‘German Theory’: Cosmopolitan geographies, counterfactual histories and the (non)travel of a ‘German Foucault’

In this paper, I propose the notion of ‘German Theory’ to allude to the processes of translation and circulation of theoretical ideas discussed in the German humanities and German geography across and beyond the linguistic boundaries of its origin. However, these ideas did not travel, at least not to Anglophone geography. This paper investigates why not, and it maps out the lost theoretical potential that has been foregone as a result. ‘German Theory’ is thus read here as a potentiality that has not actualised. More specifically, this paper studies the reception of Foucault, as it emerged in two distinct territories of thought, each with their own interpretation: first, Kittler’s work on the materiality of discourse in German humanities, and second, the ‘discourse school’ in German geography. The latter’s insistence on the textuality of discourse disconnected their Foucault reception not only from Kittler’s, but equally so from ongoing debates in Anglophone geography. By reflecting on why the ‘German Foucault’ did not travel to Anglophone geography, I raise a speculative epistemological question about the im/potentiality of ‘German Theory’: could a more cosmopolitan theory have emerged from a circulation of these ideas had they travelled across linguistic boundaries?

In this paper, I study the emergence of a 'German Foucault'. While the 'Italian Foucault' was a re-interpretation of Foucault's work on biopolitics, the 'German Foucault' offered a re-working of Foucault's discourse theory. The German Foucault emerged in two distinct territories of thought. Both propagated different Foucault interpretations, giving two distinct forms to his discourse theory: first, Friedrich A. Kittler's work in the Geisteswissenschaften. Kittler's notion of the materiality of discourse circulated widely within Anglophone media and cultural studies as 'German school' of media studies (Sale and Salisbury, 2015: xiv). Second, the 'discourse school' in German geography, which formed more than a decade later in the field of Neue Kulturgeographie, following Anglophone's new cultural geography's reading of Foucault, which did not travel to Anglophone geography.
The 'German Foucault' did not travel to Anglophone geography, because, by insisting on the textuality (instead of the materiality) of discourse, German geographers disconnected their 'German Foucault' from Kittler's and from ongoing debates in Anglophone geography. These dynamics of reception foreclosed the potential circulation of Kittler's work among Anglophone geographers, although his notion of materiality of discourse could have resonated well with Anglophone geographers' interest in non-representational theory and materiality.
In seeking to understand why the 'German Foucault' did not travel to Anglophone geography, this paper resonates with ongoing debates about the uneven power geometries of academic knowledge production in human geography and the personal and institutional dilemmas that non-Anglophone geographers face around questions where to publish, which theories to engage with, how to write and what topics or cases to select for study (Aalbers and Rossi, 2007;Best, 2016;Hannah, 2016;Houssay-Holzschuch and Milhaud, 2013;Houssay-Holzschuch, 2020;Kitchin, 2005;Korf et al., 2013;Minca, 2000Minca, , 2013Paasi, 2005;Rossi, 2008;Schlottmann and Hannah, 2016). More specifically, J€ ons and Freytag (2016: 4) identify two conditions of possibility for the flow of knowledge and ideas across linguistic boundaries: first, the work of 'boundary spanners', who 'facilitate knowledge transfer', and second, the willingness of academic peers to engage with ideas outside of their usual school of thought. 5 Yet, my case also shows that the dynamics of reception in the territories of thought of non-Anglophone geographies are equally relevant to explain the (non-) circulation of theoretical ideas beyond linguistic boundaries.
The ambition of this paper is to go beyond seeing these questions only through the lens of power geometries, however. By reflecting on why the 'German Foucault' did not circulate to Anglophone geography, I pose a speculative epistemological question about the im/potentiality of 'German Theory': what could we potentially have gained theoretically from a circulation of these ideas had they travelled across linguistic boundaries? Perhaps a more cosmopolitan theory, which 'confronts itself with the varieties of places, . . . with the variety of alterities making up the discipline' (Minca, 2000: 289). Exposing the (counterfactual) intellectual histories of 'German Theory' is thus not a 'provincial' project (Houssay-Holzschuch and Milhaud, 2013), but a 'cosmopolitan' one if understood as an imaginary that can transcend the particularisms (Cheah and Robbins, 1998) of its origin.
Geisteswissenschaften and was popular more outside of the academy in informal theory reading circles in major German cities such as West Berlin, Hamburg, Frankfurt or Cologne (Birnstiehl, 2016;Felsch, 2015;Holub, 1992;Raulff, 2014). The opposition to poststructuralist deconstruction was so strong among leading scholars in the Geisteswissenschaften that the only possibility to engage French Theory was in the form of an oedipal revolt (Gumbrecht, 2007) against the 'old guard'.
