12. You Don’t Get Scared of Monsters, You Get Scared for People: Creating Suspense across Versions in Stephen King’s IT

Vincent Neyt

©2024 Vincent Neyt, CC BY-NC 4.0 https://doi.org/10.11647/OBP.0426.12

The world may view Stephen King as a horror novelist, but he regards himself as a suspense novelist. ‘A suspense novel is basically a scare novel’, he told an interviewer in 1979; ‘I see the horror novel as only one room in a very large house, which is the suspense novel’ (Underwood and Miller 1989, 91). Suspense, according to King, can be seen as ‘diluted horror’ (81): a horrific scene can amplify the readers’ emotional state from ‘tense’ to ‘terrified’ by triggering deep-rooted fears or phobias in addition to their anxiety, hopes and fears for the characters involved. As to ‘what makes a good horror story’, King said in 1980:

In this essay, I will explore this central position assigned by King to the characters he places in harm’s way.

The study of suspense traditionally has two main focuses: the narrative that creates suspense, and the reader, viewer or listener who experiences it (Bálint 2020). I propose to widen the text-oriented focus to include a genetic approach, to study the drafts of such captivating texts. Authors of suspense-driven narratives revise their work with the goal of enhancing the reader’s experience in suspenseful episodes or of strengthening the reader’s engagement with the narrative, resulting in a more intense overall experience. King has referred to the craft of rewriting as ‘a nuts-and-bolts kind of operation’, like ‘adjusting the carburetor […] to make it right’ (Underwood and Miller 1989, 169). His valuable insights into the mechanics of suspense shine through in the adjustments he makes ‘under the hood’.

From On Writing, his memoir of the craft, it becomes clear that King is an intuitive writer who requires minimal preparation to begin a new work. He starts from what he calls a ‘situation’, which usually arises in the form of a ‘what if’ question (King 2012, 190). The situation that King explored in his epic novel IT (1986), the test case of this genetic analysis, can be formulated in its simplest form as: ‘what if a group of children came face to face with a monster at the age of eleven, and were then forced to face that same monster again as adults?’

After the ‘what if’ question, King explains, ‘the characters—always flat and unfeatured, to begin with—come next. Once these things are fixed in my mind, I begin to narrate’ (ibid.). From here, he moves forward solely on instinct. The plan is to put the characters in a predicament and watch them try to work themselves free (189) and he goes about this without any outlining or plotting beforehand. ‘For a suspense novelist’, King adds,

this is a great thing. [...] [I]f I’m not able to guess with any accuracy how the damned thing is going to turn out, even with my inside knowledge of coming events, I can be pretty sure of keeping the reader in a state of page-turning anxiety. (King 2012, 190)

In the eighties, King imposed a high tempo of writing on himself. Six pages a day, no more, no less (Underwood and Miller 1989, 75), with no rereading or revising of what was written the day before, ever moving the story forward until it’s complete. This is necessary ‘to keep up with my original enthusiasm and at the same time outrun the self-doubt that’s always waiting to settle in’ (King 2012, 249). For each work, King habitually does ‘two drafts and a polish’ (248). In this polish (which sometimes becomes a third draft), he is mainly concerned with language, with giving the work a unified stylistic feel.

King wrote three drafts of IT between 1980 and 1986: the first on a typewriter, the second and third on a personal computer.1 The original typescript of the first draft is missing, but there are photocopies. The second draft, however, is nowhere to be found. Two printouts of the third draft are kept at King’s archive. Because of the missing link in the dossier, it’s impossible to tell whether the variants between the first and third drafts entered the text at the second or third draft stage. So, as a shorthand, I will speak of changes in ‘the second/third draft’, taking the two together out of necessity. Chuck Verrill, King’s editor, gave his editorial feedback both on the first draft and the third (after King had submitted it for publication in early 1986). Verrill primarily raised issues in chronology and continuity, suggesting cuts and revisions regarding language and overall pace. On his set of proof pages, King made only cosmetic changes.

In the following two sections, I present a concise overview of relevant methodological publications on suspense theory and the narratological concepts of pace, characterisation and focalisation. The section ‘Revising IT for Suspense’ describes four patterns of revision discovered in King’s work on IT across versions, and the section ‘Meeting Pennywise’ combines these patterns in the analysis of a passage from the first chapter of the novel.

From studies in the cognitive (reader-based) approach (Zillman 1980; Gerrig and Allbritton 1990; Vorderer et al. 1996; Beecher 2007; Smuts 2008; Hakemulder et al. 2017) the consensus has arisen that an equivalent-to-reality representation of events in a narrative can trigger suspense, which is a pleasurable experience that has an emotional component (interest, hope, fear, thrill, anxiety, restlessness, empathy, sympathy) and a cognitive component (uncertainty, anticipation, prospection, gap-filling, the dynamic calculation of possible outcomes and the probability of these competing scenarios). Without the emotional component, there can be no suspense.

