Ingmar Bergman’s literary output comprises dozens of books and hundreds of articles. Admittedly, this is fewer than Balzac, though considerably more than Flaubert. Despite this prodigious corpus, Bergman asserted more or less aggressively throughout his life: ‘I myself have never had ambitions to be an author.’¹ Whether we believe this affirmation or not (I would advise against it), the key word here first demands a definition. The question ‘What is an author?’ has, of course, been famously asked—and answered—by Michel Foucault, who dates the nascence of authorship to the Renaissance as a response to the need to attribute distinctive copyright or, alternatively, to allocate personal responsibility.² The result was the ‘author as creator’, whose works began to be viewed less in their own right than as emanating from this or that individual originator.

As a creator, Ingmar Bergman has certainly benefited from this transfer of power from work to author. Officially, however, he lamented this state of affairs, aspiring to be a nameless artisan rather than a celebrated artist and identifying himself with the creators of cultural artefacts of the past, before the Renaissance elevated them to the status of artists. In an essay published in the mid-1950s, by which time Bergman was really starting to make a name for himself, he mused over the anonymity of the medieval artist:

The essay in question may very well be Bergman’s most quoted, and I suspect that no one quoting from it has ever failed to mention the name of its author. While it is easy to jest about Bergman’s false modesty—a world-famous artist yearning for anonymity—I believe his desire for obscurity to be reasonably sincere. Still, there is a paradox here: Bergman claims that anonymity is an ‘invulnerable assurance’ of artistic freedom; yet this freedom is, in turn, contingent on the promotion of the artist into a ‘name’. Somewhere in this contradiction, we find the explanation as to why Bergman refused (or tried to refuse, or pretended to try to refuse) to be laurelled as an author.

While Bergman’s aversion to being called a writer may have been sincere, I will nevertheless do my best to prove him wrong. In this chapter, I will attempt to disregard his films (or, at the very least, decline to agree with his own view) where these are the end result of a process that is rather uninteresting in itself. That is to say, instances where the written works upon which the films are based are, at best, sketches. I will instead view his writings as works in their own right. Where I have considered them at all, the films are to be regarded as interpretations. Bergman’s own adaptations of his works may remain authoritative, much as when a playwright stages one of his own works to critical acclaim. Although a contemporary audience viewing such an adaptation may consider it an unsurpassable benchmark of the play in question, this too shall pass. Interpretations come and go while works remain. After all, Shakespeare’s main occupation was that of theatre manager, and he probably considered his works merely a necessary means to attract an audience. Moreover, were we able to travel back in time to watch the Bard’s own staging of Hamlet, we would in all probability regard it as a sub-par production with confusing direction, overdramatic acting, and poor set design (and even poorer lighting), to say nothing of the seating and the horrible audience. Media forms (be they films, stage performances, or books) are historically contingent, some more so than others. In short, my (presumably) controversial position is that Bergman the writer has a strong chance of outliving Bergman the filmmaker.

Writing for the screen

As a writer, Ingmar Bergman excelled in genres such as essays, memoirs, diaries, and letters. In this instance, however, I will limit myself to considering the bulk of his authorship—his screenplays.⁴ If, indeed, that is the correct term for a form of writing that sometimes looks suspiciously like a traditional drama, sometimes like the prose of a novel, and sometimes even like poetry. Only rarely do they resemble a conventional screenplay. I would argue that the term coined by their first Swedish publisher, filmberättelser (‘film stories’ or ‘film narratives’), is a more apt description, a rubric that better captures their genre-defying qualities (cf. the ciné-romans of Marguerite Duras or Alain Robbe-Grillet). Rather than discussing thematic features (which, in principle, do not differ from the corresponding films), I will focus on the screenplays’ specifically literary aspects. My aim is, quite simply, to show that Bergman’s film scripts, although written for the screen, are literary works rather than screenplays.

