Erin Harrington
The Swedish silent film Häxan (1922), “The Witch,” is a remarkable account of witchcraft that offers an early intervention into how we conceptualize the effects of the interplay between truth and fantasy in horror. Danish director Benjamin Christensen’s film is a hallucinogenic documentary-narrative hybrid, presented in seven chapters, which expanded the scope and conventions of cinematic storytelling at the time. It draws from the Malleus Maleficarum (1487) and other witch-hunting treatises to exemplify the appalling outcomes of superstition and paranoia. It opens with historical context, in the form of an illustrated lecture. The bulk of the film, set in 1488, dramatizes this account, following the actions of a sorceress Karna, who makes potions for townsfolk from animals and bits of corpses, and the persecution of a poor weaver Maria, who is blamed for the illness of a wealthy printer. Crooked clergy roll into town at the behest of the Pope, performing torturous interrogations that provoke cascading accusations until entire villages are scourged.
History’s horrors are braided with phantasmal ones – and everything is threaded with an anarchic sense of humor. Shocking, jaunty scenes of Satanic temptations, demonic incursions, and witches’ sabbaths are presented with extraordinary cinematic flair, making use of a rich, deep mise-en-scène and innovative special effects, and taking pleasure in the macabre. Women birth devils, witches fly to meet on the Brocken, and a skeletal horse lumbers around in the background. Christensen himself appears as a gurning, tongue-waggling Satan, and once, in superimposition, as Christ. And yet, the film is deeply sympathetic. Its final chapter suggests we reconsider superstitious phenomena as repressed and projected desire and nervous exhaustion, using then-contemporary psychological models of hysteria. This situates executed ‘witches’ as victims of illness, oppression, marginalization and misogyny – although it fails to recognise that same misogyny in the hysterical diagnosis.
Figure 1: Devils and witches
Häxan was released in 1922 in Sweden and Denmark to great acclaim. International releases were met with simultaneous praise and horror – and in some cases significant censors’ cuts or outright bans. Critics noted its sweeping vision and virtuosity, as well as its startling gore, nudity, violence, and anticlericalism (Fujiwara); after a screening in London in 1923, one reviewer noted that “wonderful though this picture is, it is absolutely unfit for public exhibition” (‘Häxan [Review]’). The film was re-released in 1941 with an additional prologue from the director (Kendrick), but Häxan became more widely known once re-released in the United States in 1968 in a version titled Witchcraft through the Ages. Antony Balch’s edit, which cut the film from 104 to a brisk 76 minutes, is narrated drolly by William S. Burroughs and features an unruly jazz score from Daniel Humair. It teases out the humor of the original – perhaps at the expense of much of the film’s emotional nuance – and was popular on cult viewing circuits. The Swedish Film Institute has undertaken three restorations of the original Swedish-language film. Its 2016 digital restoration, informed by laboratory instructions on the film’s tinting for its premiere screenings (Mair) and notes on the original orchestral accompaniment (Anderson), was released by the Criterion Collection in 2019. These official versions, with their institutional backing and their concomitant sense of cinephilia and connoisseurship, have cemented the film’s position as a cinematic masterpiece. The film has also found new audiences through screenings accompanied by live performances of original scores.
But how might we position this film? My intention here is not to re-litigate, from a film-historical perspective, the extent to which Häxan fits within the horror genre’s canonical development. Häxan has been largely excluded from the “master narrative of horror’s lineage” (Leeder 8). Its release sits alongside touchstone works such German Expressionist films The Cabinet of Dr Caligari (1920) and Nosferatu (1922), which directly influence the ‘classical’ Hollywood horror of the 1930s, in part because of the exodus of creatives from Europe to the United States. But its style and form do not fit within, nor come to inform, a specific filmic tradition, even if we can clearly see its influence on later cycles of Satanic and folk horror (Buckley; Scovell 102-3), psychological horror films that dramatize neuroses and their treatments, and contemporary films such as Robert Eggers’ The VVitch (2015) (Hall).
It is not even that the film itself was marginal. Häxan was, at the time, the most expensive film ever to be produced in Sweden. Nonetheless, it is also an oddity within the burgeoning Danish and Swedish film industries, in particular Sweden’s ‘golden age’ (Tybjerg). Christensen himself had had an idiosyncratic career across industries and was seen by his contemporaries as an interloper (Baxtrom and Meyers 2). Häxan is also a fascinating addition to the history of documentary cinema. Its release sits alongside that of the cinematic milestone Nanook of the North (1922), widely considered to be the first feature-length documentary, yet again it remains a marginalized curiosity in accounts of that form’s development. What we can say for certain is that the film is unusual not only for its own time but for our own. There hasn’t been anything else like Häxan at all in the hundred years since its release. It is a rare example of something truly unique: a film that braids didactic rationality and sober science with cinematic innovation and lurid nightmare.