Foundational to the text-oriented approach in suspense theory is the work of Meir Sternberg (Sternberg 1978; 2003a; 2003b). He discerns three ‘universals of narrative’, three narrative techniques that produce enjoyment in readers: suspense, curiosity and surprise. Suspense and curiosity both derive from a lack of information, drawing the reader’s attention forward (Sternberg 1978, 65). They differ in that suspense is the emotion experienced with regard to temporary gaps in story events situated in what Sternberg calls the ‘narrative future’, and curiosity is the emotion with regard to events from the ‘narrative past’ that have not yet been related in the narrative (65). A third category is surprise, where a hidden gap is opened and the reader discovers retrospectively, at the point of closure, that there was a gap or an ambiguity (244). The three lines of narrative interest in Sternberg’s model can occur simultaneously, and usually do, on all textual levels.

Brewer and Lichtenstein based their influential ‘structural-affect theory of stories’ on Sternberg’s universals. They turned suspense, curiosity, and surprise into three separate ‘discourse structures’ that account for the ‘entertainment force’ of stories, and they state that these structures will produce three different affect curves in readers (Brewer and Lichtenstein 1982).

The ‘suspense discourse organisation’ must contain an initiating event, early in the narrative text, an event which could lead to significant consequences (either good or bad) for a character. The initiating event causes the reader to become concerned about the consequences for the relevant character and this produces suspense, which is later resolved by the outcome event. Between the initiating event and the outcome event, there is ‘additional discourse material’, or ‘outcome delay’ (what Sternberg calls ‘retardation’), to encourage the build-up of the suspense. Chronological narration is prevalent, with the exception of the technique of foreshadowing (482). In addition to the narrative-spanning suspense and resolution structure, there are ‘mini’ suspense and resolution episodes along the way, which, in the case of a long narrative, results in an affect curve that makes ‘a saw-toothed climb towards a climax’ (Brewer 1996, 116).

The initiating event is important in engaging the reader. The protagonist must by this point be characterised as likable and good and the event itself must have a considerable impact on both protagonist and reader. As a result, the reader will empathise (feel with the character) or sympathise (feel for the character) and become concerned, which is experienced as a ‘clash of hope and fear’ (Sternberg 1978, 65).

Stephen King’s IT adheres to the macro-organisation of Brewer and Lichtenstein’s ‘suspense discourse structure’. In the horror genre, suspense is the dominant line of narrative interest, with most stories moving chronologically towards a final confrontation between the protagonist and the monster. Curiosity comes into play when, in order to successfully defeat the monster, the protagonist must first discover its true nature, where it came from, and how it can be killed.

Noël Carroll states that what drives readers forward is their desire to find answers: an early scene will raise a question which is then answered in a later scene, which he calls ‘erotetic narration’ (Carroll 1990, 130). Suspense, in his view, ‘is generated as an emotional concomitant of a narrative question that has been raised by earlier scenes and events in a story’, a question that must have only ‘two possible, opposed answers which have specific ratings in terms of morality and probability’ (137). The suspense is highest when the morally just outcome is the least likely (138). It can easily be seen, Carroll concludes, how horror narratives trigger suspense in readers. Monsters are irredeemably evil, are generally immensely powerful and often operate in secret. Almost from the onset the odds are stacked heavily against the human protagonist(s) in the inevitable confrontation, and ‘the situation is ripe for suspense’ (139). Horror narratives typically will spend more time on establishing the improbability of the humans being successful against the monster than on establishing the monster’s evilness (142), suggesting that improbability—not morality—is the more important factor in suspense creation.

Suspense has also been studied at the smallest textual level. Richard Gerrig (Gerrig 1996) describes an experiment on triggering suspense through a single sentence. From the results he distils two categories that proved successful. First, sentences that suggest a lack of knowledge, either on the part of the narration, the character, or the reader, are suspenseful because they activate the problem-solving cognitive mechanism (99). A second category is ‘classic suspense schemas’ (98). Concepts that recurred in the suspenseful sentences were danger, darkness, potential physical harm, doors, fear and despair. Words associated with these concepts, Gerrig believes, evoke ‘prototypical scenes in which readers are likely to have experienced suspense in the past’ (98). This dovetails nicely with the results of Mark Algee-Hewitt’s project ‘Suspense: Language, Narrative, Affect’ at the Stanford Literary Lab, which traced suspense at the word level. They found that ‘suspenseful passages were characterised by words relating to the imagination (e.g., “thought”), the senses (“saw”), and movement (“struggled”) and topics such as “assault”, “guns”, “crime”, and “dramatic weather”’ (Ueda 2016). The presence of words that convey how things appear to be rather than how they really are, such as ‘seemed’, ‘perceived’, or ‘observed’, generate ‘epistemological uncertainty’, which translates into suspense (Ueda 2016).

