Narrative representation and fictionality in performative media

1 Fictionality vs. representation

Since Lyndsey Turner’s staging of Hamlet at the London Barbican Theatre in 2015 we know that Hamlet looks exactly like Benedict Cumberbatch and wears sneakers as well as a David Bowie T-shirt. On top of this, Shakespeare’s Hamlet obviously looks like Sherlock Holmes and Alan Turing, too! This is, of course, an exaggerative interpretation that completely neglects the rules of representation and no recipient of these performative media would actually think in such restrictive terms. However, both understanding the scope of representation properly, and being aware of it when interacting with different media, seem to constitute a challenge – not only for common movie or theater goers, but also for scholars of narratology. On the one hand, we understand that we only see an actor embodying Hamlet and we do not get confused by the fact that Hamlet looks like Cumberbatch in one staging, and Maxine Peake in another.^[1] On the other hand, this self-evident understanding comes to an end when we look at aspects of the mise en scène, stage design, props, or costumes. As spectators, we intuitively think (without really reflecting upon it) that Hamlet’s clothes must look exactly like the one we see: the same cut, the same color, the same material, the same size. In the particular Hamlet storyworld (cf. Herman 2005) of Lyndsey Turner, Hamlet is obviously wearing a David Bowie T-shirt. Or is he not?

In the discussion about the way in which one interacts with fictions, Coleridge (1983: 6) famously theorized the “willing suspension of disbelief”, and connected it to the phenomenon of imagination, which he describes as “[t]his power [that] reveals itself in the balance or reconciliation of opposite or discordant qualities: of [...] the individual, with the representative”. Walton (1978) refers to and rejects Coleridge’s standpoint when he writes about

traditional ideas that the normal or desired attitude toward fiction involves a “suspension of disbelief,” or a “decrease of distance.” These phrases are unfortunate. They strongly suggest that people do not (completely) disbelieve what they read in novels and see on the stage or screen, that, e. g., we somehow accept it as fact that a boy named “Huckleberry Finn” floated down the Mississippi River – at least while we are engrossed in the novel. The normal reader does not accept this as fact, nor should he. Our disbelief is “suspended” only in the sense that it is, in some ways, set aside or ignored. (Walton 1978: 23)

And he goes on claiming that “[r]ather than somehow fooling ourselves into thinking fictions are real, we become fictional. So we end up ‘on the same level’ with fictions” (Walton 1978: 23). The issues being dealt with here can be described with the German term Fiktionsvertrag (translatable as ‘contract of fiction’), which indicates a theory that connects the contexts of a representation to the notion of fictionality. Simply put, these contexts are genre conventions that mark the representation as a “staged discourse” rather than a discourse itself (Iser 1991: 35; my translation from the German). The question of whether something is imagined or not becomes irrelevant, because the context already activates a specific mode of reception (cf. Anz 2007: 13). Ignoring this specific “fictional” context leads to a Fehlverhalten (translatable as ‘wrongdoing’ or ‘misbehavior’), as Iser (1991: 36) points out, as exemplified by Partridge in Fielding’s The History of Tom Jones (1749), when he witnesses a staging of Hamlet and thinks the represented events are real. Does a viewer of Turner’s Hamlet staging accept the fact that Hamlet wears a David Bowie T-Shirt – although it is clearly a reference to the time of the telling and not to the time of the told – because it is part of a fictional setup anyway?

