Facebook’s Hate Speech Policy and the ‘Mystery of Elche’

On Sept. 17, Facebook announced new steps it is taking to combat hate and extremism. The topic is a pressing one for all of us, since Facebook’s platform reaches something like a quarter of our world’s population. In some corners of that world, its platform has been used to mobilize mass violence: for example, by members of Myanmar’s Buddhist majority against its largely Muslim Rohingya minority. In others, it’s been deployed to sway voting behavior: On Sept. 12, Facebook suspended a chatbot operated by Benjamin Netanyahu’s official campaign account for messaging that Israel’s Arab politicians “want to destroy us all.” The range and power of the company’s networks place Facebook’s community standards and hate speech policies among the most important codes of civility and rules for communication on the planet, even if almost no one has read them.

Facebook is presumably feeling a great deal of pressure, not just in the United States but around the world, to develop more interventionist policies. One can understand why: How else to prevent these social media platforms—of a size unparalleled in the history of humanity—from serving, not “to connect” people (as goes Facebook’s corporate motto) but to polarize, separate, and even kill? Presumably there is also a great deal of pressure—from free speech advocates, for example, or from politicians whose discourses might run afoul of such standards—for Facebook not to develop more interventionist policies. Facebook’s efforts to revise its hate speech policies have been the subject of media attention.

What advice can a historian offer the guardians of the largest communications platform on earth as they take up that important work? Historical advice is needed. Ideas that are powerful enough to motivate violence or community—ideas about religion, nationality, race and ethnicity, gender and sexuality, and many others—are deeply historical. It is because these ideas are an important part of our inherited culture, because they seem intimately connected to a past we consider ours and to values we hold dear, that their deployment has the power to move us. We shouldn’t doubt the relevance of history, but we might doubt our ability to ban history’s potential to motivate violence and exclusion, as Facebook’s definition of hate speech seems to require.

Here are Facebook’s current definitions:

We define hate speech as a direct attack on people based on what we call protected characteristics—race, ethnicity, national origin, religious affiliation, sexual orientation, caste, sex, gender, gender identity, and serious disease or disability. We also provide some protections for immigration status. We define attack as violent or dehumanizing speech, statements of inferiority, or calls for exclusion or segregation.

The definition is laudable. The problem, as always lies in how to interpret it. Applied too strictly, you cut off much of the communication and debate to which democratic societies are committed. Apply it too loosely, and you permit the mobilization of prejudices with a proven capacity to promote violence.

After reading this definition, you might be surprised at what you can currently find on Facebook. Look up #Rothschild, for example, and you might be treated to postings blaming members of that family (sometimes by first and last name) for everything from the Sept. 11 terror attacks to the collapse of the world financial order, mass immigration to Europe and the United States, and the takeover of the planet by aliens from outer space. You can upload English-language videos calling the “God of Islam” a terrorist and torturer. And if you know Arabic, you can find plenty of material taking aim at “Zionists” and “Crusaders.”

One could wish that none of this material were part of the inherited mental furniture of humanity. But it is, and often enough, it is part of our most influential histories.

Consider the case of Robert Bowers, who entered the Tree of Life synagogue in Pittsburgh last year and shot 13 worshippers, killing 11. Shortly before pulling the trigger of his assault rifle, Bowers had posted a paraphrase from the Gospel of John on his social media page (which for the record was on Gab.com, not Facebook): “jews are the children of satan. (john 8:44)” The gospel’s actual words: “You belong to your father, the devil, and you want to carry out his desires. He was a murderer from the beginning, refusing to uphold the truth, because there is no truth in him. When he lies, he speaks his native language, because he is a liar and the father of lies.”

Should hate speech policies ban the posting of any passages from Scriptures, or from any religious text or teaching, that have mobilized violence in the past and have a potential to do so in the future? Should any mention of the Buddhist texts, temples, and teachings that are now being invoked to justify anti-Muslim violence be stripped from social media in Myanmar? Should any citation of passages from the Quran that have ever been invoked to mobilize anti-Christian movements across the millennium and a half that Islam and Christianity have coexisted in Egypt be banned?

The challenge is not only a religious one. So many different episodes of history have been invoked in the past to justify violence—think of contemporary white supremacists’ use of symbols from the Civil War in the United States, or of Brenton Tarrant’s invocation of the Balkan wars in his New Zealand massacre of Muslims—how do we decide which aspects of the past are safe enough to be posted, and which are not?

The memes by which prejudice is transmitted and violence mobilized in a society are not easily separated from that society’s highest values.

Consider the unlikely example of the Mystery of Elche, which I confess would never have occurred to me if my friend Stephen Greenblatt had not invited me to witness that UNESCO-designated “Masterpiece of the Oral and Intangible Heritage of Humanity.” The small town of Elche on Spain’s Mediterranean coast is known for two things: the magnificent palm groves planted a thousand years ago, when the region was under Muslim rule, and the religious drama with which its Christian “re-conquerors” have celebrated the festival of the Assumption of the Virgin Mary to heaven every Aug. 14 and 15 since shortly before Columbus sailed. The Elche Mystery—the word means liturgical theater—is the oldest continuously performed drama in Europe, older even than its more famous cousin, the Oberammergau Passion Play, performed in southern Germany since roughly 1634, and made notorious by Hitler’s praise for its depiction of Jews as enemies of God.