In view of the structuralist division of the study of narrative texts into three levels, story, narrative and narration (Herman and Vervaeck 2019, 43), it is apparent from the preceding research that suspense is triggered by elements at all levels. An initiating story event puts the protagonist in harm’s way and a series of discovery and confrontation events leads to a climactic event in which good triumphs over evil against great odds. Mieke Bal states that suspense is mainly evoked in readers on the level of the narrative; by how story material is manipulated into a narrative sequence (Bal 2009, 76). It is not so much the action of the confrontation that is suspenseful, but the build-up to the action (the outcome delay), and this is where the manipulations occur. A good deal is also achieved through the narration itself, in the actual phrasing, as the studies on suspense at the smallest textual level have shown.

In the case of the narrative aspect of time, the manipulation lies in the speed with which the narrator goes through the story events, in the text’s ‘pace’ or ‘pacing’. Brian Gingrich gives the most comprehensive definition of the concept in his PhD dissertation ‘The Pace of Modern Fiction’: ‘large-forward-rhythmic-shifting-dynamic-temporal narrative movement’ (Gingrich 2018, 6). Pace is akin to rhythm (a pattern of varying units of narrative speed), but not identical to it: ‘what distinguishes pace from rhythm in general is that it moves forward toward senses of endings (projected moments of closure, climax, or nonnarratable resolution)’ (9).

In terms of the relationship between the duration of the narrative (reading time) and the duration of the narrated events (story time), one would expect the action in confrontations to be narrated in scenic mode, and the reading time in the suspenseful build-up to these confrontations to be longer than the story time, either by narrating a scene in slow-down, ‘a sort of scene in slow motion’ (Genette 1980, 95), or by retarding a scene by inserting descriptive pauses—two options that are often hard to distinguish from each other. Gérard Genette states that pauses in scenes can be ‘concealed’ by presenting them through the perception of a character (107): a room is described as a character sees it. Story time does not stop but runs on as the character is looking around. Alfonso de Toro proposes to distinguish between ‘static descriptions’, given by the narrator, interrupting the sequence of narrated events; and ‘dynamic descriptions’, given by a character, which only insignificantly affect the narrative sequence in its flow (de Toro 2011, 133).

Karin Kukkonen provides an additional approach to the dynamics of time in narrative texts with her notions of ‘plot speed’, ‘storyworld speed’ and ‘discourse speed’ (Kukkonen 2020). Two episodes might be narrated in the same scenic mode; for instance, a tranquil conversation scene transitioning suddenly into a high-speed car-chase scene, but the reader will experience this transition as an acceleration, purely on the basis of verbs like ‘rushing’, ‘running’, ‘chasing’ and so on. For this experience of the reader that ‘the novel is speeding up’, Kukkonen uses the term ‘storyworld speed’. Readers can have a similar experience of acceleration when a page-filling paragraph with long sentences and difficult words is followed by a series of simple one-sentence paragraphs, which is a change in ‘discourse speed’, defined as: ‘readers’ sense of how swiftly they get through a stretch of narrative in relation to its perceived length in terms of mediation’ (75). ‘Plot speed’ has to do with readers’ expectations or predictions. Changes in plot speed arise when events ‘make a projected outcome more likely (acceleration) or less likely (deceleration)’ (77).

Relevant to suspense are also the narratological aspects of characterisation and focalisation. Fotis Jannidis defines characterisation as ‘the process of ascribing properties to names which results in agents having these properties in the storyworld’ (Jannidis 2013, paragraph 3). A character’s traits can be described directly by the narrator; the reader can indirectly deduce them from the character’s actions, discourse and other metonymic elements; or characterisation can be done through analogy (Rimmon-Kenan 2002, 59–70). Philippe Hamon distinguishes four different principles that work together in the dynamic process of constructing a character throughout a narrative text. The first presentation of a character is typically followed by: the continuous repetition of its most relevant traits; an accumulation of traits that together form a whole; relations to other characters, in the form of similarities and contrasts; and transformations that a character undergoes on the level of its traits (Hamon 1977, 128).

The process is equally dynamic from the reader’s perspective. When experiencing the representation of a character in a narrative, Ralf Schneider states, readers dynamically form (and update) a mental model of that character, a process that is ‘a complex interaction of what the text says about the characters and of what the reader knows about the world in general, specifically about people and, yet more specifically, about “people” in literature’ (Schneider 2001, 608). The model is fed bottom-up from the text and top-down from the reader’s knowledge. When readers first meet a character, they may form a mental model of it based on categorisation (when they recognise it as a stock character or stereotype) or a personalised model, which ‘is constructed more laboriously in the bottom-up mode, and the result will be a more complex structure that is kept “open” for a longer time to allow for the integration of further, potentially conflicting information’ (Schneider 2013, 123). Categorised models may transition into personalised models (which Schneider calls ‘individuation’), and the opposite transition is possible as well (‘de-personalisation’). In Schneider’s analysis, only characters in the personalised category can trigger the emotions in readers that are necessary for them to experience suspense (124).

