Introduction: Theorizing Transtextual Characters in Ancient and Medieval Literature

1 Mapping the Field: Approaches and Contexts

When ›King Arthur‹ is mentioned, most people will immediately have an idea of what, or rather, who he is. The same is true for characters such as Achilles, Medea, Kriemhild, or in modern contexts, James Bond or Batman. ›King Arthur‹ is likely to resonate with a broad audience because he is not the product of a single text or medium or context, but a character that has appeared across a wide range of literary texts, from the medieval period to the twenty-first century, and in an equally diverse spectrum of media, from TV series and films to opera and graphic novels. No matter which cultural product might be the origin of one’s encounter with him, Arthur seems to retain his identity as Arthur across the wide range of potential sources. He is a transtextual (as well as a transmedial) character, that is, Arthur exists beyond the boundaries of one text (or medium) and instead can be found in various textual and medial environments that are not directly linked. Much of the nature of a character’s transtextuality is a cultural convention: we seem to have accepted and agreed that the various Arthurs refer to one and the same Arthur, or perhaps an abstracted Arthurfigure who remains unaffected by the changes that inevitably occur when you transpose a character into a different medium or historical context. Yet there is more to it, narratologically and theoretically speaking, than taking recourse to audience reactions, especially when one assumes a historical perspective.

In premodern contexts, transtextual characters were the norm rather than the exception. In both ancient and medieval literature, characters appear and re-appear in different environments (genres, versions, sequels and prequels, retellings and reworkings), even in environments with varying degrees of commitment to what could be known about the ›real‹ world. How can transtextual characters be defined? What distinguishes them from characters that are not transtextual? Why are they so common within premodern (but also in current) literary cultures? And why do we need a different theoretical and methodological framework for premodern transtextual characters (that is, characters in a manuscript culture)? These are the questions we aim to answer in the course of this special issue. Our approach is historically sensitive, comparative, and diachronic; we pay attention to the historical grounding of the phenomenon and the examples we scrutinize, but we do so in a decidedly comparative manner by considering ancient, medieval, early modern, and modern contexts.

What, then, is a transtextual character? While the current research landscape is defined by a range of different approaches to characters from literary, cultural, and media studies, we can take as our starting point the relatively uncontroversial observation that characters are often conceptualized as human or human-like entities represented through a medium and situated within storyworlds. Yet, when we engage with the existing research on historical as well as contemporary characters and their representations in different media further, it becomes clear that virtually every element of this initial conceptualization of characters refers to a complex theoretical problem (or set of theoretical problems), some of which need to be unpacked here at least briefly.

The kind of characters we are interested in, are variously described as transtextual characters (e.g., Philipowski 2019; Richardson 2010; Thon 2019), transfictional characters (e.g., Haugtvedt 2022; Pearson 2018; Wilde 2021), or transmedia(l) characters (e.g., Bertetti 2014; Kunz/Wilde 2023; Thon 2019). Each of these terms has different connotations, but they all share the assumption that the representation of a given character is not bound by the limits of a single text (work?) in a single medium, which at least at first glance seems to challenge the notion that characters always are or can be located in a (single) fictional storyworld, instead suggesting what has been described as »transworld identity« (e.g., Doležel 1998; Eco 1979; McHale 1987; Pavel 1986) and perhaps even leading to »transworld characters« (e.g., Lăcan 2019; Lowes 2005). In addition, transtextual characters can be said to »become common cultural property« through »a process of culturalization« (Margolin 1996, 116), a process that can also already be observed in the context of premodern literature.

