The Limits of Epistemic Hygiene

Perhaps the most impressive (and measurable) achievement of technological modernity has been the drastic reduction in infectious disease mortality. It is difficult to exaggerate the magnitude of this victory. It is one thing to say that half, or a third, or a quarter of children used to die before their fifth birthday from infectious disease, and more adults besides. It is another thing (and quite difficult) to imagine what it was like to live under this alien (to us) regime of death. Cholera, diphtheria, smallpox, typhoid, polio, influenza, tuberculosis, pertussis, dysentery, measles, plague, yellow fever, and more besides, claimed the lives of human beings, leaving behind disfigurement, suffering, grief, and fear. There was almost nothing to be done:

The little child of Newton and Etta Riggs Loomis was removed to the home of its grandparents, Mr. and Mrs. J. C. Loomis, after diphtheria was pronounced to be in the home of Mrs. Ann Riggs, in the hopes that it might escape the dread disease. But the monster followed it and the child died Monday, aged 2 years.

Badger State Banner, January 15, 1891, collected in Wisconsin Death Trip, Michael Lesy, 1973.

The burden was not borne equally. Some families escaped; some entire families died:

Mrs. James Baty of Merrillan, while visiting the family of John Baty at La Crosse, died suddenly of a hemorrhage of the lungs. She leaves a husband, her family of 6 children having died of diphtheria last summer.

Badger State Banner, May 8, 1890, ibid.

The escape from the regime of infectious death can be credited to four broad categories of improvements:

1. Improvements in nutrition, which allowed humans to better resist and fight off pathogens;
2. Cleanliness practices to eliminate pathogens and segregate them from places where they might do damage (hand washing, antiseptic methods, sewage management, and a zeal for cleanliness in all its forms);
3. Vaccines; and
4. Antimicrobial drugs.

All these might be classified as hygiene: the set of behaviors that humans (and even other animals) use to promote health and, especially, to prevent infection. In the nineteenth century, the prestige of the hygiene model led disparate domains to adopt its name: social hygiene, mental hygiene, and of course racial hygiene.

The hygiene model of infectious disease promised an end to fear: something to be done to counteract the effects of invisible entities, who could enter our bodies unbidden. It promised, above all, meaningful agency.

Since the early days of the germ theory of disease, there was the suspicion that not only diseases, but behaviors and ideas, could be transmitted by social contagion. Nineteenth-century authors debated whether suicide, crime, and mental illness were spread contagiously. This line of thought was unified and formalized by Richard Dawkins in the late twentieth century under the name memetics.

Since then, enterprising researchers have attempted to show that a variety of behaviors and ideas are “contagious” within social networks: obesity, divorce, happiness, loneliness, sleep deprivation. (Seeing how impressed everyone was, other researchers found even more amazing contagious effects.) Your friends make you fat! Well. What can we do?

If ideas and behaviors can infect us, perhaps against our will, merely by exposure, then what can we do to regain our agency? Is there a possibility of epistemic hygiene?

Idea Hygiene

Infectious diseases cause two distinct kinds of harm. First, they are harmful to fitness in the evolutionary sense: they make those infected less likely to survive and reproduce. Second, they are harmful to flourishing, broadly construed: they cause suffering, grief, and fear, regardless of their impact on evolutionary fitness. Ideas may be harmful in either or both of these senses. Ideas, however, may be judged on yet a third metric: correctness or truth value. It is possible that even a false belief promotes evolutionary fitness and well-being. Ideas may be considered dangerous for any of these reasons.

People disagree over whether particular ideas are harmful, and to whom. Governments may judge that anti-government ideas are harmful. Political partisans may judge that the ideas of the political outgroup are harmful. Ideas that are acceptable for adults may be judged harmful to children or vulnerable people.

Let us assume, for now, that some dangerous ideas (however defined) are contagious. Following the infectious disease model, there are four hygienic responses to idea contagion:

1. “Nutrition” – increasing intelligence, knowledge, and skepticism through education
2. “Cleanliness” – censorship to limit the spread of dangerous ideas; personal avoidance of sources of harmful ideas; “entertaining a thought without accepting it”
3. “Vaccines” – exposure to weakened forms of dangerous ideas
4. “Antibiotics” – heaping ridicule and scorn on those who hold dangerous ideas; arguing on the internet

It is an open question whether education, in various forms, provides protection against dangerous ideas. I will focus on the latter three paradigms.