Focalisation, according to Mieke Bal, is the storyteller’s most effective tool in creating suspense: the manipulation of the information we receive by restricting it to the perception and cognitive functions of particular characters at particular times (Bal 2009, 76). Despite extensive critical attention (Genette 1980; Jahn 1996; Fludernik 1996; Bal 2009; Schmid 2010; Niederhoff 2011), focalisation has remained a ‘complex and elusive’ phenomenon of narrative texts (Niederhoff 2011, paragraph 18). In the combination of an extradiegetic narrator with an internal focaliser2 for instance (as is the case predominantly in IT), statements that are not explicitly linked by the narrator to the focalising character’s perception, such as ‘the house was hidden behind a tree’, are ambiguous as to whether the perceiving agent is external to the storyworld (the narrator) or internal to it (the focalising character). This uncertainty can be exploited to generate tension.

3. Revising IT for Suspense: Four Patterns

Most striking in King’s revision campaigns of IT is that he expanded the first half of the book. He added more detail to the suspense scenes in the opening chapters, the amount of added text gradually decreasing as he approached the midway point.3 The scenes in which there is little or no tension in the first half of the novel were similarly lengthened. He saw less need for expansion in the second half of the book; merely streamlining the scenes by revising for style and internal consistency.

Overall, King added almost no story events in his rewrite of the first half; the additions are mainly in dialogue and character descriptions. In my opinion, as a mechanic looking under the hood of his first draft text, King believed the key to sustaining his readers’ engagement in such a long novel would depend on the characterisation of its protagonists: lowering the pace of the chapters by giving more attention to the physicality, the character traits, the direct speech and the thoughts and emotions of these ‘paper people’.

The suspense structure of the novel was already firmly in place in the first draft. King made no changes to the overall suspense and resolution curve, nor to the tempo with which the pendulum swings between low tension and high tension; between confrontations and breather episodes; between scenes and summaries.

On the level of the suspenseful scene, however, particularly the early ones that were expanded, there are interesting patterns to be discerned in the revisions.4 Added text, of course, extends the reading time, sustaining the tension longer before the suspense is resolved. To create this extra length, the majority of King’s additions deal with how the focaliser experiences the danger. In the rest of this section, I will argue this in more detail.

To the narration of story events King added (or expanded) the character’s sensory impressions. He added indications of internal focalisation to neutrally narrated descriptions of danger by presenting them through the perception of the focalising character. ‘[The clown’s] face was deeply lined’ (King 1981, 156), for instance, becomes ‘Ben could see the clown’s face clearly. It was deeply lined’ (King 1986, 214). ‘The house, brooding and silent, drew closer’ (King 1981, 225) becomes ‘It did not seem as if his feet were moving; instead the house itself, brooding and silent, seemed to draw closer to where he stood’ (King 1986, 311). In its first draft state, this sentence contains the troubling description of an unnatural occurrence (a house drawing closer), and in its revised state the same thing is presented through the consciousness of a frightened eleven-year-old boy. King’s addition of the word ‘seemed’ (twice) confirms Algee-Hewitt’s remarks that verbs that relay how things appear can be used to generate epistemological uncertainty, leading to increased suspense.

Many such small changes occur throughout King’s revision campaigns of the suspense scenes in the first half of the book. To make the scenes more effective and to increase the reader’s engagement with the events, King placed the reader inside the body of the character in the dangerous situation, more so than in the first draft. Expanding the text with simple phrases pertaining to the sensory experience of the protagonist turns static descriptions to dynamic and adds diegetic outcome delay (on the micro-level) that slows down readers (by lowering the discourse speed) without giving them the impression that the action is being halted for description.

King regularly inserted details of the smells, sounds and tactile sensations that the protagonists experience in their predicament. In a scene where It terrorises Bill and Richie by making the pages of a photo album turn on their own, King revised ‘When [Bill] stopped turning [the pages], they turned themselves’ (King 1981, 251) to ‘He gave up after a minute, but the pages did not. They turned themselves, flipping slowly but steadily, with big deliberate riffling sounds’ (King 1986, 336). Further on, King added: ‘The inside of Richie’s mouth suddenly felt as dry as dust and as smooth as glass’ (337).

Similar in nature are King’s edits in the episode where young Eddie Kaspbrak is chased by the monster in the guise of a leper. The differences between the first draft and the published text are visualised in the following quote with omitted text struck through and added text in superscript:

Eddie raced for his bike. It was the same race as before, only now it ^now had the quality of a nightmare, where you can only move with ^{the most} agonizing slowness no matter how hard you try to go fast...and ^{in those dreams didn’t} you can ^always hear ^{or feel} something, some It, gaining on you.? ^{Didn’t you always smell Its stinking breath, as Eddie was smelling it now?} (King 1981, 229; King 1986, 314)

As suggested here by the additions of ‘feel’ and ‘smell’, King devoted conscious attention to the sensory experience of the focaliser across versions. The revision also lengthens reading time and story time, while giving the reader no information on the threat that lies behind the boy as he runs for his life.