Whether you think that you become fictional yourself or that you suspend your disbelief in the fictionality of the narrated – and thus let the fictional become factual – the notion of fictionality (used in the sense of make-believe [Walton 1990; Currie 1990], pretense, or an “as if” [Vaihinger 1965; Searle 1975; Iser 1978]) seems, however, in this case to be misleading. What is actually being discussed here is, instead, the notion of representation. If we call every representation a fiction, because it makes us believe something, the term fictionality loses its differential power: there would be no way to systematically distinguish a staging of Hamlet from the staging of real-life events (as is frequently done, for example with the refugee ‘crisis’ recently or in documentary theater in the 1960s). Iser (1996: 665) differentiates fiction and imagination and describes selection and combination as “functions” of fiction – and hence makes it independent from the notion of imagining something. Given how closely these functions resemble the functions of narrative mediation (that I will discuss in greater detail below), I suggest we speak of representation when dealing with a given narrative communicative artifact, and connect the term fictionality closely to what Iser terms imagination (cf. Iser 1996: 668–678), or to “communicated invention,” to use Nielsen’s and Zetterberg Gjerlevsen’s (forthcoming) words. In this sense, fictionality can be a feature of narrative representations, but does not have to. In other words my take on fictionality is as follows: there is one ‘fictive’^[2] Hamlet who is represented in different ways on the stage, so that these representations are never ‘identical’ with the fictive persona.

Additionally, without a similar differentiation, there would be no general way to distinguish fictive/imagined events within a representation from the representation itself as fiction. There are interesting cross-overs between these levels, whose analysis requires a precise terminology: the ghost of Hamlet’s father, for example, can be seen as an imaginationary character within the storyworld (“This is the very coinage of your braine” Shakespeare 2017: 1194, Hamlet 3.4: 134), embodied by a real flesh-and-blood actor – at least in some stagings. Moreover, the confusion between fiction and representation is already text-inherent: Hamlet himself speaks about fiction (‘fixion’) when he actually addresses a theatrical representation of events that he has just witnessed.

O what a rogue and pesant slaue am I.

Is it not monstrous that this player heere

But in a fixion, in a dreame of passion

Could force his soule so to his owne conceit

That from her working all the visage wand,

Teares in his eyes, distraction in his aspect,

A broken voyce, and his whole function suting

With formes to his conceit; and all for nothing,

For Hecuba. (Shakespeare 2017: 1173, Hamlet 2.2: 446–454)

It is an essential part of theater’s ‘magic’ to playfully merge the different communicative levels of representation (extradiegetic, intradiegetic, metadiegetic, etc.) – but of course they are still there. “Compare how ridiculous it would be for an actor playing Horatio in a performance of Hamlet to exclaim, when the ghost appears, ‘Look, my lord, it comes, in the fictional world of the play!’” (Walton 1978: 19). Although actors in performative media pretend to be someone else – which makes it possible to take the pretense principle of fictional discourse (cf. Searle 1975: 324) to be fulfilled, since the actors are part of the authorial collective of a performance – the communicative artifact does not have to be fictional at all, but rather ‘just’ representational.

Fictionality and representation can also be differentiated as two kinds of double-layered communication. Firstly, within fictionality, whilst the real authorial collective of a staging communicates with the real audience, the actions that are presented or staged are only fictional truths. Opinions, thoughts, emotions, etc. are thus communicated within the fictional frame and do not reflect the opinions, thoughts, emotions, etc. of the real actors or directors. This kind of double-layered communication is often used as an argument for taking into account the existence of a narrator or narrative instance (cf. e. g. Genette 1993: 72–73; Kuhn 2011: 69; Martínez and Scheffel 2012 [1999]: 19–20; Rimmon-Kenan 1997: 3–4; Schmid 2008: 41; Zipfel 2001: 182). In this respect, Hamlet’s T-shirt (or more accurate: the T-shirt the actor playing Hamlet is wearing) is a prop in a game of make-believe (cf. Walton 1990: 35–43). It invites us to imagine that Hamlet is alive and wearing a David Bowie T-shirt, regardless of the ontological boundary separating us as viewers from Hamlet as a fictive character. However, there is also double-layered communication that is constituted in the narrative functions of mediation, as described in the following section. In this case, we can speak of narrative representation, with all its implications. If functions of narrative mediation can be detected in a work of art – i. e. especially the ones that make it possible to differentiate between story time and discourse time – , the work is representational.^[3] We can therefore rightly assume that a narrator, a narrative instance, or – as I prefer – a system of representation is in charge of fulfilling these functions, or that there is some sort of cognitive construct which these functions can be ascribed to. In this way, it is less the moral and ideological implications expressed by fictive events, and more the functionality of the narrating, that leads to our assumption that there is a narrative instance or a system of representation at work. Although fictions almost always accompany narrative representation, fictionality is independent of narrative representation, as argued by Zetterberg Gjerlevsen (2016), among others (cf. Walsh 2007; Nielsen, Phelan, and Walsh 2015). Moreover, narrative representation can appear independently of fictional or factual discourse. In both cases – double-layered communication as fictionality, or double-layered communication as narrative representation – scholarship often speaks of a mediating narrator, thus combining functions of narrative and aspects of fictionality.