In narrative and literary studies, characters are typically defined as »usually human or human-like« (Jannidis 2009, 30) entities. In view of their ontological status, characters are variously conceptualized as mental models of actual empirical readers (cf. Schneider 2001) or hypothetical model readers (cf. Jannidis 2004), on the one hand, and as »abstract objects« that are »neither material nor mental« (Reicher 2010, 115) and that could be specified as intersubjective communicative constructs with a normative component (cf. Eder 2008; see also Thon 2019), on the other hand. To come to terms with the various layers of character depiction, Phelan (1989) distinguishes between a mimetic, synthetic, and thematic dimension of characters, which is also taken up by Eder (2008; 2010) in his distinction between characters as fictional beings, characters as artefacts, characters as symbols, and characters as symptoms.

Even though transtextual characters are widespread in premodern literatures, they have hardly ever been theorized as such (for notable exceptions, see Philipowski 2019; Glauch 2025a). In media studies, by contrast, such characters have received a much more thorough treatment, if only from very different vantage points. James Bond or Batman are often referred to as »popular hero[es]« (Bennett 2017, 1) or »cultural icon[s]« (Brooker 2013, 8), not as character, which brings to the fore a different, reception-oriented, phenomenon that is tied to popularity and the versatility of characters that have transcended a particular media context – they are transmedial characters. Both transtextuality (e.g., Genette 1992; 1997) and transmediality (e.g., Kinder 1991; Rajewsky 2018; Thon 2016) are complex multidimensional concepts, but the distinction between transtextual and transmedia(l) characters seems to be comparatively simple. Put in a nutshell, we can say that transtextual characters are usually understood as characters that are represented in more than one (media) text (as will be discussed further below), while transmedia(l) characters are usually understood as characters that are represented in more than one conventionally distinct media form across fictional (media) texts (e.g., Thon 2019). Not least because the concept of transfictionality (e.g., Doležel 1998; Ryan 2008; Saint-Gelais 2011) describes a specific case of transtextuality, transfictional characters are usually understood as transtextual characters (or potentially transmedia[l] characters) within fictional (media) texts. In contrast, the concept of transworld characters emphasizes not the relation between the media texts representing the character(s) in question, but rather the relation between the worlds within which the represented characters are located. Unsurprisingly, transworld characters can thus be understood as characters exhibiting transworld identity, with the latter defined, for example, as the »identity of a given individual through worlds« (Eco 1979, 219) or as a »relationship of identity between entities that are located in different possible worlds« (Doležel 1998, 282).^[1]

Of course, the concepts of transtextual, transmedial, transfictional, and transworld characters are dependent on the very concepts of what boundaries are transgressed here: what exactly can be regarded as ›one‹ text, ›one‹ media form, ›one‹ fiction, ›one‹ world? Much energy has been spent on these questions, but sometimes with implicitly equating ›text‹ and ›work of art/fiction‹ (e.g., Richardson 2010). Indeed, the reappearance of a character in a translation or retelling of a story, despite the fact that this clearly constitutes a new text, might not be deemed a reappearance at all. So, it seems reasonable to relate ›transtextuality‹ not to the primarily linguistic notion of ›text‹ as a specific sequence of words or, more generally, signs, but to the notion of a work (of literature or other arts), even if this notion has lost some of its currency in literary studies (but cf. Currie 1991; and see below, »Text and Work«). The concept of transfictionality, on the other hand, is meant to refer to the overlap of diegetic elements in different works, ranging from multiple texts sharing nearly the same fictional world to texts with different fictional worlds sharing only single elements.

King Arthur, then, or perhaps rather, »the Arthur-figure« (Meyer 2017, 79), is a transtextual, transmedial, and transfictional character. His role in so many narratives is cohesive because there are links between ›Arthur‹ in a new version »and the character name, narrative function, and template established by the transfiction« (Pearson 2017, 116). In order to be effective, these links must be recognized as such by the audience. A further important factor is, as Ryan points out, that »characters acquire a life of their own« if one accepts that »worlds are imagined as existing independently of texts« and therefore »escape the control of the original author« (2019, 81). The control of an original author does not exist, however, when it comes to premodern contexts, in which a narrative core (materia) – such as the matter of Britain for medieval writers – leads an independent existence and is at any author’s disposal. Given the lack of authorial and legal control and the importance of materiae, transtextual characters may perhaps be more aptly termed characters tied to narrative cores, at least for the medieval context. Material characters are by definition transtextual and transfictional because they transcend the limits of individual texts or works, and form part of a storyworld that leads an independent (prior) existence. Scholars have also used the terms »itinerant« or »migrant« characters (Margolin 1990; Rüggemeier 2021) to highlight the mobility and versatility of such characters. In fact, the very (re-)appearance of a character in a different context accounts for their strength as transtextual character: the more often a character is re-used, the easier it becomes codified, iconic, and recognizable.