If ideas spread contagiously through mere exposure, as opposed to rational learning and adoption, then preventing exposure (through censorship and source avoidance) would be the only plausible means of preventing epistemic plagues. To me, it seems incredible that human minds should be capable of being infected with ideas from mere exposure; even children exhibit discrimination in their copying behavior. If we attribute to people the ability to “entertain an idea without accepting it” – to compartmentalize new behaviors and ideas and consider them without becoming “infected” at once – then other methods of containment are possible. Censorship remains a popular method of social epistemic hygiene, but as I will explain, censors are probably more concerned with effects of common knowledge than idea contagion as such.

Epistemic vaccines – exposure to a weakened or “straw man” version of a dangerous idea – have been in use for centuries (a few contemporary examples will give you the idea). The dangerous idea is presented in a patently ridiculous or noxious form, so that after encountering the idea, the audience may simply laugh or roll its eyes in disgust rather than engage proponents.

John Stuart Mill was not satisfied with such methods; a belief held “in the manner of a prejudice,” he said, was of little more use than a superstition, even if it happened to be correct. Mill, we might say, was in favor of live vaccines at a minimum, for the purpose of epistemic health. When beliefs are sheltered from strong criticism,

not only the grounds of the opinion are forgotten in the absence of discussion, but too often the meaning of the opinion itself. The words which convey it cease to suggest ideas, or suggest only a small portion of those they were originally employed to communicate. Instead of a vivid conception and a living belief, there remain only a few phrases retained by rote; or, if any part, the shell and husk only of the meaning is retained, the finer essence being lost.

On Liberty, II.26

Ironically for Mill, many learn to parrot a belief in free speech itself in “slogan form” without being able to apply the abstract principle to specific situations.

In the treatment of infectious disease, antibiotics pose two problems. First, they are often not specific to pathogenic organisms, killing beneficial as well as pathogenic bacteria. Second, they create an environment that rewards organisms that adapt to survive in their presence (antibiotic resistance).

When a person is infected with a dangerous idea, and is “treated” for this condition in various ways, analogs of both these effects may arise. Scorn and ridicule may inhibit neutral or beneficial cousins of a dangerous idea. And an environment of hostility will benefit ideas that develop immunity to the “treatment.” Arguing on the internet, however, is a great use of your time and will definitely result in the eradication of dangerous beliefs once and for all.

Social Learning

Many authors writing on idea contagion fail to distinguish between contagion from mere exposure (like infectious disease) and contagion through the rational process of learning and adoption. Young distinguishes three models for the spread of innovations:

1. Contagion. People adopt when they come in contact with others who have already adopted; that is, innovations spread much like epidemics.

2. Social influence. People adopt when enough other people in the group have adopted; that is, innovations spread by a conformity motive.

3. Social learning. People adopt once they see enough empirical evidence to convince them that the innovation is worth adopting, where the evidence is generated by the outcomes among prior adopters. Individuals may adopt at different times due to differences in their prior beliefs, amount of information gathered, and idiosyncratic costs.

Young, p. 1900.

Young argues that these forms of innovation diffusion can be distinguished by the shape of the curves of adoption. For instance, the curve formed by social learning is two curves superimposed: first, early adopters try the idea out, and adoption initially accellerates, then decelerates or even decreases as the population of early adopters is exhausted. Then, for successful innovations, a second period of superexponential adoption follows, as the rest of the population observes the idea’s success among early adopters. An example is the adoption of hybrid corn among Iowa farmers between 1926 and 1941:

Rational social learning is a hypothesis that must be considered for all forms of alleged idea contagion. People may be learning what they can get away with – and how to get away with it. This pattern seems apparent (if you squint) in literacy rates in Europe since the fifteenth century: an initial period of increase, followed by a plateau or deceleration, and then fast universal adoption.

Aside from mathematical patterns, we can consider the situation for people making difficult decisions. Divorce has been posited to be contagious in social networks. It is possible that many people desire to get divorced, but are hesitant to do so because they fear the social and economic consequences. If a close associate gets divorced, unhappy lovers will be able to observe an example of the consequences first-hand. If the consequences do not seem so bad, unhappily married observers may learn that they can “get away with” divorcing. It is difficult to distinguish a change in morals from a change in observed consequences.

Suicide is a phenomenon that is widely believed to be contagious. One author goes so far as to posit that one can be “infected with a suicide meme” through the now-disgraced mechanism of priming. Other authors find less support for the contagion hypothesis, in that a majority of studies find no effect. As has been argued regarding obesity (and, less seriously, height), apparent “contagion” (or clustering) can be in part explained by homophily – people with similar risk factors are likely to live near each other. However, to the degree which suicide is “contagious,” it appears to operate via social learning rather than mere exposure or priming.