Kittler's revolt was directed against the hermeneutic tradition, which dominated the Geisteswissenschaften at that time. Kittler's self-declared programme was to tear 'the veil away from hermeneutics and [to dispel] its aura, its shimmering suggestion of sacral authority' (Wellbury, 1989: ix). Hermeneutics, the self-proclaimed 'queen of science', is naked, proclaimed Kittler: in 1980, he edited a volume with the provocative title 'Die Austreibung des Geistes aus den Geisteswissenschaften', which could be translated literally as 'the expulsion of the spirit from the sciences of the spirit', or, in a more liberal translation, as 'kicking the human out of the humanities' (Winthrop-Young, 2012: 365). Kittler described the academic hegemony against which he revolted as follows: Discourse Networks 6 came about when hermeneutics had established this clever alliance with Ju¨rgen Habermas. Or the other way round. It was Habermas, I believe, who in the end smuggled Hans-Georg Gadamer onto his list of winners, at which point there was no getting through anymore. (Kittler and Weinberger, 2012: 378) Kittler alludes here first to Habermas, whose criticism of the work of Derrida and Foucault was very influential in the Geisteswissenschaften. Indeed, Habermas's critical tone against Derrida and Foucault clogged the reception of poststructuralism among leftist scholars in the Geisteswissenschaften (Birnstiehl, 2016;Holub, 1992). In his influential 'Adorno lecture ' (1980/2003), and subsequently in his book Der philosophische Diskurs der Moderne (1985), Habermas criticised the anti-modernist, anti-humanist attitude that he identified in Foucault's and Derrida's writings. In his view, the latter had inappropriately embraced Nietzsche and Heidegger and given up the emancipatory potential of modernity. Habermas's par force critique poisoned the intellectual exchange during the early 1980s that could have linked French poststructuralism to Critical Theory of the Frankfurt school (Frank, 1988: 7ff.).
Kitter further mentions Gadamer, whose concept of hermeneutics was foundational in philological streams of German literary studies, history and ancient studies. Gadamer's hermeneutics also stood in tension with French poststructuralism (Frank, 1988: 7ff.;Grondin, 2009;Michelfelder and Palmer, 1989) and this cognitive dissonance had broad impacts: for example, the influential network 'Poetik und Hermeneutik' [poetics and hermeneutics], an interdisciplinary network of mostly male German scholars in the Geisteswissenschaften across literary studies, history and philosophy, 7 which, in the 1960s and 1970s, had gained a 'hermeneutic hegemony' (Wellbury, 1989: ix), kept a distance to French Theory. These scholars shared a strong commitment to philological craft and diligence, a craft that they thought would be lacking among French poststructuralists and their followers (Frank, 2017: 258). Hans Robert Jauss, often seen as key gatekeeper of the group, accused poststructuralists and deconstructivists of 'irrationalism' (Schwab, 2017: 414).
Kittler directed his exorcising spirit against this hermeneutic hegemony. His antihermeneutic affect made him also depart from Derrida's deconstructivism, which he considered to be too 'textual' (Kittler, 2012: 120;K€ ohler, 2018: 198). Instead, Kittler searched for a 'cold model of structure' (Kittler and Weinberger, 2012: 375). Kittler's epistemological apparatus therefore engaged Derrida, Foucault and Lacan as a 'joint achievement of the three' (Wellbury, 1989: xi-x) and departed at the same time from a singular Derridean, Foucauldian or Lacanian reading. Kittler's highly original fusion of poststructuralism and deconstruction with cybernetics and mathematics (Mersch, 2006: 205) exposed the materiality of discursive occurrences: in Aufschreibesysteme (1984), Kittler excavated the material preconditions of hermeneutic understanding in the practices of alphabetisation, the expansion of book production, and the organisation of the German university. In other words, he traced this understanding as 'a function of instructional practices and technologies' (Wellbury, 1989: x).
This Kittlerian twist became particularly important in Grammophon Film Typewriter (1986), which focuses much more on media technologies (Sale and Salisbury, 2015: xx). Foucault, according to Kittler, dissects epistemes but fails to explain why these epistemes emerge and maintain themselves. Discourse networks, for Kittler, are the 'network of technologies and institutions that allow a given culture to select, store and process relevant data' (Kittler, 1990: 369). Foucault, Kittler contends, simply forgot that 'writing itself, before it ends up in libraries, is a communication medium' (Kittler, 1999: 5). For Kittler, media technologies and cultural techniques form discourse networks, or, in his laconic language: 'Media determine our situation' (Kittler, 1999: xxxix). And Kittler concludes from this axiom: 'Archaeologies of the present must also take into account data storage, transmission, and calculation in technological media' (Kittler, 1990: 369).