The Elche play is highly unlikely to go viral: It lasts about five hours, spread over two days. On the first day of the festival, La Vespra, the Virgin, is moved by a desire to visit the places where her son suffered and died. She (I say she, though as in Shakespeare’s drama, all the female roles are played by boys) and her retinue enter the town’s Basilica de Santa María, dressed in blue robes, head wreathed in a nimbus of gold, moving in the stiffest of ritual cadences, as if trying to look like a painted icon rather than a living being. Blond angels (also played by boys, though the gender of angels remains an open theological question) place cushions before her as she kneels to sing in monophony at each of the Stations of the Cross. The sound of the boy’s plainchant—high, thin, Catalan syllables floating across the vast church, hushed and sweating—remains haunting while flirting with monotony.

When the retinue reaches a raised platform built for the occasion beneath the soaring cupola of the church, a trap door opens high, astoundingly high above our heads, and a singing angel begins a slow descent, lowered by a wooden winch hidden with its human operators in the uppermost recesses of the cupola. This high-wire traffic between heaven and earth is one of the marvels of the mystery, and it remains as dramatic today, in an age long accustomed to special effects, as it must have been hundreds of years ago.

The angel announces to Mary her imminent death, grants her wish that the apostles gather around her before she dies, and gives her a palm leaf to be buried with her. Now the apostles begin to arrive, singing in rich polyphony, starting with the “beloved” John and ending with “doubting” Thomas (who will show up a day late). The shift in music is momentous: It feels as if a movie that started in black and white had suddenly switched to technicolor.

The apostles (still minus Thomas) gather around Mary, who bids them farewell, gives burial instructions, and falls over backward. Five angels descend from the sky, collect her soul in the form of a small effigy, and ascend again, reminding (in song) the apostles where to bury the mother of God.

These words provide the palest idea of what it feels like to sit, as my wife and I did this year, packed among thousands of worshippers in a sweltering Mediterranean basilica as this venerable performance of Christian devotion unfolds around you. At the end of this first evening an elderly local couple seated behind us (who had kindly whispered commentary in our ears whenever they thought we might miss something, such as the Virgin’s sudden collapse) assured us that, if we had been moved by this first day, we would be even more moved by the second, which would be much more dramatic. We must not miss the Jews, they told us. According to them, it was the best part.

Day two, La Festa, was indeed dramatic. It opens with the apostles bearing the Virgin to burial on a bier. Suddenly a group of “Jews” (played by Christians: Jews have not lived in the region since the 15th century) appear at the church door, dressed like extras in Ben Hur. They are agitated, furious, gesticulating toward the scene taking place on the platform beneath the cupola. Their song is harsh: This honoring of the Virgin is a great dishonor to them, and they will not allow it. The antithesis is stark: on the one side Jews hateful and incredulous, on the other Christian piety. We could just as well be looking at a 15th-century painting such as Jan van Eyck’s “Fountain of Grace” (whose arrival in Spain is more or less contemporary with that of the Mystery), as at a 21st-century production.

The Jews run up the ramp and attack the apostles, seeking to seize the body of the Virgin. The two groups are locked in combat, until one of the Jews grabs onto the bier. Suddenly his hands are paralyzed. The other Jews, seeing this miracle, are stunned. They all agree to convert, and the choir turns into hushed harmonies, as each Jew is tapped on the forehead with the palm frond and converted to a Christian. Apostles and (former) Jews now circle the platform, processional crosses high in the air, jointly bearing the body of Jesus’ mother.

There is more, of course. You must not miss the four angels who descend, with harp and guitar, singing a melody that will stick with you for months, in order to take the Virgin’s statue aloft; nor the Trinity descending from above them to crown her, suspended high above us all, with the regalia of the queen of heaven. You should also stay for the throbbing call and response of the thousands of spectators proclaiming the Virgin’s glory after the drama’s conclusion. But it is true, as my elderly native informants had told me, that the “Judiada” (as it is called—the word means both a crowd of Jews and a bad turn, an action with negative effects) stands out as a moment of dynamic human conflict in the midst of liturgy, as if the biblical figures painted on gilded altars had suddenly come to life and begun quarreling among us. No wonder so many generations of the play’s viewers have focused on the scene.

Though the play has been performed in the Basilica de Santa María for almost 600 years, 2019 is, as it happens, the first year you can watch the festival on its Facebook page, Misteri d’Elx Oficial. So here is a question: Should the play’s Jew-scene be banned under Facebook’s hate speech policies? After all, from a historical perspective it is a member of a venerable and vastly powerful family of stereotypes, part of a Christian discourse that presented Jews as the enemies of God and of the godly.