Complementary to the focus on sensory experiences (transporting the reader into the character’s body), in his second/third draft King chiefly expanded these suspense scenes by adding the protagonist’s inner life, transporting the reader into the character’s conscious mind.5 The previous example can also serve as an illustration here. In the next paragraph the narrator informs us that, for a moment, Eddie felt ‘a wild hope’ that he was indeed having a nightmare, which clarifies that the narrator was expressing Eddie’s thought that his race had the quality of a nightmare, complete with sounds and smells. The revision from ‘you can always hear something’ to ‘didn’t you always hear or feel something’ (and the subsequent repetition of ‘didn’t you always’) turn the passage into more of a direct transcription of Eddie’s thoughts than narration by an external narrator.

King made heavy use of this technique in his revision of the first section of chapter 3, ‘Six Phone Calls’, in which Patty Uris tries to fight off panic when she realises that her husband Stan has locked himself in the bathroom and doesn’t answer her calls. The suspenseful scene follows the classic schema of ‘the horror that lies behind the closed door’. King expanded the text by 32% from its first draft form, slowing down the pace primarily by diving deeper into Patty’s panicked mind, as in this passage:

^{Yes. Only he had already locked the bathroom door from the inside and because it was too much of a bother to unlock it again he had simply opened the window over the tub and gone down the side of the house like a fly crawling down a wall. Sure, of course, sure—}(King 1981, 26; King 1986, 56; italics are King’s)

After having fetched a spare key, Patty forces herself to walk, not run, back to the bathroom to stave off panic, because, she thought, ‘running made the panic want to come back’ (57). In his second/third draft, King extended this paragraph considerably by adding:

Having arrived at the door with the key, Patty is afraid to use it because it is ‘somehow too final’ (58). King added: ‘If God hadn’t taken it back by the time she used the key, then He never would. The age of miracles, after all, was past’ (ibid.). The insertions more accurately evoke the story time involved in Patty’s walk back upstairs and hesitation at the door. The choice to expand this classic suspense scenario with Patty’s inner life instead of with new minor events, descriptions of setting or character physicality reveals King’s poetics on how best to amplify the tension.

Feeding Patty’s internal panic throughout the scene is an external sound coming from behind the bathroom door—the sound of dripping water: ‘Plink…pause. Plink…pause. Plink…pause’ (55). The sound motif, already present in the first draft, was further expanded. When Patty is standing by the phone, thinking about who to call and what to say, King replaced ‘someone had to know that Stan didn’t answer, because he was unconscious, or dead. Someone had to help her. Maybe it wasn’t too late yet’ (King 1981, 26), rational considerations on Patty’s part, with ‘[how did you tell someone] that the steady sound of the water dripping into the tub was killing her heart? Someone had to help her’ (King 1986, 57). The incessant drip of the faucet leaves no room in her mind for rationality or hope of a good outcome.

In the climax of the scene, King revised the description of Stan’s body in the bathtub from static to dynamic, presenting it through Patty’s perception, and added a concluding repetition of the sound motif. He expanded ‘Patty Uris at last found her voice, and staring into her husband’s dead and sparkling eyes, she began to scream’ (King 1981, 28) to:

That did it. Patty Uris at last found her voice. Staring into her husband’s dead and sparkling eyes, she began to scream. (King 1986, 59)

In summary, King chose to heighten the tension of this scene by amplifying Patty’s inner life and the sound dynamics between her shouts to her husband and the excruciating ‘plink…pause…’ which is the only reply she receives.

A third way in which King lowered the pace of suspenseful scenes was by adding repetitions of character traits of the protagonists. After having intuitively discovered the traits of the seven protagonists while writing his first draft, King used all subsequent stages of revision to paint the characters with a thicker brush. He did so most extensively in the low-tension episodes, but this pattern of revision is also present in high-tension scenes.

Richie Tozier was the character that underwent the most expansive rewrite, with a significant amount of new text devoted to him in the second/third draft. The eleven-year-old Richie loves doing comical voices and talking about rock ’n’ roll music. In the first draft, young Richie grew up to be a lawyer who no longer did impressions. King’s most significant alteration to a character in the novel was transforming adult Richie from a lawyer into a radio deejay who had gained success and fame by performing voice impressions between songs.

King inserted references to these character traits into several of the suspense scenes in which Richie is the focaliser. In the scene where Richie is attacked by It in the guise of a giant plastic statue of Paul Bunyan (a landmark in the town of Derry), King slows down a sentence containing straight-forward action:

^{Madly, in his mind, he heard the Dovells:} Oh the kids in Bristol are sharp as a pistol When they do the Bristol Stomp....
(King 1981, 504; King 1986, 586)

Although the extra length undeniably lowers the discourse speed of an actional passage (during which the reader will be anxious to find out what happens next), the storyworld speed is still high (because of the words ‘shook’, ‘rattled’, ‘earthquake’, ‘buried’, ‘madly’ and ‘Stomp’). Adding references to popular culture is another revision pattern in the second/third draft of IT, and although the other children also like rock ’n’ roll music, King singled it out as a distinguishing trait for Richie because of the change from lawyer to deejay.