2 Text and Work

A central argument for historicizing transtextual characters is the concept of text or work, without which the literary character (and correspondingly also the transtextual character) cannot be conceptualized. A literary character is an entity that can either have its origin in a fictional text (like Reinhart Fuchs/Reynard the Fox) or be taken from the extra-textual world into a text (like Charlemagne or Napoleon). Starting from the individual text, a transformation must be recognizable that justifies speaking of a transtextual character. But what does the prefix ›trans‹ refer to? What relationship and what distance between two texts justifies speaking of a transtextual character? Doležel bases this necessary divergence on the respective possible narrative worlds. According to him, transtextuality is »[a] relationship of identity between entities that are located in different possible worlds« (Doležel 1998, 282). But how different must these possible worlds be for their characters to be describable as transtextual?

With these questions about relationships between texts, we find ourselves in a debate that is familiar to all those who deal with premodern (and in the narrower sense with manuscript) transmission. It is the debate about unstable texts or text boundaries, about versions, redactions and stemmatological (i.e. genetic) relationships that has always (and especially in its beginnings) preoccupied philological disciplines in their efforts to identify dependencies and relationships between texts. This evidence is intended to help trace the textual history, i.e. the historical development of a literary work, and to be able to undo ›errors‹ in order to reconstruct an original and fixed textual form (as far as possible). However, these philological efforts always take place with regard to the transmission of a specific literary work.

Our theoretical interest in relation to transtextual characters is not directed at the media or transmission of a literary work, nor at the work itself, but at the relationships that arise between works through the fact that they tell of characters that can be meaningfully examined as to whether they are the same or even identical. Some basic structures of approaches, arguments, questions and problems are nevertheless comparable. For anyone who wants to understand the relationship between transtextual characters in different works must first clarify what is to be understood by a work in order not to end up in the same blind alleys into which New Philology has fallen with its conviction that every manuscript preserves not only a text, but a work in its own right. Just as some differences in wording between two manuscripts do not create two different works, minor differences do not create transtextual characters. Conversely, in some theories the transtextual character is regarded as an abstraction of all the concrete characters that constitute it. In a comparable way, the work (the Odyssey, the Nibelungenlied, the Parzival) is constituted by many manuscripts, all of which differ from one another, but which nevertheless indisputably transmit and constitute the same work.

Anyone asking about the theoretical prerequisites of transtextual characters thus adopts a similar perspective on text complexes as philologists do in order to determine the historical extent and structure of the tradition that constitutes what we call a literary work like the Odyssey, the Nibelungenlied, or the Parzival. The proximity and distance of single texts to one another – albeit here always in relation to the character – are also decisive for the investigation of transtextual characters. Too great a correspondence (as in the case of translations, for example the relationship between Yvain in the Old French novel by Chrétien de Troyes and Iwein in Hartmann von Aue’s Middle High German adaptation/translation) does not create transtextual characters. However, if the individual texts and character designs are too far apart and too loosely related to each other, so that, for example, apart from the similarity of the names of two characters (as in the case of Hagen in the Nibelungenlied and in Kudrun), there is hardly any connection or recognizability between them, there are no transtextual characters either. The prerequisite for the term to be used meaningfully is that the respective characters (who are of course not independent of their respective storyworlds) are in a specific relationship of sufficient distance and sufficient proximity/recognizability/identity to each other.

