Expansion and Densification of Fictional Spaces. Transtextual Characters in Arthurian Romances and Grimmelshausen

1 Characters and Storyworlds as Mental Models

Behind the words of the text, there is not simply nothing. A narratology informed by cognitive sciences assumes that readers make additions according to their knowledge of the world in order to fill gaps in the information structure (cf. Brosch 2018, 435–437; Herman 2014, 49 sq., 56 sq.; Palmer 2004; Schneider 2000). Within »the umbrella field of the cognitive sciences« (Herman 2014, 47), this basic notion has been elaborated in more detail, among others, in the field of situated cognition theory (cf. Smart/Heersmink/Clowes 2017, 256–274). According to the theory of »distributed or scaffolded cognition« (Heersmink/Sutton 2020, 140; cf. Sutton 2006), past reading experiences will scaffold a given reading process and the generating of meaning. Theories of »distributed cognition« presuppose the idea of a cognitive ecology that meaning in texts is generated always in interaction with its closer or wider environment, which is built, among other factors, by the »semiosphere« (Lotman 2005) of cultural memory and not least other texts.

Against this theoretical background, a position has found acceptance in recent narratology according to which textual characters can be understood as text-based mental constructs of the readers (»textbasierte mentale Konstrukte von Rezipienten«: Hillebrandt 2018, 164; cf. Reuvekamp 2014, 112–117; Jannidis 2004, 177–185; Schneider 2000, 35–98). However, the same can also be assumed for narrative worlds and the actions that take place in them as these are also constructed and received under inferential formations (cf. Jäger 2023, 151; Thon 2016, 35–70). For transtextual characters, this means that their (text)transgressive dynamics shift the semantic relations between texts and their textual environments. Thus, familiar narrative characters not only take on new contours themselves when they are evoked in another storyworld, but they also change the spatiotemporal structure of the latter (cf. Jäger 2023; Lăcan 2019, 143). The new world they enter is now one that provides the space-time setting for the places and events that are connected – remembered, associable – with a character.

For narratological analysis, these general assumptions on storyworlds with transtextual characters can be fruitfully related to a text-genetic model proposed by Wolf Schmid (Schmid 2010, 175–215). According to this ideal-typical model, the ›story‹ (b) is already the result of a selection from the ›happenings‹ (a). The story, in turn, is brought into the form of the ›narrative‹ (c) through processes of disposition (linearization, permutation), which is then presented through the ›linguistic form‹ (d). In this model, the reference point for transtextual characters would in principle be at the tier of the happenings (a), which are

the amorphous entirety of situations, characters and actions explicitly or implicitly represented, or logically implied, in the narrative work. The happenings understood in this way form a continuum that is spatially fundamentally unlimited, can be endlessly temporally extended into the past, can be infinitely divided internally, and can be concretized into infinite properties. (Schmid 2010, 190)

If characters are to be understood as mental constructs based on textual structures and inferences, then the tier of the happenings corresponds to the notion that not only explicitly narrated plot details come into play in a new storyworld, but also plot details that can be associated with a character.

In this article, I want to work out two different types of space-time structures in storyworlds that can be correlated with transtextual characters. To this end, I will take the MHG Arthurian romances on the one hand and Grimmelshausen’s 17th-century picaresque novels on the other. To be sure, the point is not to compare medieval romances and early modern novels per se, nor is it to compare the transtextual characters appearing in them. The argument is that the space-time structures in the respective storyworlds are shifted when already familiar characters are recalled in them. To shed more light on the potential effects of transtextual characters, I analyse two groups of texts which are comparable in one respect: Their storyworlds merge into the continuum of a common frame of reference; the characters enter new texts but move within the same fiction.^[1] However, quite different space-time structures are developed in these worlds. The article aims at ideal types in the Weberian sense (Weber 1988, 190–209); it does not dispense with historicization, but rather relates considerations of transtextual characters to the different historical contexts of text production and reception. I will contrast these contexts beforehand.

The courtly romances of the High Middle Ages circulated in relatively closed communication communities, while the novels published in the early modern period faced the relatively open communication communities of an anonymous book market. Both groups of texts, which do not need to be further specified for the moment, had to prove themselves and assert themselves under very different media-historical conditions.