For instance, what Stack calls “the most dramatic illustration of an imitative effect” – suicides associated with the book Final Exit – is not imitative at all. Final Exit is a book that provides information about how to complete suicide; no actual suicide or imitative target was present. Stack also mentions a suicide spike following the death of Marilyn Monroe; importantly, specific information about the method she used was published on the front page of many newspapers, down to the number of pills. In considering suicide contagion, researchers rarely attempt to distinguish the information payload (target for learning) from moral licensing or priming.

One group that seems to take social learning seriously is evangelical Christians. The practice of “witnessing” includes leading by example, trying to be a kind, righteous, happy person, so that others may see how beneficial one’s religious beliefs are. One author has criticized this practice as encouraging adherents to put up a “false front” of perfection, but I admire the way that the practice respects the rational abilities of observers. Opponents of religion might take heed.

Even in cases that appear to be hysterical contagion, the rational learning element rears its head. Hysterical contagion in religious orders begins to look like a form of rational dissent when placed in context:

Young girls typically were coerced by elders into joining these socially isolating religious orders, practising rigid discipline in confined, all-female living quarters. Their plight included forced vows of chastity and poverty. Many endured bland near-starvation diets, repetitious prayer rituals and lengthy fasting intervals. Punishment for even minor transgression included flogging and incarceration. The hysterical fits appeared under the strictest administrators. Priests were summoned to exorcise the demons, and disliked individuals often were accused of casting spells and were banished, imprisoned or burned at the stake. Witchcraft accusations also were a way to settle social and political scores under the guise of religion and justice. These rebellious nuns used foul and blasphemous language and engaged in lewd behaviour: exposing genitalia, rubbing private parts or thrusting hips to denote mock intercourse.

Bartholomew et al., p. 300.

In addition, many incidents initially believed to be hysterical contagion (because, for example, nonspecific symptoms were presented, transmission was by line-of-sight, and a high proportion of the affected were female) were later determined to be the result of poisoning (Bartholomew et al. at p. 304). Infectious contagion is one among many hypotheses that must be considered, and must be considered a diagnosis of exclusion.

Two Real Threats to Epistemic Agency

In implicit models of contagion without social learning (contagion through mere exposure), we fear the same powerlessness and lack of agency that accompanies infectious disease. In the social learning model, some of our agency is restored. However, even if the social learning model predominates in belief adoption, there remain two important sources of powerlessness over beliefs and behaviors.

The first is the power of common knowledge (see, e.g., Weaponized Sacredness). We may have a high degree of control over our own beliefs, but we have negligible control over what others believe – and even less control over second-order beliefs, what everybody believes that everybody else believes. The sacred, inviolable beliefs of a culture at a particular time are largely a given, and we are obligated to pay them at least token regard. By paying them token regard, we reinforce their obligatory nature to others. Havel’s green grocer posts his sign “Workers of the world unite!” not out of sincere belief, but from the need to signal conformity; his doing so reinforces the need for others to do so. Thus, a preference falsification regime can be maintained: fear of expressing true beliefs manifests as support for false beliefs.

In many cases, common knowledge is quite salubrious. For instance, in all populations there is some percentage who would like to steal and murder, but common knowledge that these acts are evil ensures that would-be thieves and murderers know that others will cooperate to punish their acts. Genocides frequently result from a change in common knowledge regarding the rights of outgroup members, rather than a sudden epidemic of contagious, sincere belief. The property confiscations of the Spanish Inquisition relied on common knowledge that Jews and heretics did not have property rights. The legal property rights (e.g., the right to retain property absent conviction of a crime) of certain other groups is abrogated in many modern regimes:

Dictators demand credible signals of loyalty, creating thereby observable common knowledge of their fearsomeness. The spread of common knowledge can be assisted by propaganda, as with Havel’s green grocer, or by the Rwandan radio station that broadcast assurances that it was okay to kill Tutsis and take their property. Propaganda (and even political takeover) can only go so far in changing common knowledge; the rate of genocide of Jews was very high in Romania and very low in Denmark, despite both being nominally occupied by Nazis, likely reflecting different levels of common knowledge about the permissibility of killing Jews.

The genocidal meme, in other words, is not so much concerned with convincing would-be génocidaires to have a sincere belief that killing and theft are good, but rather in giving them assurances that they will not be punished for their acts. There is little we can do about common knowledge; Havel’s prescription is to “live in truth” to what degree we can.