Determining sufficient identity and distance is also a challenge for medieval studies, because medieval narratives are not only translated in the context of retelling, i.e. the practice of narrative appropriation, but are also repeatedly reworked, abridged, continued, embellished and connected to other narratives or storyworlds (e.g., Glasner/Zacke 2020; Bumke/Peters 2005; Bußmann 2005; Worstbrock 1999). Whether an adaptation is conceptually independent of another narrative is always a case-by-case decision; for instance, a long and an abridged version of a courtly novel can represent the same work or not. One example of this is the discussion concerning the definition of versions (Fassungen). Versions describe groups of texts in a specific relationship to one another: they all belong to the same work (or constitute the respective work) that exists in different versions. Discussing their characteristics through the example of the Nibelungenklage, the continuation of the Nibelungenlied (which has four different versions), Joachim Bumke has observed that versions, unlike adaptations or translations, are independent of each other with respect to their transmission. It is therefore impossible to determine which version is primary, more original, or more ›genuine‹ than another. Versions are, hence, equally original, even though they are all variations of the same work. However, in order to be considered versions, they must each have their own creative intention, which makes them distinguishable and gives them a specific profile or identity (cf. Bumke 1996, esp. 45–49).^[2]

It is worth considering whether the basic structure of this model could be suitable for describing the specific relationship between transtextual characters in premodern narratives. In contrast to modern literature, it is, after all, hardly ever possible to identify original characters invented by an author and referred to in later adaptations. The concepts of ›primary‹ and ›secondary‹ are just as inapplicable here as that of authorship in the modern sense of originality. The balance between identity and independence, which is necessary for determining the versions of a work, seems also to be highly relevant with regard to the transtextual characters, which must be recognizable as manifestations of ›the‹ respective character, but must also show sufficient independence within their own storyworld.

3 Fictionality vs. Factuality

The notion that characters are or can be located in a storyworld has several wide-ranging implications not only with regard to their ontological status (as the worlds represented by narrative representations are likewise conceptualized as mental models and intersubjective communicative constructs; e.g., Thon 2016), but also with regard to their relation to other characters and, indeed, their relation to the actual world. When discussing the status of transtextual characters in antiquity and the Middle Ages, it is thus important to consider the pragmatic contexts of their reception. Modern concepts of transtextuality and of transtextual characters are usually developed against the backdrop of a culture of reading and writing that takes the fictionality of ›literary‹ texts (and also its concept of factuality) for granted. When we turn to the ancient and medieval periods, however, the presumed boundaries between factuality and fictionality become blurred and highly porous. This also has implications for the concept of transtextual characters.

In the following we differentiate between ›fictive‹ (fiktiv) and ›fictional‹ (fiktional), as has been proposed in narratology (cf. Schmid 2014, 31). ›Fictive‹ refers to what is depicted in the text. It is synonymous with ›invented‹ or ›unreal‹. ›Fictional‹, on the other hand, categorizes the narrative as a whole; it does not only refer to the histoire. The term denotes a representation that does not claim to refer to reality. The opposite term to ›fictive‹ is ›real‹, the opposite term to ›fictional‹ is ›factual‹.

In the field of Classics, there has been much debate on whether and when there was an understanding of fictionality as a generally accepted pragmatic mode that makes no claims of referential truthfulness (for this definition, see Schaeffer 2012, § 1). The question is highly complex, as it concerns the cultural context of authors and recipients, and the literary conventions prevalent therein.^[3] Irrespective of how one decides the matter in detail, it seems clear that in antiquity there was no equivalent to the modern concept of fictionality, as defined above.^[4] Instead of working with the dichotomy of fictionality and factuality that is fundamental in modern narratology, ancient theory distinguishes between degrees of fictivity: according to rhetorical theory, which informs literary practices, narrative is divided into historia/ἱστορία, argumentum/πλάσμα, and fabula/μῦθος (or: res vera, res ficta, and res fabulosa). These terms can be translated as ›factual representation‹, ›probable or at least possible fiction‹, and ›improbable, if not impossible fiction‹ (see Feddern 2018, 306). Similarly, the principle of probability (probabile/εἰκός), which plays an important role in ancient rhetoric and literary theory, operates with a scalar rather than a binary notion of fictivity.^[5]