As far as we know, courtly romances were written at court, sponsored by patrons, and addressed primarily to court society (cf. Bumke 1986, 638–677; Bumke 1979). Under the conditions of a manuscript culture, they were written over a long period of time, quite a few will have remained unfinished, and in individual cases we know of considerable interruptions between the start of a romance project and its completion (Veldeke’s Eneas romance). Once completed, the texts could only be disseminated to a limited extent, both in terms of the number of manuscript copies and their geographical reach. In the reception of courtly romances, the primary target audience ranked much higher than would be imaginable under the conditions of an open book market (cf. Green 1994). Of course, it is rarely possible to trace which texts were presented at which courts (and to what extent and at what stage of completion; cf. Bumke 1986, 721–725). In terms of the aesthetics of production, the romances were aimed at a rather exclusive audience, ideally with the same knowledge of the material and the texts as the poet himself.

It can be assumed that these courtly societies had a relatively clearly defined horizon of norms, and that the common standards of value were articulated in the common stories. Traditional and time-tested are the materials of the chanson de geste tradition; and heroic-epic as well as legendary or biblical-epic materials, which could be assumed to be familiar to the audience, could be reworked in a courtly manner. The courtly romances, on the other hand, represent a more recent development, especially the adaptations of older material, the Arthurian epics, and the Minne romances.

Another historical and cultural specificity of medieval courtly poetry is what might be called its epistemological function. In pre-modern culture (and in this respect there is no categorical difference between scholarly-clerical and lay groups), no strict distinction was made between fact and fiction. Characters in fictional texts could resemble historical persons (cf. Nesselroth 1996, 141). While this is also true of modern narrative (cf. Margolin 1996, 116 sq.), in the pre-modern period fictional characters were also reliable in real life. Thus, in addition to historical persons, characters from ancient mythology, but also from new fictional texts, can serve as exemplary figures; the highest dignity is, of course, accorded to biblical figures. What matters is not an ontological distinction between real persons and fictional characters, but that something can be shown through them. Exemplary figures enter a cultural reservoir from which they can be retrieved for argumentative purposes and functionalised in different contexts (cf. Friedrich 2012). The topical thinking of premodernity (essentially von Moos 1988) makes it likely that transtextual characters – once they have been given the authority to appear in another text – can also take on pragmatic functions. Cultural knowledge is bundled in them, as they offer orientation, serve as examples (worthy of imitation or warning), and demonstrate how to (successfully or unsuccessfully) behave in the world. Against this epistemological background, an aesthetics of confirmation develops, which must be distinguished from the aesthetics of deviation characteristic of modernity (cf. Fricke 1981). In the constant reference to the ›original‹ appearance of a character, semantic charges will occur to varying degrees, with its original profile remaining predominantly the standard, which itself does not undergo any change (unlike in modernism: cf. Lăcan 2019, 150 sq.; Margolin 1992, 45). What matters in these processes is not the reference to an original, but to the reality – guaranteed by tradition – of meaning and significance of the character. The function of courtly poetry is not least to secure meaning for relatively homogeneous circles of audiences under the conditions of face-to-face communication.

By contrast, early modern novels were written for an anonymous book market that operated according to capitalist rules. The stories are new (and they have to be to attract attention), so their familiarity cannot be taken for granted, all the more so since the unknown readership is to be understood as open and heterogeneous (cf. Röcke 2004). However, when old, basically familiar material is no longer retold or taken up, but new stories enter the market, the function of the texts shifts. They can no longer refer to shared stories and thus serve as a medium of participation in shared values. The point of reference, then, can only be the stories in their published form, i.e. the texts. The individual text in its linguistic form takes the place of the collectively shared dimension of meaning. The function of texts, which comes to the fore under the conditions of distance communication, is the securing of meaning.

Under these different media- and literary-historical conditions, the hypothesis goes, different narratological moments and functional-historical dimensions should also be connected with transtextual characters. In a first step, I examine the characteristic way in which characters from Arthurian romances reappear in other texts or are named as belonging to their storyworlds. In doing so, their deeds are remembered without being given a precise place in the past (Section 2). In a second step, I discuss a case in which an Arthurian event that the audience might already know is taken up and placed as a background plot alongside the main plot of a new romance (Section 3). The coordination of the two plots does not require much narrative effort. This is due to the spatio-temporal specificity of the Arthurian worlds, as I will show in a third step by contrasting some of Grimmelshausen’s texts (Section 4). In these, too, characters appear again and again; however, the storyworld is not supplemented by new plot spaces, as in the Arthurian texts, but new plots are inserted precisely in such a way that previously undefined parts of the plot are filled in.