I suggest, instead, that film’s twin impulses of horror and documentary demonstrate how horror itself has drawn from signifiers of truth-telling to expand its sense of dread and terror, right from the contemporary genre’s nascent beginnings. Even outside of the complex histories of found footage or pseudo-documentary horror, these two forms share an interest in the fabrication and manipulation of seemingly objective, grounded images and more heightened cinematic language. Others have signaled how the film’s fantastical sequences act as forms of subjective truth-telling that open up spaces within the documentary form for a rich expression of historic fear and belief (Baxtrom and Meyers; Tortoliani). My reciprocal assertion is that Häxan’s peculiar documentary form (with its emphasis upon historical and material evidence), and Christensen’s own authorial presence as director and researcher, do the same for the film’s spaces of horror. This is because Häxan’s formal hybridity and overall ambiguity lead us directly into a space of epistemic and ontological uncertainty, a disruption of the known and the real, that is key to the genre in general.
The film’s structure is key to this ambiguity. The opening sets a scholarly tone: its first titles identify it as “A presentation from a cultural and historical point of view in 7 chapters of moving pictures.”[1] Our first, red-tinted image is that of a stern-faced man looking directly into the camera; this is Benjamin Christensen, who “wrote the script and produced this film between 1919 and 1921.” Here is our author, it signals; these are his labours. The text switches to the first person. Christensen thanks two of his collaborators,[2] and notes that his sources, which he also cites throughout the film’s intertitles, “are mentioned in the theatre’s playbill.”[3] It continues: “Let us look into the history of mysticism and try to explain the mysterious chapter known as the Witch.” This didactic opening situates the film as a successor to the travelling lantern lecture – something educational, well-researched, and clearly framed. But its relative sobriety and pedagogical tone sits in contrast to the images of degeneracy that will follow, which gesture to one of the magic lantern’s other popular uses: the phantasmagoria ghost show.
This opening chapter offers the thesis that a belief in witchcraft is as old as humankind, that incomprehensible phenomena are easily pinned on sorcery and spirits, and that such beliefs are “the result of naïve notions about the mystery of the universe.” The camera irises in and out on medieval woodcuts, illustrations from Persian manuscripts, diagrams from scientific manuals, and a lovely diorama of an ancient Egyptian cosmogonic model, a pointer intruding into the frame to highlight points of interest. A stunning stop-motion animated model of the earth and heavens at work illustrates the geocentric worldview of the Middle Ages, which saw God and His angelic hosts beyond the stars, and demons rising from the earth’s core.
Figure 2: Illustration of hell with pointer
And now to the dark arts – what I suspect many of us are here for. Crowded illustrations of hell are tapped again by a lecturer’s sharpened dowel. The intertitles ask us to take direct note of details: devils stuff the damned into large pots, throwing one sinner directly into the fire and pouring burning sulphur down another’s throat, while monsters tear at the damned with their teeth. We regard a “strange old mechanical presentation of hell,” a remarkably detailed articulated model in which little demons pump bellows and tear at people’s bodies, like a Hieronymus Bosch painting sprung into action. The image is tinted bright red and smoke curls and billows in front of the camera, the fires of hell manifested. “Observe the eagerness with which the devils tend to the fire under the cauldrons!” assert the titles. They clearly aren’t the only eager ones; we are barely eight minutes in, and one can’t help but think that Christensen is having a pretty good time.
Christensen’s very sincere point is that during this era devils were considered real, and feared terribly. Witches were persecuted as their consorts. He draws from illustrations from witch-hunting manuals, offering ‘evidence’ of witches casting spells, bewitching objects and animals, attending sabbaths. One, perversely, milks an axe; for what, I’m not sure. Witches smear themselves with ointment so they might fly, dance blasphemously on the holy cross, eat banquets furnished with corpses, and perform the osculum infame – kissing the Devil’s ass. We have to wonder whose these morbid fantasies are.
Figure 3: Kissing the Devil’s ass
I dissect this opening sequence in detail because this combination of thesis, audience interactivity, clear authorial enunciation, and educational intent marks Häxan as a very early type of essay film (see Corrigan 55-6). As an illustrated lecture, it draws from an historical record that is more fantastical than indexical; we are dealing with an intensity and veracity of belief, not fact. It also sets up an argument that will resolve in the film’s final chapter, as we cast a more (ostensibly) rational lens back on past horrors. Authenticity is further entrenched as Christensen, through the intertitles’ description and direct address, references doctors and scholars as he discusses the provenance of the images.