In the (European) Middle Ages, the poetic distinctions developed in antiquity – especially the distinction between historia, argumentum, and fabula – were taken up and integrated into strands of Christian ontology and epistemology.^[6] As a result, the scope for socially acceptable fictio that was not subject to expectations of referential truthfulness was very narrow. It was essentially limited to those narrative forms that unmistakably identified themselves as allegorical and openly displayed their ›as-if‹ character: the animal fable, the integumentum, the allegory (cf. Schneider 2020, 84), which are all of course didactic and insofar claim a specific authority, validity, and truth. Ultimately, the frame of reference against which all poetry had to justify itself was the truth of the factual, in the sense of verifiable information, or information that was conventionally thought to refer to something ›real‹. »The freedom of fiction«, as Sonja Glauch puts it pointedly, »meant deficiency« (Glauch 2014, 139).

However, the normative-ideological framework in which medieval poetic invention was situated, or in which it situated itself, tells us very little about the extent to which medieval audiences were nevertheless able to receive the works in a mode corresponding to fictional texts. In fact, the extent to which a text was considered fictional or factual – if, unlike myth, saga, or legend, it was not epistemically indifferent to this opposition (cf. Glauch 2014, 97, 100, 125) – may have been largely determined by the cultural environment and social practices, and the same text may have been open to (gradually) different modes of reception.

These observations have consequences for the understanding of literary characters and their (possible) transtextuality. Due to the scaling notion of fictivity, storyworlds in ancient and medieval texts work differently from what is commonly assumed in modern theory. First (and depending on the genre), storyworlds are not completely separate from the extra-textual world. As there are no clear boundaries between factual and fictional accounts, nor between text, context, and world in premodern literatures, different types of characters and different literary genres are relevant to the study at hand. Except for their different origins in storyworlds, there is no categorical difference between characters from myths, historical accounts, and newly invented narratives, which are a rare occurrence in premodern literature anyway. They can easily meet in a narrative and thus be compared regarding their transtextuality. A second consequence of this is that a referential reading of literary characters as ›masks‹ for the author or other historical individuals is always possible in ancient literature (see the remarks on the literary persona, below). Moreover, there are no clear-cut boundaries between different storyworlds, although the storyworlds of various text types differ from each other. The (re-)appearance of a character in different contexts is therefore not as exceptional as modern theory presumes and does not necessarily produce an aesthetic shock.

The absence of a culture of fictionality comparable to that of more recent times also has consequences for the nature of transtextual characters and their variability. When characters are not part of a culturally established ontology and aesthetics of fictionality, but are either considered historical or occupy a realm beyond factuality and fictionality, their multiplicity is subject to relatively narrower limits than in the context of a modern concept of fictionality.

Modern concepts of transtextuality leave ample room for contradictions in the various representations of a character (cf. Brooker 2012; 2013; Thon 2015). Therefore, these concepts seem questionable for characters in ancient and medieval texts, whose referential status lies in the realm of probability and on a spectrum of different degrees of fictivity. These texts often claim to be true, even if they tell of something that for modern readers is a clear indication of fictionality. For this reason, the concepts of fictionality, factuality, and truth should be carefully adapted to claims to validity and the self-representation of ancient and medieval texts (cf. Putzo 2023). An example of this is the common use of allegorical characters and personifications in the context of philosophical and doctrinal writing. Whether the famous dialogue between Boethius and Lady Philosophy, as depicted in the De Consolatio Philosophiae, was considered a fictional work or not is impossible to determine.