2 Times of the Topical: Fame (Lanzelet, Erec)

An overview shows that only the protagonists of the older MHG Arthurian romances, which set the standard for the 13th century, are regularly mentioned in the later texts. Even the hero of the early ›post-classical‹ romance Wigalois only appears once, in the late romance Gauriel of Muntabel, and only there are the two Knights of the ›Blossoming Valley‹, Daniel and Garel, mentioned once each. One obvious reason for this is the chronology of the texts. Of course, the main characters of the relatively late stories have fewer opportunities to be mentioned again in later romances, but the purely statistical explanation probably cannot yet cover the findings. My thesis is that the significant preference for the ›classical‹ characters lies in a specific constitution of the Arthurian narrative world, which will be described below.

The ›post-classical‹ protagonists serve as experimental fields in which the rules and functioning of the Arthurian narrative can be tested. This is done on the basis of the well-known Arthurian inventory, which includes characters with a typified core (Arthur, Ginover, Keie, Gawein) and a few places such as Camelot, Karidol, and possibly Nantes. This ensemble is surrounded by a number of characters, most of whom are probably strongly typified, but some of whom are also individual in the sense that they are associated with singular stories. This is what I mean by narrative individuality. However, such narrative individuality is not invoked when characters from one Arthurian romance are recalled in another (on intertextual allusions in the Arthurian romance, cf. Schirok 1988). This applies also for characters who in fact can display a broader variety of actions in the course of several texts. Even with characters such as Gawein or Arthur, it is still a topical knowledge of themselves and their actions that they carry with them when becoming transtextual. I want to show this with Lancelot and Erec.

3 Coordination: Times of Events (The Kidnapping of the Queen)

Pleier’s Garel of the Blossoming Valley begins with a double challenge to the Arthurian court. First, the queen is kidnapped, and all the knights who try to prevent it fail miserably. Immediately afterwards, a messenger arrives with a declaration of feud from a giant king. The first of these two shocks to Arthur’s self-image may be familiar to an audience from Hartmann’s Iwein, and the narrator explicitly refers to this text in the Garel romance: »Now listen to a strange story. Hartmann von Aue has already told us, as a true story – in a well-known book called ›The Knight with the Lion‹ –, that Arthur was robbed of his wife, and how it came about« (Ga 31–38).^[5]

After this announcement, the scene is given only 50 verses of space; it is presented in a concise and summarising manner, literal speech is completely absent. Pleier is not simply retelling what we have already heard elsewhere; he is explicitly summarising what Hartmann had told (Ga 44: ›as I heard the story‹; »als ich daz maere hân vernomen«). Invoking the very abduction of Ginover already mentioned in Iwein and placing it within Pleier’s own storyworld creates a common reference point which is not Hartmann’s text, but the event that allegedly took place and was already reported by Hartmann (which presupposes that the storyworlds of Hartmann’s and Pleier’s romances are identical). In the quoted verses, the narrator joins with his audience to form a ›we‹, thus creating (or claiming) a common ground; anyone who does not know Hartmann’s »well-known« text must feel excluded from the exclusive circle of ›those who know‹.

The contrast to the immediately following challenge, which serves as a starting point for the foreground action, illustrates the different speech patterns of summation and narration. The scene surrounding the declaration of the feud is given five times as much space (Ga 219–476); the giant-like messenger is portrayed in detail (Ga 240–272), followed by a dialogue between him and Arthur in direct speech with several changes of speakers, the king’s reaction, the courtly behaviour of the giant, and the discussion at court about how to proceed. A well-informed audience might also recognise this scene; in it, the initial demand of a giant in Stricker’s Daniel of the Blossoming Valley has its counterpart (the almost identical names of the protagonists alone suggest that the Garel romance refers directly to Stricker’s Daniel). However, this intertextual allusion takes place beyond the narrative world, at the level of textual composition. Pleier is telling something here that is not necessarily new to the audience, but new in the epic world, and he does so in the detailed way that can be expected.