Lastly, King lowered the pace of the suspense scenes in the first half of the novel by inserting more dialogue, notably by putting more words into the mouth of Pennywise the Dancing Clown. As the antagonist, Pennywise is always the focalised object, never the focaliser; it is only at the beginning of chapter 21 that we are taken inside the mind of the monster for the first time. In the many scenes that build up to that moment, King worked on his villain by expanding the external characteristics only, and the increase in dialogue is the most notable revision pattern. When Pennywise tries to lure young Ben Hanscom towards him, for instance, King’s modifications in the speech contain a repetition of one of Ben’s character traits, his love of books:

You’ll like it here, Ben, the clown said, and now ^{said. Now} it was close enough so that Ben could hear the clud-clud sound of its funny shoes on ^{made as they advanced over} the uneven ice. You’ll like it, yes, there are all sorts of things to be here; so it here, I promise, all the boys and girls I meet like it here because it’s like Pleasure Island in ^Pinocchio and Never-Never Land in ^{Peter Pan}; they never have to grow up and that’s what all the kiddies want! So come on! See the sights, have a balloon, come with me, run away with the circus, feed the elephants, see the world, Ben, oh, Ben, ride the Chute-the-Chutes! Oh you’ll like it and oh Ben how you’ll float—(King 1981, 155; King 1986, 213)

Similarly, when Pennywise menacingly invites Richie Tozier to return to the sewers to seek him out, he reminds Richie of his childhood fear of the movie ‘The Crawling Eye’: ‘We’ve got the Eye down here, Richie… We’ve got the Crawling Eye down here’ (King 1981, 509), which King revised to resemble a radio advert:

We’ve got the eye down here, Richie… you hear me? The one that crawls. If you don’t want to fly, don’t wanna say goodbye, you come on down under this here town and give a great big hi to one great big eye! (King 1986, 591)

The changes made across versions in the first physical description of Pennywise, in the novel’s opening chapter, illustrate many of the patterns of how King revised for suspense. The chapter is focalised by George Denbrough, an innocent, vulnerable and likable six-year-old boy whose older brother Bill is the novel’s protagonist. George’s violent murder by Pennywise is the initiating event (cf. Brewer and Lichtenstein) that sets up the narrative-spanning suspense arc: the first chapter introduces and endears Bill and George to readers and after George is killed readers instinctively know that what this novel is heading towards is a confrontation between Bill and Pennywise.

With the importance of the impact of this first atrocious murder in mind, King meticulously rewrote the chapter, increasing its length by a quarter, thus lowering the pace.6 The character of George Denbrough is developed further (his fears about going into the cellar, his relationship with his brother, his love of movies and television); additional insights into Bill’s character are gleaned indirectly through Georgie’s perspective; the playful dialogue between the brothers is enriched possibly with the intention of enhancing the reader’s emotional reaction to George’s subsequent death; and foreshadowing references were added to the events of the book’s climax (necessary because of King’s intuitive first draft writing practice). Small adjustments are made throughout from external to internal focalisation, such as the change from ‘[George] put on speed, and did almost in fact catch the boat’ (King 1981, 8) to ‘[George] put on speed, and for a moment he thought he would catch the boat’ (King 1986, 12).

Barely visible in the shadowy hole, he could see a clown. He ^{There was a clown in the stormdrain. The light in there was far from good, but it was good enough so that George Denbrough was sure of what he was seeing. It was a clown, like in the circus or on TV. In fact he} looked a bit like Bozo, who had been on TV until last year; his ^{a cross between Bozo and Clarabell, who talked by honking his (or was it her?—George was never really sure of the gender) horn on Howdy Doody Saturday mornings—Buffalo Bob was just about the only one who could understand Clarabell, and that always cracked George up. The} face ^{of the clown in the stormdrain} was white, there were funny tufts of red hair on either side of his bald head, ^and there was a big clown-smile painted over his mouth. ^{If George had been inhabiting a later year, he would have surely thought of Ronald McDonald before Bozo or Clarabell.} ¶7 He held a bunch of balloons like gorgeous ripe fruit in one hand. ¶ In the other he held George’s boat. (King 1981, 9; King 1986, 13)

This descriptive paragraph has been quite heavily extended, from 72 to 175 words. The word ‘clown’ has been moved to the front in a much simpler sentence, suggesting that King found this phrasing to be more effective as the slap in the readers’ face that initiates the tension in the scene. King added three repetitions of the word ‘clown’ in the paragraph. The observation that it was dark inside the hole, needed for verisimilitude, is moved, and expanded along the lines of the revision patterns discussed above.