Kittler's important role in importing 'French Theory' into Germany's Geisteswissenschaften is now widely acknowledged (Birnstiehl, 2016;Gumbrecht, 2013;K€ ohler, 2018: 171ff.;Mersch, 2006;Wintrop-Young, 2007). Kittler is held to be one of the founding fathers of Medienwissenschaft [German media studies] and celebrated as one of its central theorists (Mersch, 2006). After initial struggles with his Habilitation, 8 Kittler started a successful career at German universities and held a chair in Kulturwissenschaft at the prestigious Humboldt-University in Berlin from 1993 until his death. His 'material turn' inspired a broader network of scholars in the Geisteswissenschaften around Hans Ulrich Gumbrecht and Ludwig Pfeiffer, who organised a series of colloquia around the notion of 'materialities of communication' (Gumbrecht and Pfeiffer, 1994). Kittler had gained the nimbus of a public intellectual, often being interviewed in the media and, after his untimely death, being remembered in numerous obituaries.
Kittler's influence has travelled beyond Germany's academia into Anglophone media studies, where his work has been discussed in a number of collected volumes (e.g. Ikoniadou and Wilson, 2015;Sales and Salisbury, 2015;Winthrop-Young and Gane, 2006). Many of his works have been translated, starting with Aufschreibesysteme that appeared as Discourse Networks in 1989, and is now often considered an early classic in media studies (Birnstiehl, 2016;Gumbrecht, 2013;Mersch, 2006;Wintrop-Young, 2007). Through the work of important boundary spanners in literary studies, such as Gumbrecht, who moved from Germany to Stanford in 1989, or Werner Hamacher, an influential German literary theorist, who edited the prestigious Meridian series at Stanford University Press, Kittler's work became more prominent in Anglophone media studies, where Kittler is considered a key figure of a 'German school' of media theory (Sale and Salisbury, 2015;Winthrop-Young, 2008). In Anglophone geography, Kittler's work has not travelled much, however, although Thrift references his notion of discourse networks in his influential Spatial Formations (1996: 258, 302; also: Thrift and Olds, 1996), and Craine gives him a passing reference in an editorial on media geography (Craine, 2007: 147).
Even less did Kittler travel to German geography, however: Neither in the late 1980s, when first works on postmodernity appeared in German geography (e.g. Hasse, 1989), nor later in the 2000s, when the 'cultural turn' took firm roots in German geography, did German geographers take notice of Kittler's work and his reception in the Geisteswissenschaften. Even the few German geographers who, in the 1980s, engaged explicitly with the hermeneutic tradition (e.g. Pohl, 1986) or postmodern thinking (e.g. Hasse, 1989) did not take up Kittler's impulse. Pohl, for example, propagated his hermeneutic geography through the work of Gadamer, Lu¨bbe and Waldenfels (Runkel, 2017) without referring to the intellectual milieu of P þ H. Hasse, in turn, engaged with French postmodernism through reading Lyotard and Baudrillard, as well as authors from Anglophone cultural studies and geography (e.g. Jameson and Jackson). Nor did Kittler feature in Hasse's monograph Mediale R€ aume [Medial Spaces] (Hasse, 1997), which engaged explicitly with the question of mediality. Even in an edited volume published much later with the title Mediengeographie Two dynamics may account for this intellectual disconnect. First, since the 1960s, German geographers had increasingly detached themselves from the milieu of Geisteswissenschaften, including literary studies and history (Dix, 2005;Steinkru¨ger, 2015;Wardenga, 2005). The quantitative turn propagated since the 1960s (Michel, 2016;Paulus, 2017), and the pressure to implement new degree programmes tailored towards the job market provided an incentive for applied and quantitative strands of geography rather than hermeneutic craft (Wardenga et al., 2011). Hermeneutic approaches therefore had a difficult standing at the expense of more applied and quantitative work, and German geographers had very little knowledge of what was going on in the Geisteswissenschaften (Korf and Verne, 2016). Kittler's exorcising attacks against hermeneutics simply might therefore have been unintelligible to German geographers.