2 Technological and narrative mediation in verbal and (audio-)visual representations

For a representation to be narrative, mediation is an important and often-postulated prerequisite. In relevant scholarship, discussion of narrative mediation has long been preoccupied with differentiating between narrative literature and non-narrative literature. The former, which is also defined as fiction, is frequently discussed under the notion of fictionality, which has hindered the creation of a clear definition of narrative mediation independent of fictionality (or factuality). With reference to ancient sources such as Aristotle’s Poetics or Plato’s Politeia, narrative studies have claimed from their very beginnings that mediation is the most fundamental aspect of narrative discourse. Whereas Friedemann (1910) merely talks about the (anthropomorphized) narrator as the evaluating, feeling, and perceiving instance – and thus describes aspects of mediation without using the term yet – it is Stanzel (1955, 1979) who establishes the term and differentiates between two modes of mediation, namely that of the teller and that of the reflector, the latter constituting an “illusion of immediacy” (1979: 190; my translation from the German). This differentiation goes back to the distinction between telling and showing made in Anglo-American research (1990, 1988). Furthermore, Genette (cf. Lubbock 1968; Friedman 1955) sees mediation as a twofold phenomenon that appears on the levels of récit (narrative discourse) and of narration (the narrator’s act of telling).

Alber and Fludernik (2014: § 2) misconstrue Genette’s understanding of the concept of narrative mediation in describing visual or audio-visual narrative mediation: “the narrating instance represents events and existents (story), and they are thereby mediated in a particular (verbal, visual, or audio-visual) sign system (narrative).” For Genette narrative is “stricto sensu [...] a verbal transmission” (Genette 1988: 16) and can therefore be oral or written, but not – in his definition of the term – visual or audio-visual.

Chatman was the first to widen the scope of the definition of narrative mediation by concentrating on narrative functions instead of reactivating an anthropomorphized understanding of the narrator. His approach is based on a general assumption of the “doubly temporal logic” of story and discourse as essential for narrative: “Narrative entails movement through time not only ‘externally’ (the duration of the presentation of the novel, film, play) but also ‘internally’ (the duration of the sequence of events that constitute the plot)” (Chatman 1990: 9). On this basis, it became possible to establish a “covert narrator”, in charge of fulfilling narrative functions when there is no overt narrator figure claiming responsibility for the narration. Narrative functions include, for example, the selection, ordering, and presentation of events as well as commentary on the latter. Covert narrators can also be identified on the basis of (linguistic) markers of subjectivity. Using the concepts of “chrono‑logic” and narrative function as the basis of narrative mediation, which Chatman calls “narrative transmission” (Chatman 1993 [1978]: 22), we see that narrative is no longer solely language-based, but also possible in (audio-)visual representations: “narratives can be said to be actualisable on the stage or in other iconic media” (Chatman 1990: 114). With this transmedial widening of the scope of narrative research, it has become possible to use postclassical narratological approaches for the analysis of film (Bordwell 1985; Kuhn 2009, Kuhn 2011; Kuhn and Schmidt 2014) and television (Thompson 2003), comics (McCloud 1994; Kukkonen 2013 a, Kukkonen 2013b), dance (Foster 1998), drama (Jahn 2001) and theater (Horstmann 2016, Horstmann 2018) to name just a few. Prince (2003: 58) writes: “The narrative media of representation are diverse (oral, written, and sign language, for example, still or moving pictures, gestures, or any ordered combination thereof).” He thereby rectifies his former take on theater stagings (cf. Prince 1987: 58), no longer excluding them from narrative representations. It is thus a well-defined position to postulate narrative representation in various media including theater.