Screenplays typically adhere to a strict form in terms of dramaturgic elements, literary style, and even typography. Bergman often complained that these conventions restrained him, at least until the late 1950s. All the same, his screenwriting was considerably independent and unusually, even unnecessarily, literary from the outset. When we first meet the knight in The Seventh Seal (1957), he ‘stares directly into the morning sun which wallows up from the misty sea like some bloated, dying fish’.⁵ Even as a supposed source of inspiration for the actor or cinematographer (which is how Bergman explained himself when writing things like this), the metaphor of the sun as a dying fish is, quite literally, too extravagant—at least for a screenplay. This passage is also addressed by Birgitta Steene, who cites it as an example of ‘non-cinematic’ features in Bergman’s screenplays, including the early ones, in which he ‘used metaphors and similes that [gave] literary significance to the text but were hardly transposable to the screen unless transformed into a piece of visual surrealism’.⁶ This is a form of excess that can only be credited to the account of the writer, rather than to that of the filmmaker. It is in instances such as these that Bergman reveals himself (perhaps even to himself) to be an author.

While cultured people will not hesitate to assert that ‘the book was better’ after having seen a film based on a celebrated novel, such an opinion may seem blasphemous in the case of a Bergman film vis-à-vis its literary origin. In the case of Hour of the Wolf (1968), however, I stand by this assessment. As a written work, it ranks among Bergman’s greatest; it is also, I submit, nigh on unsuitable for filmic adaptations—by its author or by anyone else. Above all, the text is as much a closet drama as Goethe’s Faust or Ibsen’s Peer Gynt. That is to say, while evidently possible to adapt for the screen or stage, this is first and foremost a work of literature belonging to the dream-play genre. Hour of the Wolf takes place in the twilight zone between objective and subjective, exterior and interior, waking and dreaming, reason and madness. Though this is not impossible to capture on film, text offers other and less demanding possibilities. In a work of literature, a character can interact with an unreliable outer world without the reader ever having to determine the ontological status of that world. In cinematic works, however, the filmmaker must make up the spectators’ minds for them: objective reality or subjective state of mind. Indeed, even when these boundaries are blurred, as in the Bergman films Persona (1966) and Cries and Whispers (1972), we are at least aware that they are obscure. In Hour of the Wolf, even the question of obscurity is obscure; strange things happen, such as when a character walks ‘up the wall to the ceiling and is standing like a fly, head downward, apparently without the slightest inconvenience’.⁷ Note how nimbly this peculiar state of affairs is described. On film, this same unusual course of events would be more awkward, and the scene more easily reduced to a reliance on special effects. Another reason why Hour of the Wolf works better in writing lies in the way it often refers to vague characters without making the reader aware of who is being mentioned—in the Swedish original, that is. Interestingly enough, the English translation makes things much ‘clearer’, and hence less intriguing, demanding, and original. The following passage may serve as an example: ‘Then I catch sight of Johan. He is huddling behind the stump of a tree, trying to hide, but quite visible.’⁸ Here, the translator has seen fit both to change the narration from the third to the first person and also to name the character behind the tree stump, Johan, although Bergman never gave him away so easily. A more faithful translation of the first sentence might instead read: ‘Then she catches sight of him.’⁹ Although circumstantial evidence might lead the reader to suspect that ‘him’ is Johan, the fact of the matter is that the text never reveals that, which is only fitting for a story about demons (or, as they are referred to in the text, ‘those others’). This ambiguity is also difficult to capture on film.

Furthermore, as literary fiction Hour of the Wolf adds a meta-literary quality to the film, since it is a text (the book we are reading) about a text (Johan’s diary, as integrated into the story). That said, it is difficult to separate one from the other at times, as the narration moves seamlessly between Johan’s ‘subjective’ perceptions of reality, as rendered in his diary entries, and seemingly (but only seemingly) ‘objective’ events as retold by Alma. Hour of the Wolf is Bergman’s version of a récit in the tradition of Maurice Blanchot—a narrative that is not the narration of an event, but the event itself.