Second, when German geographers started to engage with Foucault and other poststructuralist authors in the late 1990s and early 2000s, Foucault travelled to German geography mainly through his reception in Anglophone new cultural geography of the late 1980s and early 1990s. In the internal turf battles of German geography, reference to Anglophone geography could be strategically deployed as a source of innovation and 'internationalisation' against other schools of thought within German geography. The 'Foucault' of new cultural geography became a central pillar in the formation of a so-called discourse school in German geography, which gave Foucault's discourse theory a textual twist at odds with Kittler's notion of discourse networks. These reception dynamics are discussed in the next section.

The discourse school of Neue Kulturgeographie
Foucault started to become a key reference and theoretical celebrity in German geography in the late 1990s and early 2000s when a group of scholars started to advance Neue Kulturgeographie, a new field of study in human geography that was influenced by the cultural turn in Anglophone geography and the reception of Foucault in new cultural geography. An influential group of scholars within this network of Neue Kulturgeographie propagated and further developed a textual interpretation of Foucault's discourse theory, which became highly influential in a specific German network of geographers and social scientists but did not reach out to Anglophone audiences, nor did it resonate with Kittler's material turn to discourse theory. In this section, I will explain the scholarly struggles that produced this reception pattern.
Two events and a reader were key in popularising Foucault in German geography and marked the emergence of Neue Kulturgeographie: In 2001, during the Deutscher Geographentag [the biannual conference of German geographers] in Leipzig, Ute Wardenga and Andreas Dix organised a panel that took stock of the 'cultural turn' in Germany (Blotevogel, 2003;Natter and Wardenga, 2003;Pu¨tz, 2003;Werlen, 2003). Sahr (2003: 231f.) described the Stimmung [atmosphere] of the event as a sensation: 200 geographers squeezed together in a tiny room in a creative, almost revolutionary mood, discussing the epistemological turn that was going on in Anglophone geography around the 'cultural turn'. According to Sahr, the 2001 Leipzig session provided the basis for the revolutionary Stimmung at a subsequent event, again in Leipzig: On 29-31 January 2004, more than 150 scholars attended a conference with the title 'Neue Kulturgeographie in Deutschland' [new cultural geography in Germany]. 9 Leipzig 2004 also provided the stage to popularise an influential reader on Kulturgeographie (Gebhardt et al., 2003a).
In his opening address at Leipzig 2004, Paul Reuber, one of the organisers of the event, emphasised that Neue Kulturgeographie was not new in a strict sense, as a new cultural geography existed already in the Anglophone world, but that it was new in German geography. 10 Implicit in this narrative was the anxiety of lagging behind the international standards of human geography. Berndt and Pu¨tz (2007: 7) expressed this feeling when they wrote that German geography risked being a 'late-comer', lagging behind ('den Anschluss verlieren') the latest international debates around the cultural turn(s) in Anglophone geography, where in the late 1980s and early 1990s, 'conference sessions on the theme of postmodernism [had been] overflowing' (Creswell, 2013: 263). This framing was also reflected in the Kulturgeographie reader, in which several key contributions engaged with Anglophone cultural studies (e.g. the work of Judith Butler, Stuart Hall and Edward Said) and new cultural geography (e.g. the contributions of Belina, 2003;Lossau, 2003;Reuber and Wolkersdorfer, 2003;Stru¨ver, 2003;Woods, 2003, and the introduction to the reader in Gebhardt et al., 2003b) and included translations of two seminal texts from Massey (2003) and Soja (2003) to account for 'the international debate'.