Theater, however, typically is assumed to be unmediated – but the basis of this generally true assumption is a different type of mediation. On stage, there is no technical process that transforms the actual actors, props, and costumes into for example a digitized, visual representation like it is done in film. The spatiotemporal co-presence of actors and viewers is the main distinctive feature of theater or performance in general with regards to other media (cf. e. g. Fischer-Lichte 2004: 50). One could therefore be tempted to ask, how it is possible that Hamlet’s shirt – the fictive artifact within the storyworld – is in the very same room as the viewers. What we are presenting here is, of course, a misinterpretation of the ontological status of the costume. Not only the actors, but everything else that appears onstage, is part of our real world and represents another world, a storyworld. Thus, Shakespeare’s Hamlet does not have thousands of different faces, bodies or – for that matter – costumes; each of the actors who have played him since 1602 and each of their costumes merely signify Hamlet’s appearance and clothing in his world. From this we can deduce that there is a process of mediation at work in theater, between our actual world (which includes the auditorium as well as the stage space) and the represented storyworld. Due to the fact that theater is sense-physiologically unmediated, this mediation is not a technological mediation (which I define as the transmission of content by technological channels including writing), but nevertheless a narrative one. However, many narratologists (e. g. Rajewsky 2007) and representatives of drama or theater studies (e. g. Pfister 1993) have claimed that it is not possible to talk of mediation in theater performances. This is due to an assimilation of – thus confusion between – the two aspects of mediation: the narrative and the technological.

Although not from a narratological perspective, theater studies still come to a very similar result. In this area of research it is common to talk of the potential iconicity of everything that takes place or can be perceived onstage (cf. Kott 1969; Pavis 1976; Elam 1980; Fischer-Lichte 1992). With reference to Peirce’s semantic theory of signs, the icon has become the most popular sign within theater studies, due to its relation of similarity to its signified. Signs that are originally used as symbols or indices can be transformed into icons onstage and thus fulfill an aesthetic function (cf. Fischer-Lichte 1992). Differently from the signs of written narratives, that are all symbols, the iconic signs of theater have a stronger connection to the actual or the real (which can be understood as the extratextual world), because of their similarity and their non-arbitrariness. Consequently, props as well as costumes are employed to convey similarity to actual phenomena. Once again, it is important to notice that a door or a coat onstage only represents the story-door or the story-coat and that their actual appearance on stage is merely more or less similar to that which is represented – and this general principle is no less valid in a naturalistic stage design.

Having come to differentiate fictionality and narrative representation as well as technological and narrative mediation, let us ask whether the discussed principles hold true for other performative media than theater as well. We do not only find performance in theater, but e. g. in film, too. And film can also represent fictive events. I use “performative media” as an umbrella term that is able to cover rather different media like narrative films and narrative theater stagings beneath the unifying aspect of performativity. At first glance, the differences seem to prevail: films are pre-produced then received a posteriori, whereas a live performance on a theater stage is performed and received at once. A film is a digitized audiovisual artifact, whereas a narrative theater staging is, sensu stricto, merely an intentional communicative artifact, as Currie (2010: 6) uses the term in his definition of narrative. Moreover, Berns’ (2014) twofold understanding of performativity – actual performativity (I) as opposed to textually-prepared and thus imagined performativity (II) – does not cover film either. “Performativity I,” which is “the embodied live presentation of events in the co-presence of an audience at a specific place and time” (Berns 2014: § 1), however, can be broadened in scope in order to include embodied events in media other than theater. It is this broadened understanding of performativity that informs the concept of “performative media” as employed in this article: performative media present (potentially narrative) events that are embodied (i. e. presented corporeally) by actual actors or other forms of experiencing entities (e. g. animals) that exist in our actual world. Instead of drawing a line between narrative artifacts that are either experienced live or technologically reproduced, the umbrella term “performative media” allows us to compare narrative media that present real-world objects and entities within a storyworld. All of these performative media can fulfill the functions of narrative mediation. For film this has been widely acknowledged in research (cf. e. g. Chatman 1990; Kuhn 2009, Kuhn 2011; Kuhn and Schmidt 2014; Thon 2016; Wolf 2002), and I want to emphasize that this holds true for theater as well.