There is a note within parentheses on the very first page of Hour of the Wolf that recurs throughout the manuscript: ‘(From here on, the text is to be accompanied by images ad libitum.)’¹⁰ What on earth is such a note doing in a film script? Here, Bergman (who is often quoted as saying how much he distrusts language in general, and his own verbal capacity in particular), relying instead on his audio-visual gifts, is considering images as a supplement to the text, rather than the other way around. Besides, how can one possibly make a film whose images are to be decided ad libitum, ‘at one’s pleasure’? After all, the shooting of an expensive film requires somewhat more planning than that, and Bergman was not exactly known for his willingness to improvise. So what should one make of this proposed ad-libbing? The only reasonable explanation is that the Bergman who wrote Hour of the Wolf was writing a work of literature rather than a film. (The fact that the work was eventually adapted for cinema should not deter us from drawing this conclusion.)

It should be mentioned, however, that these notes are not included in the published English translation of Hour of the Wolf. In fact, this is far from the only case where translators, editors, and publishers have taken it upon themselves to ‘improve’ Bergman’s writings by purging them of excesses and perceived semantic, syntactic, and grammatical idiosyncrasies. Their reasons are fairly obvious: if even the author himself regards these texts as mere sketches (‘half-measures’ is a word Bergman often used to refer to his scripts), as long as the story comes across, the philological filigree with which proper literature is ordinarily invested is simply unnecessary.

Rhythm and reduction

Moving on to another example, very different from the textual peculiarities of Hour of the Wolf, Bergman’s Autumn Sonata (1968) is an example of ‘pure’ drama. Along with Scenes from a Marriage, this work is Bergman’s most reduced text. It hardly includes any instructions of any kind, such as what the scenery should look like or how a character should express something, and virtually no information about what characters feel or think. It is just dialogue, pure and simple. And yet it is very consciously written. The passage below, which illustrates this feature particularly well, may give the impression of being either unfinished or an instance of stream of consciousness. Whatever the case may be, it is quite spectacular, not least when one considers (as one always should) that this is a line meant to be heard rather than read, even if it works well in print, too. Despite running to 207 words in total, which makes it the longest sentence in Bergman’s collected prose works, the passage has a breathing rhythm that testifies to Bergman’s profound understanding of performances of the written word:

It is a rare treat for a Swede to be afforded the opportunity to be snobbish about his native language. Hence, I have seized the chance presented here to quote a lengthy passage in what eighteenth-century poet Esaias Tegnér called ‘the language of heroes and glory’. My intent is not to embarrass readers of Bergman who are not proficient in Swedish (in fact, foreign readers have advantages over native speakers which I will address later), but to emphasize two things: (1) the rhythm of Bergman’s language, and (2) the perils of translation. To illustrate these points, let us compare the beautiful Swedish sentence above with the English translation, where we read the following:

Although this is by far the best of the available English translations of Autumn Sonata, once again Bergman’s language is domesticated to the brink of banality. Not only is this ‘vomit’ of lamentation in a single, run-on sentence broken down into several sentences, but the translator also chooses to use quotation marks where Bergman, as he often does, has left them out, thereby subtly reminding his readers to be on their guard: who is speaking? To what end? The quotation marks indicate that Eva’s account is accurate. Without the quotation marks (as in the original Swedish text), the question remains more open.

While we will soon return to the minutiae of Bergman’s writing, such as quotation marks, I would first like to compare Autumn Sonata’s clear-cut, reduced style with that of another Bergman screenplay from the same period. Whereas Autumn Sonata is a downsized, traditional drama, Fanny and Alexander (film released in 1982) is prose, and at times almost like a novel. This is especially true of the prologue, with its long descriptions of the grandmother’s apartment, in which Bergman fragments Alexander’s perceptual faculties by describing them one at a time. First, the narrator tells us what the protagonist is seeing:

Note the use of the verbs see, glimpse, and regard: here, vision is in focus. Alexander is observing, but also sensing himself being observed by the marble statue. This passage continues for almost three pages, in visual impressions of furniture, paintings, photographs, light, darkness, and so on. Next among the sensual data is smell, with descriptions of olfactory impressions ranging from the odours of cabbage soup to those of the outhouse, and, not least, the various aromas of people: ‘symphonies of odour’ composed of sweat, tobacco, perfume, powder, soap, urine, etc. Finally, hearing is introduced with a call for silence—‘if you stand quite still and hold your breath, you can hear the silence’¹⁴—followed by descriptions of clocks ticking, pens scratching, and dishes rattling, until the catalogue of sounds ends when a housekeeper fills the stove with coal, and the ‘noise breaks the spell’.¹⁵