The result is an odd appeal to authority. The film establishes itself as a serious account of deep-seated superstition that ruptured social structures and resulted in significant numbers of deaths, mostly of women.[4] And yet, we are invited to take perverse pleasure in the images. Christensen whips us in and out of the “diegetic absorption” of immersive narrative that was developing in early cinema, having things both ways (Baxtrom and Meyers 40). I am reminded a little of the way vivid social hygiene videos would attempt to duck censorship laws by virtue of their educational intent (see Schaefer), offering grotesque or anti-social pleasures before urging moderation.
This not-quite-dispassionate framing is all the more playful as we head into the film proper. After a series of vignettes dramatizing beliefs about witches’ activities, including demonic temptations and misleading dreams, the central chapters tell a story about the treatment of those accused of witchcraft. The story of Maria the weaver’s terrible persecution would near stand alone as its own religious horror film. As she is tortured by monks we see images of her weathered, pitiful face; these sequences clearly influence the vivid and intimate close ups in The Passion of Joan of Arc (1928) (Baxtrom and Meyers 90). These affecting shots are cross cut with increasingly carnivalesque dramatizations of her purported occultist activities, and further fabricated accusations. Witches eat babies, women and demons dance naked, and the Devil watches on, furiously churning butter in a masturbatory frenzy. It’s wild.
Yet the truth-claims of these scenes are enigmatic. Re-enactments are, themselves, a contentious aspect of nonfiction storytelling; they “occupy a strange status in which it is crucial that they be recognized as a representation of a prior event while also signaling that they are not a representation of a contemporaneous event” (Nichols 73). But these sequences, though framed by evidence of a sort, do not depict a prior event but offer up an unsteady composite based on mad illustration, false testimony, slice of life realism, and burlesque. In their blend of artifice and evidence, horror and entertainment, their authenticity is not contingent on historical accuracy, but on how they communicate terror and desperation to a contemporary audience.
And then, two more documentarian flourishes. The narrative is arrested as we take a detour through the torture chamber. Spiked collars and pincers are held up to the camera, and tilted in the light. The camera drifts across the rack’s apparatus, and tools designed to crush arms and legs, before we see more woodcut illustrations of atrocity. “Wouldn’t we confess too?” ask the intertitles. One of the actresses has insisted on giving the thumbscrews a go – we have now ducked behind the scenes – and we see her face, at first playful, contort with pain. Then we then slip, in the final chapter, into the present day. The film identifies most of the victims of these tortures as poor, vulnerable old women. And we are pulled out again as Christensen (through intertitles) outlines a discussion he had with Maren Pedersen, the elderly actress playing Maria, in which she tells him the devil is real. He flicks through the little prayer book she has brought to show him, which includes instructions on how to identify devils, illustrating the persistence of belief. The camera again shows her weathered face in close up, blurring the line between contemporary human portrait and historic fictionalized victim, both now bound up in superstition.
Figure 4: Maria Pedersen, the actress who plays Maria
It its closing moments, the film notes that though “[w]e no longer sit in church staring terrified at the frescoes of the devils… isn’t superstition still rampant among us?” It juxtaposes images of contemporary fortunetellers and a hysteric receiving hydrotherapy at a clinic with footage of three bodies burning at the stake, presumably dummies, but perhaps not. Häxan thus hovers in an ambiguous space, where fact, fiction, and fantasy overlap. The lines between fictional and real horrors are so blurred as to be impossible to tease apart.
Just as Nanook of the North expresses tensions that remain core to the documentary (Saunders 88), I offer that Häxan embodies a sense of epistemic uncertainty that permeates the horror genre and that prefigures many of the genre’s rhetorical devices. This is the effect of its attempts to authentically express the elusive ‘truths’ of fear and irrationality through a combination of innovative trickery, (quasi?-) documentary footage, and narrative storytelling. And we can recognize this friction elsewhere, such as in supernatural films like The Amityville Horror (1979) and The Conjuring (2013), or lurid dramas such as The Devils (1971)[5] that tell us (truthfully or no) that they draw from actual events or testimony. We can also recognize it in the self-referential fabrication of evidence and frames, as in the opening of The Texas Chain Saw Massacre (1974), or the presentation of archival footage alongside fabrications, as in mondo shockumentaries like Faces of Death (1978). It’s there in the use of signifiers and frames of (faux-)authenticity in found-footage films. It’s also there in the playful bricolage of authentic and faked archival materials in comedies, such as the extended opening of What We Do in the Shadows (2014). Ultimately, Häxan shows that horror is always indebted to the past–that the terrors of today may persist in unexpected ways, and that we might successfully bear witness to terrible truths through fiction.