The appearance of the protagonist is also new; Garel is a nephew of the king who arrives at the helpless court immediately after Ginover’s kidnapping. He advises his uncle to take the disgrace in stride and is himself persuaded not to pursue the kidnapper but to remain at court; this scene alone is more than twice as long as the account of the kidnapping (Ga 91–218). After the declaration of the feud, Garel will follow the giant messenger to explore the challenger’s land. Pleier’s romance begins with an event that continues in the background of the main plot, and which may also take on a topical character in the knowledge about the Arthurian world. At the end of Garel’s story, Ginover will have been freed by Lancelot. Pleier connects his romance to an already known narrative world and expands the latter with his adaptation. But apart from this nexus, the two storyworlds remain separate. The peculiar convergence of the two storyworlds leads to a specific asymmetry: Hartmann does not know that immediately after the kidnapping, a giant demands the complete subjugation of the Arthurian court. Within a cognitive ecology, it is not only cultural knowledge and past reading experiences that scaffold a given reading process. Conversely, fresh impressions can also reshape knowledge from previous readings. This is the case here: not only is the storyworld of the Garel romance the same as the one in which Hartmann’s Iwein learns of the abduction of the queen, but retrospectively the world in which Hartmann’s Gawein frees Ginover is also the same as the one in which Garel follows the giant messenger in the meantime. The structures of a new text reshape already familiar storyworlds and their space-time orders. Under the media-historical conditions of a semi-oral aristocratic culture and reception via the spoken word, however, the audience will hardly have realised this specific asymmetry between earlier and later texts. It is more likely that later additional information was integrated into the knowledge of the epic world as mosaic stones, without any contradictions between the respective textual information being perceived.

Chrétien has already linked two of his romances in a similar way. In Yvain, he has two characters summarise the events surrounding Ginover’s abduction, which he had already recounted in detail in the Lancelot romance. In the later Yvain, the brief reminiscences serve primarily to explain that Gawein/Gauvain is absent from Arthur’s court: he is currently pursuing the kidnapper. Here, the fact that the focus of the Lancelot romance was on the unknown Knight of the Cart is completely pushed into the background. Lancelot does not appear and is only mentioned twice as belonging to the storyworld. Since Chrétien already had no use for Lancelot in this context, it is not difficult for Hartmann to omit the Knight of the Cart completely in his adaptation of the romance, and he presumably did so because he could not assume that his audience had any knowledge of the Lancelot romance.^[6] In Chrétien’s construction, as in Hartmann’s, the action sequences of Lancelot and Yvain take place simultaneously; Pleier, for his part, will insert the plot of his romance into this virtual time continuum (and the fact that it is Lancelot who saves the queen in Garel shows that in the later 13th century the German public could be assumed to know the Knight-of-the-Cart plot, be it directly from Chrétien or transmitted by oral tradition). However, it is but to a very limited extent that one can speak of a space-time continuum; the protagonists of the two – or three, respectively – action sequences do not cross each other’s paths, nor do they arrive at places that the others have already passed.

On the one hand, these examples demonstrate a skillful coordination of plot sequences, which is rather the exception in Arthurian romance. On the other hand, these narrative integrations remain peculiarly inconsequential in that the epic worlds, apart from the single connections, remain separate. In converging, the storyworlds are but loosely connected; this is a specificity of Arthurian epic which is to be further elaborated in the following. In order to describe the range of such coordination and its prerequisites in more detail, I will make a leap from the Arthurian narrative cosmos to Grimmelshausen’s picaresque world.

4 Coordination: The Time of Storytelling (Grimmelshausen)

In Book V of Simplicissimus Teutsch (1668), the protagonist has accompanied his best friend to a spa in the Black Forest; himself in perfect health and to escape boredom, he pursues the ladies. Only a few lines of text are given to one of these love affairs: a lady who pretends to be a noblewoman but apparently is not. So Simplicius soon breaks off the liaison; it thus remains a mere episode in the novel. Simplicius, by this time married and the owner of a manor, is more concerned with something else: an unexpectedly rich blessing of children.

[…] when my wife was delivered, the maid was also brought to bed: and the child that she bare was like unto me, but that which my wife had was like unto the servant, as it was cut in the image of his face. Nay, more! for the woman of whom I have written above, the same night she caused one to be left at my door, with a written notice that I was the father: and so I had a family of three at once, and could not but expect others to creep out of every corner, which caused me not a few grey hairs. (ST 481,16–26)

Here, then, is the protagonist with a child who is not his own and another one who is his but should not be. The mutual infidelities follow a familiar pattern of tricks and counter tricks; the focus is on Simplicius’ conflict with his wife, who quickly proves to be an ineffectual housekeeper but an even more sturdily drinker. The fact that the »lady« foists another child on him follows the logic of escalation and provides the preliminary final tableau with the threatening vision of a crowd of children, which may remind today’s readers of a scene from Monty Python’s The Meaning of Life.