Of course, this does not mean that ancient and medieval texts did not produce different, sometimes even contradictory, character profiles under one and the same character name. However, while ancient authors used the potential of the variatio of recurring characters by presenting new perspectives on the well-known stories, medieval texts seem to leave less room for such variation. Because medieval characters were generally conceived as somehow grounded in the real world or even prehistoric, and thus did not belong to a second, invented world, but ontologically to the same world as that of the recipients – even if it was an existentially, spatially, or temporally distant world –, variations in their representation often became the subject of a controversial ›work on the truth‹ of the respective version.

4 Character and Genre

It might at first appear a truism that the transtextuality of characters also affects generic aspects and vice versa. It is evident that some genres have always relied more heavily on transtextual characters than others. Certain genres of contemporary popular culture, in particular, are rich in such phenomena – first and foremost, the ubiquitous superhero comics and their respective film adaptations (cf. Thon 2019 as well as Thon 2025/in this issue). In contrast, other genres – perhaps because they rely on tightly closed narrative worlds (see below) – do not typically feature transtextual characters, making them an exception. This applies, for example, to the European novel of the 18^th, 19^th, and early 20^th centuries (cf. Richardson 2010). The highly variable intensity of this phenomenon from genre to genre correlates, as expected, negatively with the aesthetic impact that the transtextuality of characters holds in a given genre. Statistical frequency plays a significant role: it is crucial if at all and how often transtextual characters appear, less so which characters within the narrative world they are and which narrative roles they fulfill (protagonist, antagonist, supporting characters of the first, second, nth rank, etc.).

At least one other dimension of transtextual characters must be distinguished from this first, obvious generic application, which has received less attention in previous research because it is typical especially of older forms of storytelling. This other dimension does not depend on statistical expectations but instead on the fundamental narrative characteristics of certain genres that are inseparably linked to specific transtextual characters and their narrative roles. Medieval secular epic offers a rich repository for this. A 13^th-century Arthurian verse romance cannot exist without Arthur, nor without Guinevere, Gawain, and Kay (see Reuvekamp 2025/in this issue). Similarly, a late medieval German heroic Dietrich-epic requires Dietrich of Bern, Hildebrand, further characters from the circle of Bern heroes, and dark characters such as the notorious traitor Witege (cf. Kragl 2025/in this issue). The names alone or the personal identity of the characters are not enough. In the individual texts, the characters must always fulfill similar, though sometimes varied, narrative functions: the absolute yet weak king, the latently unfaithful queen, the best knight, the court-stabilizing jester; the hero, his mentor, his allies, the traitor, and so on. In these cases, then, transtextual characters are formative for their respective genre; they are more firmly inscribed into the narrative worlds than characters that merely travel superficially and optionally between worlds or texts. As they form part of the basic equipment of their genres, their transtextuality becomes aesthetically incidental: it is not a specific appeal but obligatory and self-evident (cf. Glauch 2025/in this issue; Kragl 2025/in this issue).

The aforementioned generic dimension of transtextual characters and their strong effect relies on narrative worlds of high stability which are clearly separated from one another (cf. Harweg 2004). The inclusion of a transtextual character automatically constitutes a narrative act of transgression with a temporal logic (a later text refers to an earlier one; cf. Margolin 1990, 865–868; Richardson 2010). This transgression primarily occurs at the level of the storyworld. The second dimension, on the other hand, relies on the stability of character designs around which the texts of a specific genre coalesce. These cases are characterized by the similarity of the individual storyworlds across different texts, or their capacity to be made similar, thereby forming a kind of narrative universe. The more permeable the boundaries of the storyworlds become – especially regarding the repertoire of characters –, the more the transtextuality of characters loses its transgressive nature. The relationships between the texts lack clear directional vectors; in extreme cases (e.g., the Dietrich epics), the texts form a quasi-synchronous sum of generally available »narrative knowledge« (Philipowski 2019, 126). Once again, the transtextuality of a character is initially a phenomenon of the histoire; however, certain genres seem to associate metonymically with central transtextual characters, leading to generic discourse-specific features, such as stylistic expectations tied to the characters.