Notes
[1] Direct quotes come from the English subtitles of Swedish-language intertitles in the 2016 Criterion Collection edition, as available on the streaming site Kanopy.
[2] He specifically thanks art director Richard Louw and cinematographer Johan Ankerstjerne, although not his (female) editor Edla Hansen. This is representative of the way women’s labour has been minimised in horror and cinematic history – even in films, like this, which are deeply invested in the disproportionately unfair treatment of women.
[3] For a full list of these sources, see the appendix to Baxtrom and Meyers (249-252).
[4] The film suggests that 8 million were executed across centuries, although a source for this isn’t directly supplied; other scholarly estimates sit significantly lower.
[5] A sequence set in a convent near the end of Häxan shows a community of nuns experiencing a collective madness and claiming to be possessed; the film posits that their somnambulism and strange compulsions are a result of hysteria. This sequence alludes to the same real-world events as Ken Russell’s film, which stylishly dramatizes an incident in the French town of Loudun in 1624 in which a group of Ursuline nuns claimed they had been possessed by demons. The Devils also deals with a grisly witchcraft trial; the two films make an excellent double feature.
Works cited
Anderson, Gillian. “Häxan: About the Music.” The Criterion Collection, 15 Oct. 2001, https://www.criterion.com/current/posts/148-h-xan-about-the-music.
Baxstrom, Richard, and Todd Meyers. Realizing the Witch: Science, Cinema, and the Mastery of the Invisible. Fordham University Press, 2016.
Buckley, Chloé Germaine. “Häxan: ‘Let Her Suffering Begin’.” The Criterion Collection, 15 Oct. 2019, https://www.criterion.com/current/posts/6634-h-xan-let-her-suffering-begin.
Corrigan, Timothy. The Essay Film: From Montaigne, after Marker. Oxford University Press, 2011.
Fujiwara, Chris. “Häxan: The Real Unreal.” The Criterion Collection, 15 Oct. 2019, https://www.criterion.com/current/posts/147-h-xan-the-real-unreal.
Hall, Jacob. “The Influences Of ‘The Witch’ Part Two: Director Robert Eggers On ‘Häxan: Witchcraft Through the Ages’.” SlashFilm, 17 Feb. 2016, https://www.slashfilm.com/542705/the-witch-influences-haxan/.
Häxan. Benjamin Christensen, director. Janus Films, 1922. Kanopy, https://www.kanopy.com/en/product/112959.
“Häxan.” Turner Classic Movies, http://www.tcm.com/tcmdb/title/508216/haxan/. Accessed 19 June 2022.
“Häxan (Review).” Variety, 1922, https://variety.com/1921/film/reviews/haxan-1117791535/.
Kendrick, James. “A Witches’ Brew of Fact, Fiction and Spectacle.” Kinoeye: New Perspectives on European Film, vol. 3, no. 11, 2003, https://www.kinoeye.org/03/11/kendrick11.php.
Leeder, Murray. Horror Film: A Critical Introduction. Bloomsbury Academic, 2018.
Mair, Chrisopher. “Om Häxans digitala restaurering.” Svenska Filminstitutet, 19 Dec. 2016, https://www.filminstitutet.se/sv/nyheter/2016/om-haxans-digitala-restaurering/.
Nichols, Bill. “Documentary Reenactment and the Fantasmatic Subject.” Critical Inquiry, vol. 35, no. 1, Sept. 2008, pp. 72–89.
Saunders, Dave. Documentary. Routledge, 2010.
Schaefer, Eric. “Resisting Refinement: The Exploitation Film and Self-Censorship.” Film History, vol. 6, no. 3, 1994, pp. 293–313.
Scovell, Adam. Folk Horror: Hours Dreadful and Things Strange. Auteur Publishing, 2017.
Tortoliani, Erica. “Troubling Portrayals: Benjamin Christensen’s Häxan (1922), Documentary Form, and the Question of Histor(iography).” Journal of Historical Fictions, vol. 3, no. 1, 2020, pp. 24–41.
Tybjerg, Casper. ‘Dreyer and the National Film in Denmark’. Film History, vol. 13, no. 1, 2001, pp. 23–36.