One way of thinking about stereotypes is as culturally transmitted cognitive shortcuts, prejudices (literally “pre-judgments”) by which humans try to make sense of a complex cosmos. From that perspective, stereotypes about Jews must surely count as one of the tools with which Christians (and Muslims, and plenty of secular cultures as well) have sought to distinguish good from evil in a morally ambivalent and confusing world. Rosemary Ruether’s Faith and Fratricide: the Theological Roots of Anti-Semitism, is only one of the many learned books that have presented this negative use of Jews and Judaism by Christian theology as one of the causes of violence against Jews, and even of the Holocaust.

An awareness of this potential connection between modern genocide and ancient representations of cosmic enmity is presumably what prompted the sponsors of the Oberammergau Passion Play to revise their depiction of Jews in the decades after World War II, taking steps such as referring to Jesus as rabbi, and replacing Jewish guards with Roman ones. Such revisions are challenging. It is not an easy task for a community to reinterpret what it understands as its most foundational revelations, even when it knows that those teachings mobilize violence.

When it comes to the Mystery of Elche, the task should be easier. The New Testament says nothing about Mary’s last days. Stories about Mary’s death, dormition, or assumption did not appear in our sources until the fourth or fifth century. As soon as they did, they were quickly provided by writers such as Pseudo-Melito of Sardis with validating episodes of Jewish enmity. Those episodes are not part of the earliest Christian teachings. They have certainly contributed to the ways in which many centuries of the Virgin’s worshipers have thought about Jews and Judaism, and have even mobilized violence.

Could such episodes be removed without damaging any of the central tenets of devotion? Should they be? Is it desirable, or even possible, to purge our inherited cultural practices—our languages, religions, arts, industries, morals, etc.—of their history, that is, of the prejudices and habits of thought of the past worlds in which they arose?

In the late 18th century, at roughly the time the Bill of Rights was being ratified in the newly founded United States, an enlightened bishop suppressed the “Judiada” scene in Elche’s play. The bishop was supposedly scandalized by the violence the scene provoked within the basilica when it was performed: Apparently the actors playing the Jews were being beaten up by audience members eager to join the apostles in the defense of the Virgin. The scene was written back into the script in 1924, as part of an effort to “restore” the festival. Like many such restorations, this one says as much about the period doing the restoring, as it does about the “historical origins” being restored. It has been pleasing audiences ever since.

Even if we all agreed that the “Judiada” represents a “hateful stereotype” of Jewish enmity toward the good and godly, we might still ask: Does it threaten violence, dehumanization, or exclusion in a way that Facebook’s policies forbid? That isn’t an easy question to answer. Viewing the scene you could (as some scholars do) interpret it as a celebration of inclusion: Once converted, the Jews participate next to the apostles in the burial of the Mother of God. On the other hand, the representation of Jews as enemies whose inclusion in society requires miraculously coerced conversion can reasonably be understood as dehumanizing and exclusionary, all the more so in a society with a past history of forced conversions, expulsions, and inquisitions.

The question of threat is no easier to answer. After all, there are very few Jews living in Spain today, and even if there were, the representation doesn’t call for any attack, though it may encourage existing stereotypes. One cleric we encountered on the street during the morning procession that opens the second day of the festival commented to us that the Jews had been very powerful in Mary’s time, and remain so in our own: Polling data suggests that the latter view is very common in Spain. But why worry about such stereotypes, if so few Jews are left in Spain to be harmed by them?

There are today virtually no Jews living in Iran, Iraq, Egypt, Poland, Hungary, and many other countries where Jews and Judaism once flourished. Does that mean that the active redeployment of stereotypes in those lands has no potential to cause harm? Of course not. The potential for harm remains, and not only to Jews or to Israel, but also to all the citizens of those lands whose politics is being manipulated by politicians deploying discourses of enmity to consolidate their own power.

We have no way of knowing what effects ideas and ideals from our past will have in the unpredictable future, but we should not doubt that they can and will have effects. To pick an example closer to Elche: There were virtually no Jews living in Bavaria in the early centuries of the Oberammergau Passion Play’s performance, but it would be difficult to deny that the Christian anti-Judaism represented in those performances played a role in the prejudices with which Jews were received when they started immigrating in large numbers to German-speaking lands in the decades around 1800. Hence Hitler’s (and before him, Wagner’s) praise for the play when he sought to mobilize those prejudices for purposes of total extermination.

I recognize that my medieval thought experiment does not provide Facebook with much of what military types call actionable intelligence. Companies and polities must act with the tools at their disposal, whereas my musings suggested that stripping what Facebook defines as “hateful stereotypes” from our minds, our media, and our cultures would require far more than those tools. The problem is not only that the corporate social media of our new age are motivated primarily by profit and growth, or that democratic societies depend upon debate and the free exchange of ideas. It is also that the memes by which prejudice is transmitted and violence mobilized in a society are not easily separated from that society’s highest values. Our most persistent stereotypes and our most powerful (and powerfully contested) ideals exist in close proximity, not only in our religions, but in our languages, philosophies, ideologies, and all the other behaviors we learn from the cultures into which we are born and whose past we inherit. Algorithms alone cannot save us from our history. We must attend more critically to the past.