New cultural geography's interest in Foucault in the 1990s made his work an important reference for Neue Kulturgeographie scholars. Foucault's discourse theory was particularly prominent early on among feminist geographers (e.g. Bauriedl et al., 2000;Kutschinske and Meier, 2000; but also Hasse and Malecek, 2000: 104f.) and became a central reference of the so-called discourse school within Neue Kulturgeographie (Hannah, 2016;J€ ons and Freytag, 2016;Weichhart, 2008: 374ff.). 11 This 'discourse school' painted a particular notion of discourse that it constructed from a reading of Derrida and Foucault that centred around Derrida's notion that 'there is nothing outside the text' -often reproduced as an English phrase rather than in French or German (e.g. Gebhardt et al., 2003b: 11). In her plenary lecture in Leipzig 2004, Lossau made a strong case for a textual notion of 'discourse', which would heavily influence subsequent discourse-theoretical debates among German geographers. Engrained in this conceptualisation is a theory of the subject as a fictitious projection (Lossau, 2002: 37ff.). Gebhardt et al. (2003b: 15) and Lossau (2002) legitimate this theory through Foucault's early writings and conceptualisations of 'discourse' and 'subject' in 'The Order of Discourse' and the 'Archaeology of Knowledge' (Marquardt and Schreiber, 2012). 12 This 'Foucault' came to German geography through a conceptualisation of 'discourse' and 'subject' that had influenced many scholars in Anglophone new cultural geography in the 1990s, although for Anglophone geographers, Foucault's Discipline and Punish seems to have been more central at the time (Hannah, 2007(Hannah, , 2010Philo, 2012). There has been prominent work in discourse theory in Anglophone geography though (Philo, 2011) that became a central inspiration of German geographers, in particular the discourse analytical work of critical geopolitics of Dalby, Ó Tuathail, Dodds and Sharp (Wolkersdorfer, 2001). Reuber and Wolkersdorfer (2003: 51f.) became central protagonists of a rebirth of Politische Geographie along the lines of critical geopolitics, especially in the aftermath of the Balkan wars and the attacks of 11 September 2001. In a conference they organised around the theme in Mu¨nster in 2001, discourse-theoretical approaches celebrated a kind of breakthrough (Reuber and Wolkersdorfer, 2001). Similarly, Lossau (2002Lossau ( , 2003 had developed her 'ANDERE Geographie' [a 'DIFFERENT geography'] through an engagement with critical geopolitics and postcolonial studies, as Glasze had done on Francophone geopolitical imaginations (Glasze, 2007(Glasze, , 2013, followed by work on neoliberal subjectivities in urban geography (e.g. Glasze et al., 2005;Mattissek, 2008).
Through this engagement with critical geopolitics, 'discourse' was primarily framed as 'method' (Gebhardt et al., 2003b: 15), which resulted in an early engagement with a couple of German-speaking authors who had pioneered the method of discourse analysis, especially Keller et al. (2001) and J€ ager (1999), both of whom Gebhardt et al. (2003b, footnote 3 on page 16) explicitly mention. As Mattissek and Glasze (2016: 44f.) emphasise, the discourse school 'clearly [distinguished] itself from governmentality studies' -with the former having a focus on methodological questions, whereas the latter on epistemological ones and questions of Zeitgeist. 13 Subsequently, a group of younger geographers devoted great efforts to a more systematic conceptualisation of this discourse analysis as method (Mattissek, 2007;Mattissek and Reuber, 2004), combining Foucault with Laclau and Mouffe's theory of hegemony (Glasze and Mattissek, 2009b). Financially supported through a network grant of the German Research Foundation (DFG Forschergruppe), their work resulted in the Handbuch Diskurs und Raum [Handbook Discourse and Space] (Glastze and Mattissek, 2009a), which compiled not only chapters on theories of subject and discourse, but also meticulous elaborations on different methods of discourse analysis (e.g. lexicometric methods, argument analysis and visual analysis).
Mattissek and Glasze (2016), editors of the Handbuch, claim two key innovations that set the school apart from Anglophone new cultural geography: first, a more elaborate theorisation of the ontology of the subject; and second, a more pertinent reflection on the question of methodological implementation of discourse theory. While conceding that the original impetus came from Anglophone human geography, Mattissek and Glasze (2016: 40) emphasise how from this initial impetus, intense interdisciplinary exchanges within a distinctly German (language) thought collective developed that resulted in a treatment of discourse analysis distinct from Anglophone human geography. Hannah lauded this 'discourse school' in a recent review as follows: 'the systematic and differentiated German-language handling of critical discursive methods, and the 'Neue Kulturgeographie' discourse in which it has developed, is thus a tremendously important and distinctive fruit of recent Germanlanguage human geography' (Hannah, 2016: 73).
And yet, this 'distinctive fruit' of German geography did not travel beyond the Germanspeaking world into Anglophone geography: The discourse school was highly influential within German geography and, to some extent, in neighbouring disciplines in Germany, but did not find much resonance among Anglophone geographers. For example, the Handbuch, albeit popular and much cited in German-language scholarship, has not been cited much in Anglophone geography. 14 Two dynamics may explain this dynamic of reception. First, the discourse school's focus on textuality ('the end of the subject') was out of touch with the most recent theoretical turns that became fashionable in Anglophone geography around 2000 (Philo, 2000;Thrift and Dewsbury, 2000): non-representational and more-than-representational approaches, emphasising notions of materiality, network and affect. Key representatives of Neue Kulturgeographie, however, defended the textual notion of 'discourse' against nonrepresentational theory's critique of this discourse theory among Anglophone geography, most prominently Julia Lossau in her keynote in Leipzig in 2004 (cf. Redepenning and Wardenga, 2004: 3 and my conference notes; see also Lossau, 2002: 36, Lossau, 2008. As a result, the 'German Foucault' of Neue Kulturgeographie was confined to stand for an anti-humanistic, textual notion of 'discourse', and the reception of morethan-representational and posthumanist approaches started only later among German geographers (Kazig and Weichhart, 2009).