Do the postulated differences between fictionality and (narrative) representation and between technological and narrative mediation hold true for the medium of film, too? In film, we only see a digitized image of Hamlet and his cloak, a visual representation and thus a mediated version of the actual cloak the actual actor is wearing on the actual set. Is there a categorical difference from the David Bowie T-shirt in the mentioned staging of the play in the theater? I want to emphasize the representational similarities here: After the play (or the movie shooting) is over, Hamlet disappears, as do Horatio, the ghost, and the deaths. After all, theatrical performance is a transitory art form as Lessing affirms (1954: 10), and this holds true for the movie shooting itself, too. The actors do not usually continue to impersonate the characters once the performance is over and they are just ‘themselves’ again, performing their everyday identities. But Hamlet’s T-shirt (or his cloak) remains – a ‘real world’ artifact that we can touch, smell or even wear – although it is meant to be a fictive character’s clothing. It seems like at least some artifacts keep their representational ‘power’ longer than humans do.

Whereas a stage version is technologically or sense-physiologically immediate, a film version of Hamlet uses technological as well as narrative mediation. At the same time, however, both versions are narratively mediated, since performative media usually utilize a wide range of narrative functions. It is therefore important to keep in mind that, although the respective actor is wearing ‘real’ clothes in the ‘real’ world, it is still merely a representation of Hamlet’s dress, and thus does not necessarily look exactly like the one he supposedly wears in his world. It is furthermore possible that, for example, the colour of his costume is referred to verbally as different from what we see (“Tis not alone my incky cloake, coold mother / Nor customary suites of solembe blacke” (Shakespeare 2017: 1144; Hamlet 1.2: 77–78). If we as spectators see a David Bowie T-shirt instead, this thus visually contradicts the verbally transmitted content. This can be seen as a commentary on the narrative as a whole, which in turn fulfils one of the narrative functions of mediation discussed above.

That which theater studies refer to as ‘iconicity’ can be applied to performative films as well. Regardless of whether a particular film is more naturalistic (i. e. displays a great similarity of representative and represented entities like e. g. Franco Zeffirelli’s film version of Hamlet [1990] with Mel Gibson in the leading role) than a theater staging (like Turner’s Hamlet staging with Cumberbatch) or not: everything we perceive in the mise en scène represents a storyworld. The representation can only ever be similar to that which is being represented. This has been discussed and is widely acknowledged with regard to verbal representation: Fludernik, for example, states that “representation is not verbatim” (Fludernik 1993: 356; cf. Walton 1990: 181–182). Thon (2016: 86) has argued recently that this general principle is applicable not only to verbal, but also to transmedial representations, and that, therefore, the recipient of a given representational work must apply “medium-specific charity” (2016: 87).^[4] With reference to Currie’s representational correspondence (“for a given representational work, only certain features of the representation serve to represent features of the things represented” [Currie 2010: 59]), Thon suggests distinguishing between “presentational and representational aspects of a given narrative representation” (Thon 2016: 60; original emphasis). He demonstrates that recipients do two things when they interact with representations: “not only [do they] ‘fill in the gaps’ (as it is described by Ryan’s principle of minimal departure), but also routinely ‘ignore’ some aspects of narrative representations in order to intersubjectively construct the represented storyworlds” (Thon 2016: 60–61). In doing so, he writes, they look for external explanations such as the conventions of representation and authorial intention before deeming a storyworld implausible, contradictory or impossible (cf. Thon 2016: 61). The iconicity of performative media (in contrast to the symbolic use of language signs in written narratives) anchor the represented somewhat more closely to reality. This is not an insignificant differentiation: if fans of Peter Jackson’s film version of The Lord of the Rings (2001–2003) buy “the one ring” (as they can do), they expect the replicate they get to look exactly like the ring Elijah Wood wears in the movie – or otherwise they would be more than disappointed if they received, instead, a piece of paper with the words “this is the one ring” on it, although both versions are representations of the fictive ring that has to be destroyed within the storyworld, and although the original “one ring” (Tolkien’s invention) was exactly that: a word on a piece of paper. Fans expect to buy an icon, not a symbol. Even though iconicity, with its similarity principle, links the represented more closely to the actual world, it is nevertheless important to apply the rules of representation and be aware of the global mediation process that is at work and that potentially affects each and every aspect of the performance.