In the sequels to Grimmelshausen’s novel, the somewhat burlesque confrontation is taken up again, and the focus shifts from the drinking farmer’s daughter to the actual antagonist, the Courasche, after whom Grimmelshausen’s first Simplician novel with a new first person narrator is often named. Courasche uses her report to accuse Simplicius, because she wants to take revenge for him making their relationship known to the whole world in his life story (i.e. Simplicissimus Teutsch), making her pregnant in the spa und finally getting rid of her by a malevolent trick (cf. Cou 22,16–23).

Courasche was thus the supposedly noble lady whom Simplicius in his biography had mentioned only in passing. Her entry into the storyworld of Simplicissimus Teutsch sets different accents from those seen in the examples from the Arthurian romances. In both cases, the information provided retrospectively changes the storyworld, in which more actions had taken place than previously known. In Arthurian epic, however, they are organised in parallel plots (background plot, main plot). Thus, at one point in the plot, two storyworlds are connected that otherwise remain separate. The entry of a transtextual character expands a given storyworld; in the audience, new listening impressions will reshape the already familiar storyworlds from earlier texts. This is possible without great narrative effort because the Arthurian storyworlds are not systematically structured, but primarily provide the plot stations for the protagonist without their dimensions and all components having to be defined. In Grimmelshausen, on the other hand, the individual stories, those of Simplicius and Courasche, are confronted with each other. The storyworld is not expanded but, on the contrary, condensed, insofar as the same space has to contain more action. Here too, the new information retrospectively reshapes what is known from previous readings. However, because the happenings do not take place in parallel, but the stories are based on the same happenings (Schmid), the perspectivity of narrating plays a greater role here.

Different perspectives can be combined with meticulous distributions of information, allowing Grimmelshausen to continue the trick-and-countertrick sequence. Courasche reports that in fact it was her who deceived her lover at the time:

My maid had bundled up […] a little son whom she […] also brought happily into the world outside the town; she had to have named him Simplicium, although Simplicius had never touched her in his life. As soon as I learnt that Simplicius had married a farmer’s daughter, my maid had to wean her child and carry it […] to Simplicius’ house; she then laid it outside his door at night, with an enclosed written report that he had procreated it with me. (Cou 133,19–34)

From a distance, Courasche, thirsting for revenge, is delighted to see how Simplicius not only receives a severe punishment from the authorities, but above all has to put up with the whims of his duped wife. The first-person narrator of Grimmelshausen’s next novel, Der seltsame Springinsfeld (›The curious Hopalong‹), tells the same story. He is a burnt-out scribbler to whom Courasche had dictated her life story in the indefinite past (presumably some four months ago). In Springinsfeld, he meets Simplicius and his old companion, the eponymous Springinsfeld, whom readers have already met in Book III of Simplicissimus Teutsch and who in turn also has a past with Courasche. To these two, the writer repeats what Courasche had told him about her deception of Simplicius (Sp 182,25–29).

Grimmelshausen has his first-person narrator repeat exactly what was already known in the chronology of the storyworld and also in the real life chronology of the three novels. He does not limit himself to allusions, as was possible facing the relatively closed circle of addressees of the Arthurian romances, but recalls the relevant information and makes it available to those of his readers who had not read the previous novels. Knowing exactly what Courasche knows is a prerequisite for the point that follows. When the scribe reports Courasche’s glee because her former lover »adopted and raised the child to his great mockery: and also let himself be severely punished by the authorities« (Sp 182,31–33), Simplicius interrupts him. In fact, he says, everything was different back then: Even if Courasche imagined that she had tricked him, she had actually »done me the very best service […]; since when I had an affair with her, I have slept with her chambermaid more often than with herself; and I would rather have my Simplicius […] born of this chambermaid than of a loose gipsy woman« (Sp 183,7–16).