Second, Lossau's verve in defending this notion of 'text' can also be understood as a response to German geographers who had been critical of Neue Kulturgeographie and its discourse theory (e.g. Blotevogel, 2003;Ehlers, 2007;Klu¨ter, 2005;Werlen, 2003: 49). In Leipzig 2004, two eminent German geographers, Hans Heinrich Blotevogel and Peter Meusburger, criticised Neue Kulturgeographie to fall victim to a limited and superficial reading of poststructuralist theorists (Blotevogel) and the theory debates in Anglophone cultural studies, e.g. around the Birmingham school (Meusburger). 15 The enormous intellectual labour that went into conceptualising discourse theory as a method in the Handbuch might also be understood as a response against those internal critiques within German geography, but these appeared less topical in Anglophone geography.

To read what was never written
Let us briefly pause after this lengthy analysis of why geographers did not read Kittler and instead think about what could have happened, if German geographers had read Kittler? Imagine that German geographers had been in touch with Kittler's Foucault; that they had critically interrogated his notion of discourse networks. Imagine German geographers taking up Kittler's work to infuse the notion of materiality of communication into Anglophone geographers' debates on non-representational and more-thanrepresentational geographies, post-humanism and materiality. Thrift's occasional reference to Kittler in Spatial Formations and elsewhere could have given them an entry-point to connect the 'German Foucault' with Anglophone geography.
In posing these counterfactual questions, I take inspiration from Matthew Hannah, who once speculated about what could have happened to Benno Werlen's Handlungstheorie (highly influential in the 1990s and thereafter in German geography), had his work been confronted with a more vigorous Marxist tradition in German geography (which it was not) (Hannah, 2016: 76f.). Hannah is convinced that Werlen's subjectivism would have been softened. Similarly, a Kittlerian twist might have had effects on the intellectual trajectories of German geography and its 'discourse school'.
First, within Neue Kulturgeographie, a dose of 'Kittler' might have softened the discourse school's strong interest in the method of textual analysis, e.g. lexicometric methods, argument analysis and coding (Dzudzek et al., 2009;Glasze et al., 2009;Mattissek, 2009) rather than the material conditions of possibility for the circulation of these discourses that Kittler was interested in. Instead, the discourse school resisted for a significant time the nonrepresentational, actor-network and posthumanist approaches that were and still are discussed in Anglophone geography. Kittler's notion of the materiality of communication could have produced some fruitful theoretical conversation with other materialist and posthumanist approaches, e.g. Latour's notion of 'inscription device' or Donna Haraway's 'informatics of information' (Schmidgen, 2019: 108). In this conversation, Kittler could have offered a conceptual lens to theorise the materiality of communication through Lacan rather than Latour. 16 Second, Kittler's interest in cybernetics, his idea of a 'cold model of structure', brought him in conversation with the systems thinking of Luhmann (Kittler, 1985(Kittler, /2003cf. Sale and Salisbury, 2015: xxx). Luhmann somewhat formed the elephant in the room of Neue Kulturgeographie: A number of German geographers had propagated Luhmann's Systemtheorie within German geography (Goeke, 2015;Goeke et al., 2015;Lippuner, 2007;Pott, 2005;Redepenning, 2006), but Luhmann's self-referential theoretical language placed them in the position of distanced observers (Pott, 2005;Redepenning, 2007). With a self-understanding as 'serious and ongoing attempts to relate a systematically structured social theory to the purportedly anti-systematic tenets of constructivism, poststructuralism, etc.' (Hannah, 2016: 72), Luhmann's thought style was incompatible with Anglophone geography's theory debates and thought styles. 'Luhmann' thus risked to be seen as 'provincial', locked within a distinctly German-language (geographical) scholarship. Kittler's Luhmann reception could have potentially helped to make German geographer's 'Luhmann' intelligible to an Anglophone audience of geographers.