Despite the obvious representational differences distinguishing visual narratives such as graphic novels, caricatures, animated films, computer games, films, or stage performances, they all share functional similarities – i. e. all of them can actualize narrative mediation. Irrespective of whether a visual narrative representation is painted, drawn, digitally produced, or performed live on stage, it can still display a doubly temporal logic of story and discourse.

Considering all of the above, we can surmise that the David Bowie T-shirt worn by Benedict Cumberbatch embodying Hamlet or the black cloak worn by Mel Gibson do not all of a sudden become fictive. On the contrary, the shirt as well as the cloak sit in a relation of similarity to the fictive item of clothing in the two specific Hamlet storyworlds. By virtue of their iconicity, they represent the fictive piece of clothing that could also look different – if there were reason to believe this is the case. Hence, the film or the staging we see is not a univocal presentation of a storyworld, but a narratively mediated representation of it – regardless of its status as factual or fictional. Consequently, (audio-)visual and other multichannel representations – whether live or not­ – just like verbal representations, require the recipient to remain skeptical: they can be just as deceiving as verbal narratives.

3 A transmedial system of narrative representation

How can the narrative functions of mediation be conceptualized in performative media that narrate through several channels? In this section, I will suggest a way to operationalize^[5] the interplay between different channels of a composite narrative medium^[6] for the analysis of single (performed) narratives. Performative media always use multiple channels to deliver their messages and to tell their stories. In filmic artifacts, these channels are, for example, the verbal, the visual, and the auditory. On top of these, live performances in theater can make use of the three additional physiological channels: the olfactory, the haptic and the gustatory, so that no less than six channels can be at work at once in the production of one narrative.^[7] Since a ‘chrono‑logic’ (Chatman’s term) of story and discourse is required to talk about narrative transmission or mediation, we can see that only the verbal and the visual are capable of establishing both a story and a discourse level, while simultaneously making the difference between these two levels (the doubly-temporal logic) apparent. Much research has been carried out on the narrativity of music or sounds (cf. Chenoweth 1986; Longacre and Chenoweth 1986; Abbate 1991; Levinson 2004; Seaton 2005; Hansen 2010; Modrow 2016), and it has become clear that, although music contains some sort of discourse, it cannot at the same time establish a story level with a different temporal logic, and thereby fulfill the narrative functions of mediation as discussed above. Similar observations can be made of the other three channels, which have so far received no attention from narratologists. The reason for this might be that they are rarely used in the production of narratives and narratologists therefore do not feel the urge to subject them to theoretical analysis. But like music and sound (which establish the auditory channel), haptic sensations, smells, and tastes can influence the multichannel narrative representation, although they cannot be representative on their own. It is this influence – the interplay between single channels and the narrative effect of this – that is the focus of my interest here.

The narrative effect of a single channel on a multichannel narrative representation is contingent upon its degree of narrativity. The narrative effect is not to be confused with other effects; in fact, olfactory or haptic perceptions, for example, might be overwhelming enough to overpower all the other information that the channels transmit. But this does not necessarily have a narrative effect, since the degree of narrativity of both olfactory and haptic channels is actually rather low. What are the degrees of narrativity in particular? It is possible to distinguish four levels here.

