“Do you agree with the view that sociology can never be a science? What limitations need to be taken into consideration in the scientific study of social phenomena?” (1984)
- The claim that sociology can “never be a science” rests on real and specific limitations in studying social phenomena, but it overstates the case by measuring sociology against a narrow, strict natural-science yardstick that even parts of natural science do not fully meet.
- Max Weber’s insistence that social reality is qualitatively different from physical reality — because it is meaningful, not merely observable — anchors the strongest form of this scepticism.
- Against this, Emile Durkheim’s demonstration, through his study of suicide, that social facts display genuine statistical regularity open to systematic, comparative study, anchors the case for sociology’s scientific credentials.
- The position taken here is a qualified one: sociology cannot be a science in the strict sense of universal, ahistorical, experimentally verified laws, but it can be — and largely is — a rigorous, systematic empirical discipline combining elements of natural-science method with an irreducibly interpretive dimension.
Limitations in the Scientific Study of Social Phenomena
- Value-ladenness of the subject matter. Unlike a chemist studying an inert compound, the sociologist is themselves part of the society under study, shaped by the same class position, culture, and historical moment as the object of research; Alvin Gouldner’s critique of the “myth” of value-free sociology argued that background value assumptions inevitably shape what gets studied, how questions are framed, and which findings feel intuitively plausible, even when researchers sincerely believe themselves neutral.
- The reflexivity/observer-effect problem. Studying people changes their behaviour in a way studying rocks does not — respondents alter their answers when aware of being observed, and, more fundamentally, Anthony Giddens’s notion of the “double hermeneutic” captures how sociological findings feed back into the society being studied, as people adjust their behaviour once they learn what social science says about people like them; there is no comparable feedback loop between a physicist’s equations and the particles they describe.
- The near-impossibility of controlled experimentation. Precise causal isolation, the hallmark of laboratory science, is rarely available to sociology on both practical grounds (society cannot be held constant while a single variable is manipulated) and ethical ones (deliberately manipulating people’s life circumstances to test a hypothesis is normatively unacceptable), forcing reliance on comparative and correlational methods instead of controlled experiment.
- The absence of strict universal laws. Human behaviour shows genuine cultural and historical variability that physical matter does not — a “law” of kinship, religion, or class formation derived from one society routinely fails to travel to another, so sociological generalizations tend to hold as tendencies or probabilities rather than as invariant laws comparable to, say, the laws of thermodynamics.
- The meaning-dimension problem. Weber’s central point — that social action is oriented to subjective meaning, so an adequate account requires interpretive understanding (Verstehen) alongside external observation — implies that a purely behaviourist, externally measured account of social life is necessarily incomplete, no matter how precisely it is quantified.
The Counter-Case: Sociology’s Genuine Scientific Achievements
- Durkheim’s study of suicide remains the founding demonstration that even the most intensely personal, seemingly random act shows robust statistical regularity when examined at the level of rates rather than individual cases — egoistic, altruistic, anomic, and fatalistic suicide rates correlate systematically with degrees of social integration and regulation across groups, religions, and historical periods, evidence that social facts are not simply reducible to individual idiosyncrasy.
- This comparative, correlational method — studying naturally occurring variation across societies and groups rather than manipulating variables in a laboratory — is precisely how other well-regarded sciences without experimental access, such as astronomy or evolutionary biology, generate reliable knowledge, undercutting the assumption that the absence of laboratory experiment alone disqualifies a field from being scientific.
- Modern sociology deploys increasingly rigorous empirical techniques — large-scale representative surveys, longitudinal panel studies tracking the same individuals over decades, natural and quasi-experiments exploiting policy changes as exogenous shocks — that achieve genuine, if probabilistic, predictive and explanatory power, for instance in explaining patterns of educational attainment, migration, or intergenerational mobility.
- A useful contemporary illustration of the discipline grappling honestly with its own scientific limits is the ongoing debate within the social sciences over a replication crisis, where researchers have begun systematically re-examining whether published findings hold up under repeated testing — a self-correcting response entirely consistent with, rather than a refutation of, scientific practice, since it is precisely the kind of internal quality-control mechanism a mature empirical field is expected to develop.
A Qualified Verdict
- Sociology is not a science in the strict positivist sense Comte originally envisioned — a discipline yielding universal, ahistorical, experimentally verified laws with the precision of physics — because its subject matter is meaningful, reflexive, value-entangled, and historically variable in ways physical matter is not.
- But this does not make it “never” a science in any useful sense; it is better understood as a distinctive kind of empirical discipline, combining nomothetic ambitions (seeking patterns, correlations, and probabilistic generalizations through systematic, evidence-based method) with an irreducibly idiographic and interpretive dimension (attending to the specific, meaningful, historically situated character of social action).
- Judging sociology a failed physics sets the wrong standard; judging it a rigorous, self-correcting, evidence-oriented discipline in its own right — one that has produced durable, replicable findings (Durkheim’s integration-suicide link chief among them) even while remaining honest about its own limits — is the more accurate and more useful verdict.