Thus, Simplicius has the last word in the end. The poetological, ethical-moral, and gender-theoretical perspectives of this constellation need not be pursued further. In the perspective adopted here, it is more important that the sophisticated play with the information once given is not limited to these three novels, which Grimmelshausen has elsewhere closely related one to another (cf. Berns 1988). It takes place in essentially the same way in Grimmelshausen’s Ewig-währender Calender (›Everlasting Calendar‹; 1671), which does not belong to the ›Simplician Cycle‹ nor even to the genre of the picaresque novel. The fiction that the manuscript of this calendar was created by Simplicius and later discovered and published by a visitor to his manor coincides exactly with the chronology in Book V of Simplicissimus Teutsch (cf. Breuer 2020, 99–102). Simplicius (as well as Courasche and Springinsfeld) will also appear in other texts by Grimmelshausen; each of these appearances enriches the storyworld with further details. Again, an asymmetry develops insofar as the later texts provide information about the storyworld that the earlier texts had not yet contained. Yet, under the conditions of a reception in the reading process, this may have been easier for the reading public to perceive than the analogous case with the Arthurian romances discussed above.

Even without tracing these textual strategies in detail, the previous analyses allow for a contrastive comparison of the transtextual characters in the Simplician novels with those in the medieval Arthurian romances. Finally, they will be used to draw conclusions about their respective media-historical significance.

5 Characters Shifting Storyworld Structures

Like Pleier, Grimmelshausen coordinates (more precisely, co-constructs) actions from different texts that refer to the same storyworld. In principle, the same thing happens narratively. Since never everything can be told (cf. Schmid 2023, 3–12), storyworlds are necessarily incomplete. They require mental completion, based on inference and pre-structured by the textual information. Both Pleier and Grimmelshausen make use of this principle in terms of production aesthetics. In his romance, Pleier subsequently fills in these gaps, adding further spaces to the epic world and placing new plots within them.

In the Grimmelshausen episode, on the other hand, there is only a very limited expansion of the content, and one gets the impression that the new punch lines are only created by revisiting the same scenes from a different perspective. In Simplicissimus Teutsch, there was no mention of the protagonist having an affair with the maid of the supposedly noble lady, nor of the fact that he got rid of her with particular malice. Both details are added in the sequels. It would be difficult to plausibly add much more to the story. This makes it possible to characterise the narrative worlds of Arthurian poetry and Grimmelshausen more precisely.

In the Arthurian romances, one has the impression that in each text the characters act in a new space that contains fixed points, but whose topological order is not transferred into topographical consistency. This begins with the first Arthurian romance: Erec has never heard of Tulmein Castle, its annual tournament, its rules, or the sparrowhawk as a prize. Just a day’s ride from the Arthurian court, he enters a world completely unknown to him. This, in turn, plays no further role in the series of Arthurian romances; Tulmein is mentioned only once more in Arthurian poetry, when the narrator of Parzival recalls the events of the first part of Erec. No other Arthurian knight ever enters this castle again, nor Brandigan Castle, nor the fountain kingdom of Iwein’s wife. It was not until the 13th century, when the romances increasingly constituted meaning by assembling set pieces from the texts of the first generation, which were already perceived as canonical, that the stages of the first protagonists’ actions regained a place in the new epic worlds. For the most part, they are remembered as places where action took place; for example, the tournament at Tulmein Castle. Not only the chronological when, but also the topographical where of the actions can remain underdetermined, because the characters are always to be understood as exemplary and carry out actions in which exemplary things can become visible.

The Arthurian characters bring very little with them when appearing in other texts; often only their names, and if more, then striking plot connections (essentially always the same ones). These can be recalled at will, but then they serve less to characterise the characters with whom they are associated than to illustrate general facts. What a courtly audience already knows about the Arthurian protagonists is to be confirmed as reliable. The single elements of the topical are intended to be consensual, i.e. reliable; accordingly, the display of perspectivity is precisely not a characteristic of Arthurian retellings. The plots are to a certain extent detached from their characters; when Erec reappears, the narrative individuality he had acquired in his story is in most cases simply irrelevant. Of course, one knows about his story, both inside and outside the text. The allusions of the narrated characters themselves are directed at the interlocutors in the storyworld, and within the fiction it is only logical to assume that they have knowledge of past events. At the same time, however, they are addressed to the listeners of a lecture situation, who, as witnesses of the allusions, must be able to classify them. For the presumably small and exclusive group of listeners at court who came into contact with the Arthurian stories (be they written or disseminated orally) and/or who were able to become acquainted with the romances that had been written since the 12th century, a familiarity with the plots can be assumed. In any case, the narrators’ brief allusions prove that such familiarity is expected; for the courtly audience, the characters of the Arthurian romances were »common cultural property« (Margolin 1996, 116).

