approach that isn’t entirely quantitative and heavily mathematical. While these approaches can be informative, there are also many other modes of systematic inquiry that are equally if not more useful. Multiple inquiry paradigms are important.
I don’t think that we have to decide if some kind of activity IS or IS NOT “a science” to acknowledge that its practitioners can and do behave like scientists. Take as mundane an operation as the Department of Motor Vehicles – everyone’s favorite bureaucracy. Is DMV “a science”? Hardly – although in many of their activities they do use “scientific methods” such as sampling and statistical inference; not to mention their heavy reliance on sophisticated information technology. In fact, they “do science”. Would it be preferable to have them decide on the timing of street signals on the basis of freshly slaughtered pigeon entrails? You don’t have to BE a “scientist” in some ontological sense in order to behave like one, to good effect – usually better than when not behaving like a scientist.
Science is behavior, not essence – and the study of human activity can embrace as much “behaving like a scientist” as any other study domain. Trying to decide whether psychology is or isn’t a science is much less useful than deciding whether or not any given study appropriately adopts the norms of scientific inquiry, or is driven by faith and/or heuristics. There’s far too much of the latter passing itself off as science.
[NOTE: This post has been slightly edited, to better lead into Part 2 of the discussion.]