A single channel’s degree of narrativity depends on the number of narrative conditions it fulfills. Narrative conditions can be assigned to four different levels that result from different, pre-existing conceptualizations of narrativity. The first level consists of cognitive conditions that, for the most part, describe reader-related cognitive frames allowing us to ascribe narrativity (or “experientiality”, as Fludernik [1996] terms it; cf. Caracciolo 2014) to certain artifacts – we can define this as narrativization through the recipient. Since this kind of narrativity is not intrinsic to an artifact but rather ascribed by the reader,^[8] there are rarely any conditions the artifact has to fulfill to be potentially narrative with regards to cognitive frames. Levels two and three are story-based definitions of narrativity, which have been described as changes of state and eventfulness or, in Hühn’s (2013: § 1) words, event I (“any change of state explicitly or implicitly represented in a text”) and event II (“with certain features such as relevance, unexpectedness, and unusualness”). Locating a change of state or eventfulness means to describe features that appear more or less identifiable in the text. But in these cases, we assume that a certain artifact can be ascribed narrativity within a cognitive frame, and thus fulfill the conditions of narrativity level one, as well. Hence, levels two and three also contain the conditions of the respective levels below them, whereas level two fulfills more conditions than level one, and level three fulfills more conditions then level two. The highest level (or level four) on this narrativity chart is the one discussed in the first section of this paper: narrative mediation. Since narrative mediation requires a story and a discourse level, it fulfills the discussed narrative functions of selecting, ordering, presenting, and commenting that take place on the level of discourse, as well as the conditions that are required to locate eventfulness, changes of state or narrativization through the recipient. Level four thus comprises the highest degree of narrativity. Consequently, channels fulfilling these kinds of narrative conditions in a narrative representation have the highest narrative effect on the whole multichannel narrative.

As already argued, narrative mediation can only be achieved verbally and visually. Auditory signs usually are able to signal a change of state. Conversely, olfactory, haptic, or gustatory signs have either no narrativity at all, or are narrativized by the receiver and thus fulfill the conditions of basic readerly ascribed narrativity. If different channels narrate the same or similar content – and thus have a “complementary, meshing, polarizing, illustrating, [or] paraphrasing” relationship instead of being “contradictory,” or “disparate” (Kuhn 2009: 265) – their respective degree of narrativity can be affected by this interplay. This “analogous” relationship, which can be “overlapping” or “complementary,” as I have described it elsewhere (Horstmann 2016: 233), is the default case in a multichannel narrative representation. Disparate relationships usually require prominent marking to fulfill the intended effect, and stop the recipient from finding naturalizing or harmonizing explanations and thus activating the principle of minimal departure as well as medium-specific charity, as discussed above. Consequently, the auditory channel can achieve eventfulness (level three) by means of analogous relationships between the single channels, and olfactory, haptic, and gustatory channels can achieve a change of state (level two) with the help of those channels that lead the narrative representation. The following chart (cf. Table 1) displays a narrativity matrix of multichannel narrative representations that shows the highest possible degree of narrativity for each channel within a system of narrative representation. The chart can be reduced to the number of columns each narrative representation requires: contemporary films usually work with the first three channels, ballet usually with visual and auditory channels, and so on. The system is also applicable to non-performative media, like e. g. comics, that use the verbal and visual channels.

Table 1

Narrativity matrix of multichannel narrative representations

By conceptualizing or operationalizing the narrativity of multichannel narrative representations in this way, we can not only determine each channel’s degree of narrativity, but also generate an overall degree of narrativity and calculate the narrative effect a contradictory channel has on the whole narrative representation.

