Explaining supernatural cures for poorly understood illnesses: Revisiting classic questions with new evidence

Why do people invoke supernatural beings and forces to explain illnesses that do not have obvious natural causes? This article is a significant contribution to the literature on this question—a venerable literature in anthropology, sociology, and psychology going all the way back to the beginning of those disciplines and continuing today in evolutionary anthropology and evolutionary psychology. It is significant for three reasons: 1) it derives four intuitive hypotheses from the literature about why people invoke supernatural forces and beings to explain illnesses (like cancer) that do not have clear causes compared to things (like a broken leg) that have more obvious causes; 2) it tests these hypotheses with a unique set of data; and 3) it sets up a program of research to continue the search for more refined answers to this age-old problem. We treat these in turn.

First, about the hypotheses. From the beginning of social and psychological sciences, there was intense interest in and a small handful of reasons for the existence of belief in supernatural beings and forces: 1) it reduces anxiety about death and other things we cannot control (1–3; and the following: 4, 5); 2) it provides order to society (6–8; and the following: 9, 10; and 3) it fills in the gaps in our knowledge about why people invoke the supernatural as the cause of illness, disasters, and other misfortunes when a natural explanation is missing. Famously, David Hume (11) saw this as a failure of reason, but current scholarship sees the belief in supernatural beings and forces as a product of biological and cultural evolution (see, e.g., refs. 12–15).

Barra et al. (16) derive four intuitive hypotheses about the use of supernatural treatments for disease: 1) People will use supernatural treatments for any disease whose cause is not clear. 2) Severe illnesses are more likely to be treated with supernatural treatments than with naturalistic treatments. 3) Ritualistic treatments are more likely to be used for illnesses that cause high anxiety. And 4) Illnesses that cause disability and result in the need for care will provoke explicitly religious treatments (as opposed to other types of supernatural treatments).

Second, they test these hypotheses by analyzing a unique dataset: the Schools Collection, a database of folk medicine in Ireland in the 1930s. Some 50,000 children (born in 1925 or later), interviewed parents, grandparents, and elders (born in 1850 or later) about cures for 35 illnesses. The clever part: The authors of this paper got two physicians to rate the 35 diseases for severity (e.g., What proportion of patients would die of this disease, assuming no treatment?); for level of pain (a 9-point scale from no pain to the worst pain possible); for the level of disability (a 9-point scale from no assistance needed to total dependence on others); for uncertainty about the cause of the disease (from 1, patients would likely be very uncertain about the cause of the disease to 9, patients would likely to be very certain about the cause of the disease); and for the anxiety of level of the typical patient (from 1, not anxious/distressed at all, to 9, extremely anxious/distressed).

The main result: There was support only for hypothesis 1 and no support for any of the other hypotheses. Supernatural or religious treatments were more common for diseases with unclear causes, but not for illnesses that produced high anxiety or for illnesses that had severe outcomes or disability requiring long-term care.

Third, with exemplary restraint in not overstating the significance of their results, the authors set up a program of research for more refined answers to this ageless problem. For example, their data say that people turned to supernatural remedies, but not necessarily to Christian practices. This may have been, they speculate, because the survey asked for “local cures” and also because children collected the data, both of which might have discouraged informants from reporting things like prayer or consulting priests. They also recognize that others have found a relationship between ritualized actions (e.g., do this three times on successive days) and the reduction of anxiety and say this might be because “we did not capture anxiety well in our measure.”

In addition to these caveats, the authors preregistered their hypotheses and describe exactly what they did to code and analyze the texts, making the work replicable. Then, in discussing their results, the authors ask: “Why should uncertainty about the mechanisms or causes elicit more supernatural content?” They suggest that: “When the causes and mechanisms of a disease are a mystery, the illnesses present few affordances for action and folk models of biology and healing fail to suggest plausible ways to intervene,” so “perhaps it is in these contexts that people draw upon more supernatural treatments grounded in more diffuse and elastic conceptual frameworks, such as sympathetic magic.” This could be tested in controlled lab experiments.

This article draws our attention back to classic findings from anthropology and psychology that can inform how we test the roles that religion and the supernatural play in anxiety reduction and decision-making under risk and uncertainty.

This research agenda has clear applications to medicine, but it can and should inform the broader field of decision-making under risk and uncertainty. Scholars in this area have advanced theories to understand human behavior under various risks, including climate change, economic investments, political conflicts, and the role of AI. Yet, despite considerable theoretical advances, the role of religion and supernatural beliefs has not been foregrounded (17–19). This article draws our attention back to classic findings from anthropology and psychology that can inform how we test the roles that religion and the supernatural play in anxiety reduction and decision-making under risk and uncertainty.