21(2): 316–40.
Science
Working Definition
A method of gaining valid and reliable knowledge of the world based on testing theories against collected evidence.
Origins of the Concept
The concept of science originated as a description of knowledge as such, but, by the fourteenth century in Europe, science or ‘natural philosophy’ was used in a more limited way to describe knowledge that was written down and recorded. During the seventeenth-century ‘scientific revolution’, which included many breakthroughs, such as Newton’s discovery of the force of gravity, science came to be seen more as a method of inquiry. By the nineteenth century the term came to be used only in relation to the physical world and the disciplines which studied it, among them astronomy, physics and chemistry. At the end of that century, debates in the philosophy of science focused on what kind of methods were ‘scientific’, how scientific knowledge could be verified as true, and, eventually, whether the emerging social subjects could match the kinds of evidence produced in the natural sciences.
In the twentieth century, various schools of positivism argued the relative merits of deduction or induction and verification or falsification as principles to which all sciences, not just the natural science disciplines, should adhere. However, gradually sociologists came to see their discipline as scientific but in a different way to the natural sciences, on account of the intentional actions of humans and the reflexivity that exists between society and sociological knowledge. Today sociology is divided between those who continue to see themselves as scientists of society and those who are happier with the idea that they engage in social studies, rendering questions of scientific method and status obsolete.
Meaning and Interpretation
Arguably, the key issue for sociology since Auguste Comte’s positivism has been whether or not sociology is a science. How does the discipline relate to other acknowledged sciences such as astronomy, physics, chemistry and biology? And what is it that makes them so unproblematically ‘scientific’ anyway? Many people believe that scientific research involves the use of systematic methods, the gathering of empirical evidence, the analysis of data and the development of theoretical explanations for that data. Over time, the sciences can then build a significant body of reliable knowledge. If we accept this characterization, then sociology is a science, as it does involve systematic methods of empirical investigation, the analysis of data and the assessment of theories in the light of evidence and logical argument. However, a growing number of sociologists are uneasy about discussing their discipline as scientific and may be more comfortable seeing it as closer to the humanities than to the natural sciences.
Studying human beings is clearly different in some ways from observing events in the natural world, so maybe sociology and the natural sciences can never be identical. Human beings do not merely act on instinct or through some biological imperative but interact with each other in meaningful ways. This means that, in order to describe and explain social life, sociologists need to find ways of understanding why people act in the ways that they do. People generally behave according to intentions, and sociologists will often reconstruct the meanings individuals attached to their own actions. To grasp the behaviour of frogs involves no such reconstruction of complex mental reasoning. The meaningful nature of human action is both an advantage and a problem. Sociologists cannot simply adopt the methods of successful natural sciences such as biology or chemistry but must devise their own methods that are adequate for their specific subject matter – human beings and social life. One important advantage is that sociologists can speak directly with their research participants and understand the responses they get. This opportunity to converse with the participants of research studies and to confirm one’s interpretations means that sociological findings are, at least potentially, even more reliable (different researchers would arrive at the same results) and valid (the research actually measures what it is supposed to) than those from the natural sciences. Max Weber saw such gains as crucial to the scientific status of sociology. Even though its methods are necessarily different, they are no less systematic, rigorous and theoretically informed than those of any other science.
Yet sociologists face some problems that natural scientists do not. Self-aware individuals may subtly alter their usual behaviour when they know it is being studied, thereby invalidating the researcher’s findings. For instance, in daily life people constantly attempt to manage the presentation of their self to others, and this process of ‘impression management’ may occur during sociological research. Sociologists must be aware of the distinct possibility that, during interviews and questionnaires, respondents may offer the answers they believe the researchers are looking for. These various issues illustrate a key feature of studying human beings – the problem of reflexivity.
Sociological knowledge filters back into society and becomes part of the very same social context being studied, potentially altering that social context. Social reflexivity has no counterpart in the natural sciences, which means that, if it is a science, sociology cannot simply adopt the same methods as natural science but must develop its own ‘object-adequate’ methods.
Critical Points
A fundamental problem with the notion that sociology should be scientific is that it presumes agreement on what constitutes science. Although this used to mean simply looking at what the natural sciences do, this is no longer the case. Several important studies by historians of science have eroded the certainty which used to exist in relation to science. Thomas Kuhn (1970) studied breakthroughs in science – scientific revolutions – which we might expect would occur as a result of knowledge accumulation over long periods. In fact, Kuhn saw natural science operating through ‘paradigms’ – ways of doing science based on particular theories. ‘Normal’ science was essentially a continual testing and retesting of the paradigm, which did not lead to major advances. Breakthrough moments happened when someone went beyond the paradigm to resolve an anomalous finding which then led to a new paradigm.
A further blow to the ideal type of science came from historical studies of scientific methods by Paul Feyerabend (1975). He argued that many revolutionary discoveries in science had nothing to do with scientific method. Instead, they came about through simple trial and error or even by mistakes and chance occurrences which simply cannot be taught. Feyerabend’s conclusion was that there is only one important principle of scientific method – ‘anything goes’. Only by encouraging deviance from the scientific model could innovation be safeguarded. Sticking rigidly to one method was merely a recipe for stagnation and a lack of progress. Hence, after many decades of trying to work out how sociology could mimic the methods of the natural sciences, by the 1980s it no longer seemed a worthwhile exercise.
Continuing Relevance
The immediate response of governments around the world to the Covid-19 pandemic of 2019–21 was to turn for advice and guidance to epidemiologists and medical science. This was clear evidence that science is still viewed as a superior form of knowledge compared to theological knowledge or common-sense ideas. The basis of this superiority seems to be rooted in the practical successes of science rather than in a widespread understanding of scientific methods. In sociology, there has long been disagreement on whether it is desirable for the discipline to be and be perceived as ‘scientific’, which seems to have settled into an established divide over what are appropriate research methods.
Schwemmer and Wieczorek (2020) argue that this divide can be seen in studies that adopt predominantly qualitative or quantitative