The change from a casual mention of Bozo to three references from popular culture (Bozo, Clarabell and Ronald McDonald) is remarkable. In the first draft the scenic mode is not interrupted: it would only take George a few seconds to think of Bozo as he is making sense of what he is seeing. But in its rewritten form, the narrator distinctly pauses the scene when he chooses to digress into George’s thoughts on Clarabell’s gender and on Buffalo Bob. It is unclear whether George is thinking all of this as he is looking into the drain or if the narrator pauses here for a flashback. Of note is that King again opted to insert the focalising character’s inner life to slow down a high-tension passage.

The straightforwardly descriptive sentence (‘funny tufts of hair’ and ‘a big clown smile’) is left largely unaltered, and it is followed by an addition in which the narrator suggests that Ronald McDonald is a closer resemblance than Bozo or Clarabell. To do so, the narrator is forced to give up the internal focalisation, adding ‘If George had been inhabiting a later year’, because the mascot for McDonalds was only introduced in 1963 and this scene is situated in 1957. Intriguingly, by giving up the internal focalisation in this addition, King breaks the empathic link between George and the reader by temporarily letting the narrator focalise a part of the paragraph, which is at odds with the revision patterns described above.8

The paragraph in its revised form contains no extra information about Pennywise’s appearance. The changes mainly add to the characterisation of George, a typical child of 1950s America who watches TV. As readers we realise, one additional time, how young and innocent he is, and how great the danger is that he finds himself in now. Meanwhile, the pressing questions that readers have, at this point, about the nature and the intentions of the clown creature are purposefully left unanswered.

King turned a medium-sized descriptive paragraph into a long, lulling read that is dominated by a digression into George’s thoughts about The Howdy Doody show, slowing down the discourse speed, which then accelerates again in the two short one-sentence paragraphs that follow. The one-sentence paragraphs raise the tension, slapping readers awake with two new disturbing facts: Pennywise is holding a bunch of balloons in one hand, and George’s boat in the other. The balloons become a distinguishing trait for Pennywise throughout the novel.

Stephen King’s intuitive and fast-paced writing practice, in the case of IT, resulted in a first draft text that is likewise fast-paced. During the revision process, he made the necessary adjustments to align the text with his views on creating optimal suspense. The genetic dossier shows that only minimal alterations were required to the story events or their sequencing, but that King increased the length of the suspenseful scenes in the first half of the novel (albeit to a lesser extent as the narrative progresses).

There are clear patterns in the revisions (in both modified and added text), and they are fully in line with King’s statement that ‘you don’t get scared of monsters, you get scared for people’. In King’s view, it is the readers’ connection with the protagonist in body and mind that grabs them and keeps them engaged. If its place in the narrative sequence allows it, a scene that puts a character in danger can be made more suspenseful by immersing readers in what the protagonist is thinking, feeling and sensing. Such elements delay the outcome of the episode in a way that does not feel digressive or retardatory—on the contrary, it strengthens the reader’s empathy for the character, which also has a favourable effect on sustaining interest to the end.

The added repetitions of already established character traits show that even in suspense scenes King worked on what he called the ‘thickness’ of his characters (as quoted above). In a sense, those characters undergo a process of what Schneider called ‘individuation’ across versions. The unnaturalness and grotesqueness of the monster is most effectively conveyed to the reader if it is mediated internally through the perception and the bodily experiences of the protagonists rather than described by the narrator directly. Interestingly, most of the alterations made to the presentation of Pennywise mainly contribute to the characterisation of the person in danger. The additions that Pennywise speaks of Ben’s love of books, or that Richie is a radio deejay, for instance, have the effect that those characters (and readers along with them) realise the monster’s god-like knowledge and powers and, consequently, that their odds of besting the creature are very low indeed.

As King undertook the ‘nuts-and-bolts operation’ of rewriting his first draft of IT, he saw many opportunities for small edits that would put his readers more directly in touch with the harrowing experiences of his protagonists, to access not their fear of the monster but their concern for the people in danger.

Bálint, Katalin E. (2020), ‘Suspense’, in: The International Encyclopedia of Media Psychology, ed. by Jan Van den Bulck (Hoboken, NJ: John Wiley & Sons, Inc). https://doi.org/10.1002/9781119011071.iemp0178.

Beecher, Donald (2007), ‘Suspense’, Philosophy and Literature, 31.2: 255–79.

Brewer, W.F. (1996), ‘The Nature of Narrative Suspense and the Problem of Rereading’, in: Suspense: Conceptualizations, Theoretical Analyses, and Empirical Explorations, ed. by Peter Vorderer et al. (Mahwah, NJ: Erlbaum), 107–27.

Brewer, W.F., and E.H. Lichtenstein (1982), ‘Stories Are to Entertain: A Structural-Affect Theory of Stories’, Journal of Pragmatics, 6.5–6: 473–86.