This well-ordered, programmatic division of the senses is reminiscent of an assignment in a creative-writing class. It may be argued that these non-cinematic descriptions (some of which are quite impossible to convey on film) serve the purpose of inspiring the actors, set designers, costume designers, cinematographer, or what have you. It is also, however (and regardless of what Bergman would say on the matter), literature. In this context, it is interesting to note that the author sometimes seems irritated by his own uneconomical style of writing, feeling the need to remind himself that he is writing a film and nothing else: ‘Now, do not write things that can’t be translated into images’, reads a note in Bergman’s work-diary for Fanny and Alexander. ‘I’m really tired of that; it becomes a kind of semi-literary snobbery that doesn’t belong anywhere.’¹⁶ As an enthusiastic reader of Bergman, I would nonetheless argue that it does.

When the written word is spoken

There is never any doubt that the words in Bergman’s works, including those spoken aloud, are written. This is a trait for which his films have been criticized, especially by his countrymen who hear how strange the language sounds. Bergman’s alleged difficulty in hitting the right verbal notes annoyed his colleague Bo Widerberg (among others), who wondered ‘to what extent Bergman’s foreign translators are part of his success, if people simply talk less strangely in the American versions of his films’.¹⁷ Similar objections are still being voiced today. But Bergman is no realist, and never has been. Spoken language in Bergman is archaic, elevated. Above all, it is a written language, scantily disguised as spoken words. Furthermore, Bergman evinces the utmost concern not only for the meaning of his words, but also for the way they sound. In particularly inspired moments, he combines phonetics with semantics to achieve extraordinary results. For example, consider a scene that is often misquoted, despite being Bergman’s most famous scene by far. Admittedly, misquoting the dialogue is easy enough to do, since it is hard to hear what is actually being said:

The knight thus misinterprets Death (as, in all honesty, who does not?). To the question of whether he is prepared or not, he replies that his body is frightened, but that he himself is not (possibly thereby demonstrating just the opposite, being so scared that he cannot even hear what is being said). These two lines convey several of Bergman’s major themes: problems of communication, the split self, and truth and lies. That aside, the scene also demonstrates how The Seventh Seal, while being Bergman’s most famous film, also emphasizes writing. In the above dialogue, Bergman plays with the quasi-homonymity of the Swedish words for ‘frightened’ and ‘prepared’ (rädd and beredd); rädd is pronounced in exactly the same way as the second syllable in beredd, which allows for a misinterpretation of Bergman’s, the knight’s, and, seemingly, also Death’s own words (improbable though this might seem). The pun is completely lost in translation, however, and I wonder to what extent non-Swedes have pondered the knight’s strange answer to the question as to whether he is prepared to die. (Although I suppose his answer may appear less strange than Death’s walking around incarnate in the first place.)

In British film critic David Thomson’s musings over his youth in the 1950s, and Bergman’s role in it, he makes particular mention of the Swedish language ‘blooming in our mouths with its gentle, pious, slightly smug closed vowels and its swallowing syllabics. We mimed the word “Smultronstället” from the dark as Victor Sjöström and Bibi Andersson uttered it in Wild Strawberries [(1957)].’²⁰ Further, Anthony Lane writes in the New Yorker:

Following one of the first-ever screenings of The Seventh Seal in an anglophone country (in Edinburgh in 1957), a review in The Scotsman praised the film, including ‘the excellent subtitles, suggesting that it has been written in dramatic blank verse’. The reviewer concludes that, although the film’s language lacks ‘the wit of Shakespeare’s’, it is nonetheless extraordinarily apt to the subject matter of the film.²² Although it becomes apparent on closer inspection that The Seventh Seal is not written in blank verse, iambic patterns frequently occur in it. In fact, it is not entirely improbable that one reason for the knight’s strange reply to Death, besides that previously discussed, is metrics. Had the knight not misheard, he might have answered ‘Min kropp är beredd, inte jag själv’, in which case the line would be out of rhythm. As the line is written, however, it forms four iambs (as opposed to five, as in Shakespearian pentameter):