Depending on whether one regards the transtextuality of a character as a purely structuralist-narratological phenomenon or also as an aesthetic event, transtextual characters from older storytelling can be considered either paradigmatic or borderline cases of transtextuality (cf. Philipowski 2019, 120–127). Their theoretical description is challenging. Is, or in what sense is, a character transtextual if their repeated appearance establishes a sub-generic series of texts (e.g., the Hagen/Kriemhild duo in the Nibelungenlied and Klage, or Rennewart in Wolfram’s Willehalm and related epics)? Is, or to what extent is, a character transtextual if they belong to the fixed inventory of a cross-textual storyworld (as in the aforementioned Arthurian examples)? Is a character transtextual, or should they be described as such, if their behaviors and patterns define a genre (as with Dietrich of Bern, Alexander, Arthur, or Charlemagne)?

The terminological problem expressed by these questions – what does transtextuality mean (under the one or the other condition)? – reveals that the concept of transtextual character in older storytelling intersects with categories like genre or even series, especially when a genre and/or a series (these concepts are not clearly separable in older storytelling; cf. Grubmüller 2015) are primarily defined by characters and character functions, or possibly even originate from a cluster of characters. The literary history of this phenomenon remains to be written. It would have to account for diverse cases such as the types in the Commedia dellʼarte (and their diffusion into other poetic domains) or ›subliterary‹ genres like pulp fiction or the modern detective novel (e.g., Miss Marple, Commissaire Maigret, or Hercule Poirot). What is more, such a history would also need to develop a mode of description for the (historically not uncommon) case where a text or group of texts, initially classified under the first dimension, gradually shifts into the second dimension due to the habitualization of the phenomenon. These mixed forms, in particular, which resist ideal-typical categorization, demand closer scrutiny (cf. Bertetti 2014, 2350; Denson/Mayer 2018, 67 sq.; Thon 2019).

5 Conceptualizing Transtextual Characters

In the preceding sections, we have illustrated the theoretical, medial, and historical frameworks that have to be adapted for analyzing transtextual characters in premodern literature. In order to describe how characters are represented as the same in and across different works of ancient and medieval literature, we need to consider the concepts of the literary work, genre, and fictionality. In addition, theorizing transtextual characters depends on the underlying concept of (literary) character. In what follows, we introduce four theoretical approaches to determine and describe the identity of literary characters in different works.

6 Towards a Theory of Transtextual Characters

Finally, we would like to turn to the question of the relevance and systematic place of the theory of transtextual characters: does the fact that this theory is still relatively young indicate that it addresses a specialized or niche problem? We believe not only that the phenomenon to which this special issue is dedicated, i.e. transtextual characters in premodern literature, has always existed in (presumably) all literatures, but also that literary studies have always been concerned with this phenomenon. The analysis of transtextual characters has occurred, for example, when articles were written on characters that we now understand as transtextual (such as Heracles, Helen, Alexander the Great, or King Arthur), or when series of texts/genres were analyzed that are constituted by these characters, such as the Alexander romance, the Arthurian romance, or the Dietrich epic.

Yet, the question of how to define and delineate the character that connects the respective works was not approached systematically until a few decades ago. This is probably because the theory of transtextual characters is located at the intersection of theories of intertextuality, text, media, genre, transmission, and theory of character, and thus cannot be treated in isolation. An introduction to Arthurian romance can therefore not answer the question of whether all Arthurian romances tell of the same Arthur or of a different one. By now, however, the discussions within the above-mentioned areas of literary theory and literary history have become so methodologically sound and compatible that the time seems right to work towards a theory of transtextual characters, which must always be historicized accordingly. This special issue serves as an impetus in this direction.