An easy and plausible example from the medium of film may illustrate the heuristic value of this procedure. Let us imagine a situation in which the visual channel shows a party with a number of people dancing enthusiastically, while the recipient can hear a voice-over of the main protagonist telling him or her about this same party, the loud music and the dancing. The auditory channel in this case only transmits the calm voice of the character; we do not hear the loud music from the party that is being described. In this sense, the auditory channel has a disparate relationship to the other two channels. But instead of constructing a contradictory storyworld or ascribing unreliability – or even mental delusion – to the voice-over narrator, the viewer applies medium-specific charity and concludes that the music is missing due to representational economics: we do not hear the loud music because it would probably prevent us from hearing the voice-over. Furthermore, the visual and verbal narration of the party is enough to make us believe that what we perceive is a fictional fact. The auditory channel is not able to subvert the analogous relationship between visually and verbally mediated content because it has a lower degree of narrativity. The situation changes if there is no such analogous relationship between the verbal and visual channels, but, let us say, between the visual and auditory channels. In this second example, we as film viewers might see the same situation described above, but this time with only the main character dancing and no one else around her. The visual channel suddenly contradicts the verbal narration and is supported by the auditory channel that does not transmit any loud music. In this case, we are much more likely to assume that the voice-over narrator is either unreliable or deluded. Although the verbal and visual channels both fulfill the conditions for the fourth and therefore highest degree of narrativity – whereas the auditory one does not, but rather can only be narrativized by the recipient (degree one) as it illustrates what is not being transmitted – it is the auditory channel that is the deciding factor here. Let us operationalize this a little further. If each of the four levels of narrativity is ascribed a point value, the resulting basic rule of calculation is: channels that work analogously are added and channels that work disparately are subtracted. In the second version of the given example, the verbal channel that tells us about the party receives four narrativity points (narrative mediation). Since the analogously working visual (four points) and auditory (one point) channels are both contradicting the verbal channel, they together have a higher degree of narrativity and thus negate the verbal narration: 4—5 = —1.

4 Conclusion: advantages and limits of a system of representation for multichannel media

Over the course of this article, I have shown that there are two important terminological distinctions to be made, which have a number of consequences in dealing with the study of narrative. Firstly, fictionality and representation should not be confused, as they can occur independently from one another. This distinction explains the fact that real artifacts (such as a costume used in a film or onstage) do not become fictive all of a sudden (a boundary that cannot be crossed), but rather representative of fictive artifacts. The transmedial discussion of performative media establishes this as a general representational principle, applicable regardless of whether the signs are used symbolically or iconically. Secondly, we must separate narrative mediation from technological mediation. This allows us to discuss not only the differences but also the similarities between the two performative media of film and live theater on safer ground, beneath the umbrella term “performative media.” Whereas film is mediated both narratively and technologically, theater stagings are more sense-physiologically immediate. However, live stage performances can nevertheless fulfill all kinds of narrative functions constituting narrative mediation. Hence, it becomes possible to compare filmic and theatrical narrative representations within the same analytic frame.

The terms “narrator” and “narrative instance” are frequently associated with fictional discourse. Due to the fact that there is a distinction between representation and fiction, and because representation is independent of fictionality, it seems more appropriate to speak of a system of representation. The dynamic narrativity matrix devised over the course of this article can be applied flexibly to a wide range of multichannel media. After defining which channels the medium in question consists of, these channels can be analyzed according to their respective levels of narrativity. Moreover, I have shown how the interplay and influence between the different channels can be taken into account, as well as how, for example, disparate relationships can be evaluated.

However, it must be said that the use of this system is limited to the realm of representation. Only if a performance is neither fictional nor representative can we assume that the actions staged are narratively immediate or unmediated. This is the case in most performance art events, as well as in television news, talk shows, YouTube vlogs, parts of non-scripted reality TV shows, etc. In these and similar cases, it is less appropriate to discuss a system of representation, although there are a number of channels transmitting content, and technological mediation can be identified. However, in media that claim to be non-representational, we can only attempt, if possible, to detect or compare lower degrees of narrativity. It is clear, therefore, that a system of representation and a narrativity matrix would be of little heuristic value in these particular cases.