While The Scotsman’s reviewer was slightly mistaken, it seems he was still on to something that most Swedish speakers miss (as I know I did when I watched Bergman before reading him). Sometimes, not understanding a language can be an advantage in fully appreciating its distinctive characteristics. In a quarrel depicted in Wild Strawberries, Marianne ironically exclaims ‘Stackars Evald’ (‘Poor Evald’), to which he replies ‘Var god stackra mig inte’ (‘Please don’t “poor” me’).²³ Also, a little later: ‘Det här livet äcklar mig till kräkningar och jag tänker inte dra på mig ett ansvar som tvingar mig att existera en dag längre än jag själv vill.’ (‘This life disgusts me and I don’t think that I need a responsibility which will force me to exist another day longer than I want to.’)²⁴ Compare these words, or rather the sounds produced by phrases such as ‘stackra mig inte’ or ‘äcklar mig till kräkningar’ with ‘min kropp är rädd’. Or, for that matter, with seemingly controversial expressions found in Saraband (2003) and Faithless (2000) such as ‘frukostera’ and ‘nattsärk’—archaic words for ‘eating breakfast’ and ‘nightgown’, respectively, as has been scornfully noted by Bergman’s compatriots. Phonetically, these expressions are all rather similar, with the alveolar trill [r] in close connection with the voiceless velar stop [k]. In Bergman, words do not merely serve the purpose of representing this or that; he also pays the utmost attention to how they sound. Were we to listen to Bergman’s writings (and, indeed, his films) rather than reading or watching them, and without paying too much attention to semantics, we would often hear different variations of the krk sound. Whether this has any significance, I cannot tell. But the notion of listening to this ‘mysterious music’, as Anthony Lane describes it, brings me to the topic of Bergman as poet.

Poetry in motion

Together with Cries and Whispers, Faithless, Private Confessions (1996), and perhaps a few other works of his, Persona is one example of what we might call a calculated preliminarity among Bergman’s writings. This characteristic is noted by the writer himself, who opens Persona with the following words: ‘I have not produced a film script in the normal sense. What I have written seems more like the melody line of a piece of music, which I hope with the help of my colleagues to be able to orchestrate during production.’²⁵ This statement is followed by the strange passage quoted below, an equivalent of which was included in the famous prologue to the film. (That said, as anyone who has seen Persona will confirm, the film does not follow the screenplay very carefully, and one wonders how it could have done):

Persona was elsewhere described by Bergman as a poem: ‘not in words,’ he hurries to explain, ‘but in images’.²⁷ In fact, anyone who reads the script can testify that it is also written poetry. The images conjured verbally are of a nature that is not easily captured in literal images, as is illustrated in the following example: ‘short fragments of words, like sparks, begin to drip from the ceiling and walls’—imagery that is more a case of catachresis than of metaphor, in which tenor and vehicle do not match (‘fragments of words’ that ‘drip’?). Interestingly enough, the next ‘collision’ of images also deals explicitly with language, when words ‘begin to emerge like the shadows of fish in deep waters’. Then comes the tentative doubtfulness: it was not a cloud but a sheet of water. Or was it a cloud after all? Or a tree, perhaps? No, a lunar landscape.

I leave it to others to decide whether this is prose, poetry, or something in between; but I think we can agree that it is not a conventional screenplay. Regardless of genre, however, Bergman’s texts are always characterized by a strong sense of rhythm, drastic metaphors, and the phonetic effects of alliterations and assonances—in short, by the sensitivity to the form and colour of language that designates a poet.