To speculate about 'what could have happened . . .' amounts to a kind of counterfactual intellectual history of geography (Schultz, 2011), and such counterfactual histories are important to imagine more cosmopolitan geographies. Counterfactual histories do not only study successful cases of cosmopolitan theory (e.g. Italian Theory), but also cases where translations, travels and circulation across linguistic and epistemological boundaries did not take place as a basis to speculate about what could have potentially happened: Counterfactual intellectual histories make debates and translations discussed in Germanspeaking or other non-Anglophone territories of thought visible and intelligible for a cosmopolitan geography, rather than to bury them in a 'provincial' space of a language-bound scholarly community (Fall, 2013;Graefe, 2013;Houssay-Holzschuch, 2020;Houssay-Holzschuch and Milhaud, 2013;Korf et al., 2013;Minca, 2018;Wardenga, 2013).

Conclusions
To understand why 'German Theory' was never written, this paper has traced two different territories of thought in German-speaking scholarly communities, into which 'Foucault' travelled as an iconic figure of Theory and became re-packaged into two distinct forms of a 'German Foucault' -both concerned with discourse theory, but with different emphases: materiality (Kittler) and textuality (Neue Kulturgeographie). Kittler fused Foucault with Lacan; the discourse school of Neue Kulturgeographie combined Foucault with Laclau (and Mouffe). Kittler's Foucault, worked out in the 1980s, became highly influential in the Geisteswissenschaften and subsequently travelled to Anglophone media studies (albeit hardly to Anglophone geography). The German geographers' discourse school created another form of 'German Foucault' in the 2000s distinct from and without reference to Kittler's. Contrary to Kittler's 'Foucault', the geographers' 'German Foucault', albeit influential within a certain German-speaking academic milieu, did not circulate beyond linguistic boundaries.
These findings beg the question what makes theory travel in some cases and not in others -and travel to some territories of thought, but not to others. Three factors explain the nontravel of ideas in the case studied here.
First, an intellectual disconnect within German cultural studies. Debates in German geography were disconnected both, from the Foucault reception in the Geisteswissenschaften and from the latest debates in Anglophone geography. When German geographers embraced the cultural turn in the late 1990s, they absorbed the 'Foucault' of Anglophone new cultural geography under the banner of 'internationalisation'. Scholars in Neue Kulturgeographie quoted key theorists influenced by Foucault from Anglophone literary and cultural studies, e.g. Butler or Said (e.g. Lossau, 2002Lossau, , 2003Stru¨ver, 2003), but Kittler's 'German Foucault' did not travel to Neue Kulturgeographie. At the same time, as Neue Kulturgeographie was engaging Foucault's discourse theory, the latter was coming out of fashion in Anglophone geography: the textuality focus of Neue Kulturgeographie made Kittler's notion of the materiality of discourse unintelligible to German geographers unfamiliar with the terrain of Geisteswissenschaften exactly at a time when that notion of materiality of communication could have been of great interest in Anglophone geography's debates on materiality and non-representational theory, as Thrift's early (but not sustained) engagement with Kittler suggests.
Second, an organisational incapacity to make intellectual connections. In German geography, no 'cosmopolitan' geographer was in sight who could have worked as 'boundary spanner' (J€ ons, 2018;J€ ons and Freytag, 2016). The cosmopolitan geographer is someone firmly entangled in the multiple territories of thought that make up a linguistically-bound scholarly community and to juggle those 'worlds' as much as 'playing the big game of 'international' geography' (Minca, 2018: 11). Kittler's work was promoted by influential boundary spanners, especially Hans Ulrich Gumbrecht, a close friend of Kittler who had taught in German universities but moved to Stanford in 1989. Gumbrecht ensured the translation of his work into English, and Kittler was subsequently invited to teach at Columbia University and other prestigious departments in the United States, making his work influential in Anglophone media studies. In Anglophone geography, Claudio Minca might have played a similar role for Italian Theory: Italian in origin, Minca taught at prestigious departments in the UK and Australia and published a serious of influential articles on Italian thinkers, in particular on Giorgio Agamben. The crowd of German geographers gathered in the discourse school lacked such a scholar who could have acted as cosmopolitan boundary spanner.
Third, to make theories travel, the boundary spanner need not only provide a linguistic translation (from German to English), but a contextual translation of what the intellectual constellations are in which a particular argument has been made, a particular trope of theory debated. Take Kittler again: Kittler's theoretical work, argues Winthrop-Young, should be called 'German': Kittler's theory is not German because he was borne in Germany or writes in German, . . . [but] because the deep layers of his work, the bias of his arguments and the recurrence of a certain set of references and associations . . . have to be understood against a background of debates about technology, humanism, and individual as well as collective identity formation that over the last two centuries emerged in the German-speaking countries. (Winthrop-Young, 2011: 2) To attend to these references and associations requires multiple translations -practices of un/packing and re/contextualising to make them intelligible to Anglophone academic debates, and to make Theory fit for travel across the shores of language-bound academic communities.