The second agency-shaped hole in our realities is affect – the psychology jargon noun for the quale of feeling good or bad, of happiness or misery or anger. (The noun affect is unfortunately a homograph of the very common verb to affect, though it is pronounced with emphasis on the first syllable.) By virtue of existing, we are obliged to be conscious (or at least aware, as in dreams) and to experience whatever sensations and emotions our environment and brains produce. We have some control over avoiding pain, but our bodies often produce pain when there is no good reason for it, and no clear behavioral response to end it. Similarly, we have some control over our emotions, but in many cases extreme misery is unavoidable through behavioral means. Boredom, loneliness, hunger, thirst, jealousy, longing, grief, and shame are unavoidable facts of life. We respond to these experiences using the cultural tools that are provided.

A rather obvious but heretofore largely ignored hypothesis has recently gained evidential support: that suicide is caused by extreme emotional pain, within mental illness constructs and even in the absence of mental illness. Suicide is likely a last resort, when available methods to end prolonged excruciating mental states fail. The common diagnosis of “addiction” – to alcohol, cigarettes, prescription drugs, illegal drugs, gambling, sex, video games, food, excercise, or the internet – might be better conceptualized as “affect management.” People experiencing unusually high levels of negative affect search for relief; if healthy and effective means are not available, they may choose dangerous means for palliation.

If humans were not obligated to experience, or if these obligatory experiences were not in many cases unbearably aversive, then harmful palliative behaviors would not spread. The memetic harm is arguably not in the nature of those effective harmful behaviors that spread, but rather in the absence of safe, effective palliative options for the relief of misery.

Epistemic Humility and Flourishing

Belief in mind control and brainwashing is a curious feature of human cognition. The belief in the infectious contagion of ideas resembles ancestral (and contemporary) magical thinking: that like attracts like, and that thinking about something can make it real. Nick Szabo reports a belief among the Yurok, a settled hunter-gatherer culture, that mirrors contemporary belief in the “Law of Attraction” (quoting Alfred Kroeber; dentalia refers to the long, slender shells used as money):

They are firmly convinced that persistent thinking about money will bring it. Particularly is this believed to be true while one is engaged in any sweat-house occupation. Asaman climbs the hill to gather sweat-house wood always a meritorious practice, … he puts his mind on dentalia. He makes himself see them along the trail or hanging from fir trees eating the leaves. … In the sweat-house he looks until he sees more money-shells perhaps peering at him through the door. When he goes down to the river he stares into it and at last may discern a shell as large as a salmon, with gills Working like those of a fish. . . . Saying a thing with sufficient intensity and frequency was a means towards bringing it about. A man often kept calling “I want to be rich” or “I wish dentalia” perhaps weeping at the same time…

Let us not be quick to ridicule these beliefs. Memorizing a long list of cognitive biases and attempting to faithfully abstain from them is probably not a path to epistemic purity, not least because many of the biases rest on fake social science. Mose “biases” are in fact extremely useful heuristics; social proof, for example, is an excellent method of exploiting the knowledge of others. The belief in the Law of Attraction may serve a purpose – as a symbolic means of coping with envy and economic inequality, for example. A person who believes that like creates like may be less likely to rob his neighbors! The belief in social contagion, mind control, and brainwashing may serve as-yet unknown purposes.

How can we serve truth? Consider what Havel means by “living in truth,” and the nature of dissidents:

They may be writers who write as they wish without regard for censorship or official demands and who issue their work – when official publishers refuse to print it – as samizdat. They may be philosophers, historians, sociologists, and all those who practice independent scholarship and, if it is impossible through official or semi-official channels, who also circulate their work in samizdat or who organize private discussions, lectures, and seminars. They may be teachers who privately teach young people things that are kept from them in the state schools; clergymen who either in office or, if they are deprived of their charges, outside it, try to carry on a free religious life; painters, musicians, and singers who practice their work regardless of how it is looked upon by official institutions; everyone who shares this independent culture and helps to spread it…

I am reminded of Hakim Bey’s Temporary Autonomous Zone, in which freedom and flourishing are created in the small temporal and spatial zones over which we have control. In this moment, we have a great deal of freedom.

We might become joy canaries: pursuing and facilitating flourishing to the highest degree, so that we will at least know when technological and social change make these difficult or impossible. We might attempt to disrupt preference falsification equilibria by mocking nude emperors to the degree that this is possible. If we are very brave, we might insist that violence and theft are not acceptable, even when the victims are members of hated outgroups. We would do well to offer an Epistemic Serenity Prayer.