In summary: In the Arthurian romances, characters are created that have traits of narrative individualisation, but are kept exemplary in epistemic terms. The storyworlds of these romances are open for, in principle, infinite expansions. The texts keep available the Arthurian topics for a collective memory in which the characters of the storyworlds participate just like the real addressees. This constellation of character profile, space-time coordinates and the semiosphere of the Arthurian topics is a specific feature of Arthurian poetry. It is here that the conditions for the Arthurian characters to become transtextual are established.

According to this, the consideration of what constitutes the reference category for transtextual characters developed at the beginning of this article is to be readjusted for the Arthurian novel. The form of the intertextual references suggests that the basis on which the characters become transtextual is not so much the shared texts, but the stories underlying the elaborated texts. The basic category is, in Schmid’s terms, the narrative material as it has been segmented for the story. At this tier, allusions are made to the events of the earlier Arthurian romances; their reference points are the narrated actions, not their linguistic representation. When a character is named as belonging to another textual world – it does not necessarily have to appear –, the potentially »infinite properties« (Schmid 2010, 190) of the events are not brought to the fore. On the contrary, even the particular circumstances of the actions invoked recede behind the action condensed into iconic cores, so that the later romances give an »overall impression of a character«^[7] (Lancelot, the Knight of the Cart; Erec, who is warned by Enite). The form of their presentation and their linguistic design play as little a role as the »amorphous entirety« (Schmid 2010, 190) of everything that was logically implied by the characters in the romances but not narrated.

The Black Forest, on the other hand, as Grimmelshausen tells it, is a world that has already been filled. It cannot simply be expanded, just like new realms open up in the Arthurian romance. In Grimmelshausen’s work, the addition of content is achieved not by expansion, but by internal differentiation, condensation, and perspective. It is not the time of the storyworld but the time – or at least the chronology – of the narration that comes to the fore. For the reader can hardly do otherwise than to accept the most recent representation of the facts as the valid one – or he must question the reliability of the Simplician narrators in general.

In Grimmelshausen’s texts, the narrative individuality inherent in the characters is enhanced as they enter new texts by the successive addition of their hitherto untold actions. At the same time, the common storyworld seems to become denser; periods of time are filled with events, spaces are temporarily occupied by characters. Here, it becomes clear that the tier of happenings is the reference point for transtextual characters. The stories of Simplicissimus Teutsch, Courage, and Springinsfeld represent selections from the same reservoir of events. Against the background of the third novel, it can be seen that the events of the first two novels were richer than their respective narrative presentations made apparent; they are »internally divided and can be concretised into infinite properties« through renewed selections (Schmid 2010, 190). In the narration of the events, perspectives and evaluations collide, which are revealed as provisional in the linearisation of the texts. The different narrative versions reveal the provisional nature and uncertainty of all knowledge with a limited perspective.

If storyworlds represent text-based mental models (cf. Thon 2016, 35–70), which are supplemented by inference processes according to the principle of »minimal departure« (Ryan 1991), then they can coalesce for the audience/the readers in the process of storytelling when already familiar characters reappear or are named as belonging to the storyworld. Transtextual characters can shift the space-time structures of epic worlds. Thus, analysing transtextual characters can provide arguments for various theories that assume that texts do not construct meaning exclusively by themselves and it may not least support theories of situated cognition as briefly indicated above. The examples of the Arthurian epic and Grimmelshausen’s narrative cosmos show different modes of semantic charging that are based in, among others, function, media history, and epistemology.

The potential effects in historical reception processes are also different. The audience of an Arthurian romance probably had different ideas of a given storyworld. Depending on prior knowledge, the Arthurian world for the individual listener may be the one to which not only topical places such as Camelot or Nantes belong, but also the castles of Tulmein (Erec) or Pelrapeire (Parzival). Yet, such prior knowledge is not necessary; rather, the insertion of these realms into the Arthurian narrative world is to a certain extent non-binding. This is due to the specific space-time order in Arthurian fiction. On the other hand, a reader of Grimmelshausen’s texts was always dependent on the information given in the new text in order to understand that the protagonists of the various texts (from Simplicius to the first person narrator in the Ewig-währender Calender) are acting in the same storyworld. While Grimmelshausen achieves a convergence of worlds through an increase in structuring, the authors of the Arthurian romances can refer to a general frame of reference that can integrate potentially each and every new space. The audience’s inferential efforts are relieved on the one hand by an overcoding of space-time structures and on the other hand by an undercoding. Both examples show, however, that behind the words of the text, there is not simply nothing.