Writing counterfactual intellectual histories exposes potential cosmopolitan geographies. Giorgio Agamben (1999: 183) writes that what is potential can both be and not be: 'What shows itself on the threshold of Being and non-Being . . . is not the colorless abyss of the Nothing but the luminous spiral of the possible' (Agamben, 1999: 257;my emphasis). Agamben follows here the impulse of Walter Benjamin (1991: 213) to 'read what was never written'. Reading Theory that 'was never written' thereby illuminates the luminous potentiality of cosmopolitan theory in geography. Seen in this way, 'German Theory' is a potentiality that has not yet actualised. We should therefore insist that 'German Theory' is worth pursuing: Just as Italian Theory has not been left to Italian geographers (Minca, 2016), there is much to discover for geographers around the world in these multi-faceted territories of thought we might call 'German Theory'. And this discovery might be 'rather fun too' (Fall, 2013: 58).
3. The trope of 'French Theory' became popular through Franc¸ois Cusset's book French Theory (Cusset, 2003(Cusset, , 2008. Cusset used the English term 'French Theory' in the title of the French original. Cusset claims that the reception practices of America's cultural studies operated through a different logic than did debates about poststructuralism in the French intellectual milieu and thereby made French poststructuralist thinking into 'French Theory'. It was largely the Anglophone package -a 'strategically selected version' (Minca, 2016: 827) of 'French Theory'that then travelled internationally. 4. Some observers claim that Francophone geographers had long been reluctant to engage with Foucault or other French poststructuralist authors (Claval and Staszak, 2004: 319;Fall, 2007aFall, , 2007b. As a result, 'Foucault' entered into Francophone cultural geography only through the reception of poststructuralism in Anglophone geography (Chivallon, 2003;Germes et al., 2011;Holzschuch-Housssay and Milhaud, 2013). 5. J€ ons and Freytag (2016) make this point in a special issue of Social and Cultural Geography entirely devoted to the intricate relations between Anglophone and German geography. Of interest to this paper are, in particular, the contributions of Mattissek and Glasze (2016) on the discourse school (whose work will be considered later in this paper in some detail); Best (2016)  Habilitation -an additional qualification after the PhD to receive the right to teach at a university -met significant resistance from influential German literary theorists who sat on the evaluation committee. 9. The event was seen as so successful that yearly gatherings of a succession of conferences have since been organised in different places, with continuing success and resonance. The most recent was held in 2020 in Bonn. See the conference series website: https://kulturgeographie.org/ (last accessed: 25 May 2020). 10. Summaries of all presentations are documented in Redepenning and Wardenga (2004).
Furthermore, the programme of Leipzig 2004 can still be consulted at https://kulturgeographie. org/ (last accessed: 25 May 2020). In addition, I draw on my own notes that I took from the plenary lectures and various panel presentations at the event. 11. It is important to note that despite the important role of the 'discourse school', Neue Kulturgeographie hosted a bouquet of different intellectual approaches. Lossau (2008: 321) subsumed (post-) structuralist, discourse-theoretical, practice and action-theoretical approaches and linguistic, semiotic and hermeneutic methodologies under the Neue Kulturgeographie corpus, while Berndt and Pu¨tz (2007: 9) mention poststructuralist, constructivist, practice-theoretical and discourse-theoretical perspectives. 12. This is interesting in so far as Kittler was heavily influenced by Foucault's 'Archaeology of Knowledge' as well, which he considered as 'a key component of the poststructuralist program' (Schmidgen, 2019: 110), but draws radically different theoretical conclusions from this engagement. 13. Mattissek and Glasze follow here Van Dyk and Angermu¨ller (2010: 15), who distinguish two traditions of Foucault reception in Germany: discourse analysis and governmentality studies. The former engages methodical-methodological questions of operationalising the concept of discourse, often combining qualitative and quantitative approaches, whereas the latter focuses on epistemological questions. As a result, the work of Thomas Lemke (1997), a central protagonist in both German and Anglophone governmentality studies, received only limited attention in German geography (see Marquardt and Schreiber, 2012). 14. A citation search on google scholar shows that while the Handbuch is widely cited across different German disciplinary fields (citation count as of 20 August 2019 ¼ 179), only a fraction of these citations comes from Anglophone sources, and from those, rarely is one in a geographical publication outlet. 15. Blotevogel (Schmidgen, 2019: 121).