Bergman devotes a passage in The Magic Lantern to the importance of being true to the text when staging a play. There are a couple of sentences in the original manuscript (omitted in the printed version) that I believe offer an important insight into the mind of Bergman, not only as a director but also as a writer: ‘To me, interpretation is a listening to the breathing of the text: Why are these combinations of words, these commas, these hyphens right here?’²⁸ Punctuation has certain effects, not only in performed drama but also in literature (especially in literature!). Bergman often called for a system of notation for film similar to that for music. As he puts it in his essay ‘Each Film Is My Last’:

Actually, such a system of notation already exists: written language. I believe that this is how we should understand Bergman’s use of punctuation: as an attempt to emulate the supposedly more exact half, quarter, and sixteenth rests of music. Moreover, Bergman used every sign at his disposal to the full. By way of example, Private Confessions includes a pertinent conversation between Anna and her mother. Note, in particular, the various punctuation marks used after the repeated word ‘och’ (Swedish for ‘and’):

The repeated conjunction ‘och’ is followed by a colon, an exclamation mark, and a full stop, in that order. Linguistically, only the first of these is uncontroversial, yet the others are scarcely mistakes. These three distinct punctuation marks render different results that are subtle, yet perceptible. The writer knows what he is doing. Now, compare the English translation:

Translation is always difficult—some would say impossible—and too critical a scrutiny of an English version might seem unfair. Yet, in this case, the translator has made too many errors for them to pass without censure. First, she misses a whole sentence: in the original, a sentence follows ‘[…] breakdown and destitution’ that could perhaps be translated as: ‘And! He must have a sleeping pill at once.’ Instead, this rather significant detail is omitted entirely, which makes the ensuing sentence something of a non-sequitur. Second, the penultimate sentence in the quoted passage has Anna putting out the light when, at least according to Bergman, she is, in fact, turning it on. Third, in the Swedish original, Anna repeats the phrase ‘Och sedan’ at the end of the passage, with the small but significant difference that the first sentence ends with a full stop and the second with a question mark, signifying, as I read it, resignation followed by despair. Fourth, with regard to punctuation, Bergman’s idiosyncratic but stylistically logical and important ways of punctuating the three ‘ands’ have eluded the translator. In the two aforementioned cases (she ignored the third case altogether), she opted to insert an ellipsis; and thus, by ‘correcting’ Bergman’s language, she effaced the nuanced shift in tone while destroying the staccato of the phrases.

Reading readings

In closing, I will offer a reading of readings in Bergman, of which there are plenty. In fact, diaries, letters, and books abound in Bergman’s films (cf. Anna Sofia Rossholm’s Chapter 5 in this volume), and much can be said about the following examples: Alma’s reading of Elisabet’s letter to the doctor in Persona, Johan’s falling asleep while Marianne reads to him from her notebook in Scenes from a Marriage (1973), the book Isak reads to the children in Fanny and Alexander, and so on. All the same, I will focus on a book that is not just any book, but the ‘book of books’. Although there are probably already too many analyses of The Seventh Seal, I will venture to offer one more. Despite the large number of interpretations, few critics seem to address the fact that The Seventh Seal is also a film about media theory, and, more specifically, about writing, as the title itself implies: the seal of what, exactly? In the Book of Revelation, we read:

Besides being an early example of metaliterature, the Book of Revelation is the strangest of all the books of the Bible. From a literary standpoint, it only becomes more fascinating when the scroll described so suggestively turns out to be what Alfred Hitchcock used to call a ‘McGuffin’. The seven seals are broken one after the other, attended by mysterious horsemen, earthquakes, and angels blowing trumpets. Still, what is actually written in the scrolls is never revealed. Consequently, The Seventh Seal’s primary literary reference is to a book about a book that no one reads. As such, it is a fine example of Bergman’s fundamental ambivalence towards writing, an ambivalence about which, paradoxically, he has written so eloquently. When writings are included in Bergman’s works (be they texts in foreign languages, secret diaries, or private letters read by others than their addressees), they are either difficult to read, unreliable, or the mediators of unwanted information. It is hardly an accident that the heroes of The Seventh Seal are a band of jesters who state their business through performance rather than in writing. In all likelihood, they cannot even read.

It is my hope that this chapter has proved Ingmar Bergman’s status not only as an important writer, but also as a writer who insisted on making writing itself both a major theme of and a stylistically fundamental principle in his writing. Bergman’s films are good—his works of literature are great.