Dancing plagues, psychological epidemics: can we use infectious diseases to understand social influence?

In 1518 a mysterious disease struck Strasbourg, Alsace. One woman, Frau Troffea was struck down with the illness first. She began to dance manically in the street. Hours turned to days. Her dancing was relentless. Others joined her — sources suggest 34 people by the end of the week, 400 by the end of the month. Many continued until death from heart attacks, strokes and exhaustion.

This event was one of a series which are now known as “dancing epidemics”. At least ten outbreaks occurred between the late 1300s and 1518 — all taking place in Europe, and almost all near the river Rhine.

The Dancing Plague. [Public domain], via Wikimedia Commons

Why were there so many episodes in a relatively narrow patch of time and space? Many theories have attempted to explain this. One is that the manic dancers were poisoned by ergot, a fungus which can induce delusions and convulsions. However historical descriptions of the epidemics are not characteristic of typical behaviour when poisoned by ergot. Periods of dancing lasted for days, rather than hours, beyond the time one would expect the fungus to have an effect. Another idea was that those affected were part of a sect, performing a strange religious ritual. This also does not stand up well to historical analysis.

A more plausible idea is that psychological stress triggered altered states of mind. Altered states of mind as a result of stress is empirically supported and discussed in psychology. Whilst it is hard to establish a causal link with limited historical evidence, there certainly were many stressors acting on the populations affected by the dancing plague. During this time, famine was rife in the area affected by the outbreaks.

The information surrounding these epidemics is patchy at best, and I am no historian. Yet it raises some interesting questions, ones that could be interesting for psychologists, historians and epidemiogists to explore together.

The spread of social phenomena, from mass hysteria in schoolgirls to the spread of bigoted views, can be unsettling. We often think of ourselves as autonomous, consistent beings. Very unusual or unhealthy behaviours are castigated to the realm of “other”: performed by those who are consistently, pathologically weird. These “outbreaks” of behaviour force us to consider an alternative idea of ourselves - one which is more swayed by our environments and others than we’d like to admit. More positively, it is a reminder that we are not our actions. We may engage in behaviours or experience certain mental states, however we are not defined by them. They are not constant. We are not procrastinators or smokers or even psychotic. Instead we are people who procrastinate, people who smoke and/or people who experience psychosis.

The idea of a “Social Contagion” is widely used in marketing, behavioural economics and social network research. “Memes” are a similar concept. Now a common word to describe internet trends (at least in the lexicon of millenials like me), the term was originally coined by Richard Dawkins to compare the transmission and propogation of cultural phenomena to changes in frequency of genes over time, following similar principles to natural selection.

Within the machine learning world there’s been interesting work on “influence maximisation” — identifying from cascades of “contagions” spreading through a social network which individuals influence their social contacts the most. In other words, who leads the people around them to take on certain behaviours, get sick or tweet about #covefye and #saltbae. This idea could be applied to the spread of infectious diseases, twitter hashtags or (in a the highly unlikely event of having the right kind of data recorded) manic flashmob dancing in 16th century France.

Meanwhile, in epidemiology a wealth of work has been carried out modelling the dynamics and transmission of infectious disease through networks, as well as understanding the impact of interventions like vaccination. Key measures — like the number of new cases originating from each case, and the indirect effect of vaccinating a subset of the population on disease dynamics- could be useful measures in understanding the spread of other things between people. But can these principles really be applied to the spread of behaviour and ideas?

“Social vaccine” is a term used by the health promotion community, targeting social determinants of health such as diet, smoking and risky sexual behaviour. A medium article discussing the univeral basic income is an excellent example of the metaphor in action.

However, even if you can identify key “influencers” or risk groups, successfully altering behaviour is hard — way harder than convincing enough people that vaccination is a good idea (and that can be pretty hard — sometimes for understandable reasons, and sometimes not). Behaviour change or social vaccination schemes can be marred by a lack of understanding of the groups targeted. These groups often include communities already marginalized by society such as sex workers, drug users and those of lower socio-economic status. Humans are also fickle beings; ones whose behaviour can be contextual, difficult to predict and dependent on a complex interplay of influences.

There have been many discussions about “vaccinating” vulnerable young people from the influence of extremism — stopping the spread of ideas and influence. Whether you think this metaphor is tasteful or not, it is an interesting concept. I would argue that social cohesion and creating a tolerant society is key to avoiding extremism. Real disease intervention schemes can fall down when there is too much of a top-down approach and a poor understanding of the people being vaccinated and the reasons they may become infected. In the same way, schemes such as Prevent in the UK can only work when they come from a place of understanding and community ownership.