Carroll, Noël (1990), The Philosophy of Horror or Paradoxes of the Heart (New York: Routledge).

de Toro, Alfonso (2011), ‘Time Structure in the Contemporary Novel’, in: Time: From Concept to Narrative Construct: A Reader, ed. by Jan Christoph Meister and Wilhelm Schernus (Berlin: De Gruyter), 109–42.

Fludernik, Monika (1996), Towards a ‘Natural’ Narratology (London and New York: Routledge).

Genette, Gérard (1980), Narrative Discourse, trans. by Jane E. Lewin (Ithaca, NY: Cornell University Press).

Gerrig, Richard, and David W. Allbritton (1990), ‘The Construction of Literary Character: A View from Cognitive Psychology’, Style, 24.3: 380–91.

Gerrig, Richard J. (1996), ‘The Resiliency of Suspense’, in: Suspense: Conceptualizations, Theoretical Analyses, and Empirical Explorations, ed. by Peter Vorderer et al. (Mahwah, NJ: Erlbaum), 93–106.

Gingrich, Brian (2018), The Pace of Modern Fiction: A History of Narrative Movement in Modernity (Princeton, NJ: Princeton University Press). http://arks.princeton.edu/ark:/88435/dsp012f75rb738.

Hakemulder, Frank, Moniek M. Kuijpers, Ed S. Tan, Katalin Bálint, and Minura M. Doicaru (2017), Narrative Absorption (Amsterdam: John Benjamins Publishing).

Hamon, Philippe (1977), ‘Pour un Statut Sémiologique du Personage’, in: Poétique du Récit, ed. by Roland Barthes et al. (Paris: Seuil), 115–80.

Herman, Luc, and Bart Vervaeck (2019), Handbook of Narrative Analysis, 2nd edition (Lincoln, NE: University of Nebraska Press).

Jahn, Manfred (1996), ‘Windows of Focalization: Deconstructing and Reconstructing a Narratological Concept’, Style, 30.2: 241–67.

Jannidis, Fotis (2013), ‘Character’, in: The Living Handbook of Narratology, ed. by Peter Hühn et al. (Hamburg: Hamburg University). https://www-archiv.fdm.uni-hamburg.de/lhn/node/41.html.

King, Stephen (1981), IT (first draft photocopy), Stephen King’s personal archive in Bangor, box 70, folders 4–5.

King, Stephen (1986), IT (New York: Viking).

King, Stephen (2012), On Writing: A Memoir of the Craft (London: Hodder).

Kukkonen, Karin (2020), ‘The Speed of Plot. Narrative Acceleration and Deceleration’, Orbis Litterarum, 75: 73–85, https://doi.org/10.1111/oli.12251.

Niederhoff, Burkhard (2011), ‘Focalization’, in: The Living Handbook of Narratology, ed. by Peter Hühn et al. (Hamburg: Hamburg University). https://www-archiv.fdm.uni-hamburg.de/lhn/node/18.html.

Rimmon-Kenan, Shlomith (2002), Narrative Fiction: Contemporary Poetics, 2nd edition (London: Routledge).

Schmid, Wolf (2010), Narratology: An Introduction (Berlin and New York: De Gruyter).

Schneider, Ralf (2001), ‘Toward a Cognitive Theory of Literary Character: The Dynamics of Mental Model Construction’, Style, 35.4: 607–40.

Schneider, Ralf (2013), ‘The Cognitive Theory of Character Reception: An Updated Proposal’, Anglistik: International Journal of English Studies, 24.2: 117–34.

Smuts, Aaron (2008), ‘The Desire-Frustration Theory of Suspense’, The Journal of Aesthetics and Art Criticism, 66.3: 281–90.

Sternberg, Meir (1978), Expositional Modes and Temporal Ordering (Baltimore, MD: Johns Hopkins University Press).

Sternberg, Meir (2003a), ‘Universals of Narrative and Their Cognitivist Fortunes (I)’, Poetics Today, 24.3: 297–395.

Sternberg, Meir (2003b), ‘Universals of Narrative and Their Cognitivist Fortunes (II)’, Poetics Today, 24.4: 517–638.

Ueda, Akemi (2016), ‘Stanford Literary Lab uses Digital Humanities to Study Why We Feel Suspense’, https://news.stanford.edu/2016/02/18/literary-lab-suspense-021816/.

Underwood, Tim, and Chuck Miller, ed. (1989), Bare Bones: Conversations on Terror with Stephen King (New York: Warner Books).

Vorderer, Peter, Hans J. Wulff, and Mike Friedrichsen, ed. (1996), Suspense: Conceptualizations, Theoretical Analyses, and Empirical Explorations (Mahwah, NJ: Erlbaum).

Zillman, Dolf (1980), ‘Anatomy of Suspense’, in: The Entertainment Functions of Television, ed. by P. H. Tannenbaum (Hillsdale, NJ: Lawrence Erlbaum Associates), 133–63.