A WORLD BEYOND WEB OF SCIENCE: AGORA MAGAZINE’S 35 YEARS IN DUTCH- LANGUAGE HUMAN GEOGRAPHY

Understanding human geography’s development requires looking beyond high- profile academic journals. Alternative publication circuits are spaces of experimentation and play an important role in the discipline’s intergenerational reproduction. This paper narrates the 35- year trajectory of AGORA Magazine, a Dutch– Flemish popular- scientific early- career publication focusing on a broad array of socio- spatial issues. Uncovering AGORA’s history guided by a network analysis of co- occurring authors, we develop two main arguments. First, the affiliation network extracted from AGORA’s archive separates the magazine’s history into five eras. These periods provide a window to understand the transformations of Dutch and Flemish human geography and related disciplines during the past decades, critically reflecting the zeitgeist of research topics, approaches and ideologies. Second, we probe the value of AGORA as nursery institution for early- career scholars, establishing myriad inter- university and transnational connections. We conclude by assessing the value of ‘second league’ native- language publications in the context of ongoing internationalization.


INTRODUCTION
There is an increased interest among historians of geography to narrate the discipline's varied histories outside its established canonical register (Norcup 2015). This paper probes the relevance of 'grey literature' for geography's development through the history of AGORA Magazine, 1 a popularscientific, Dutch-language magazine focusing on socio-spatial issues. AGORA's editorial board consists of early-career academics and young professionals from the Netherlands and Belgium, who all contribute voluntarily. Not only the magazine's editors write articles but also solicit external authors occasionally translate the latter's contributions. In 2019, AGORA celebrated its 35th anniversary and digitized all back issues for the occasion. 2 This digital repository offers a unique perspective on the development of Dutch-language human geography and related socio-spatial disciplines. Not only does the archive offer the opportunity to easily browse through This is an open access article under the terms of the Creative Commons Attribution-NonCommercial License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited and is not used for commercial purposes.
historical issues, it also allows obtaining metadata susceptible to network-analytical methods. Analysing AGORA's historical metadata, content and key actors allows delineating an evolving discipline and evaluating the relevance of a non-anglophone, early-career institution as a vehicle for (academic) community formation.
Although the organizational structure of AGORA changed significantly throughout the past 35 years, evolving from an Amsterdam-based consultancy-related publication towards a binational early-career academic magazine, it always remained loyal to its original motto. In the very first issue (AGORA 1984-0, p. 3), the editors proclaim that: 3 AGORA's most important goal is to reveal and make accessible a large part of the research that is being conducted at universities and colleges, namely that of the students. It is our experience that people are hardly aware of those studies. Resultantly, a lot of interesting research is being lost, and spatial planning research becomes needlessly scattered and inefficient. This is why AGORA's editorial board strives to be a bridge, between research and practice as well as between the different educational institutions.
Today, the magazine's website states that: It is the ambition of AGORA to reveal underexposed and relevant socio-spatial issues in a critical and accessible manner. AGORA strives to put these topics on the agenda, to formulate an opinionated vision and to put them into a broader perspective. AGORA is thereby creating a bridge between students and senior researchers, between research and practice, and between the Netherlands and Flanders.
Although the emphasis on the magazine's 'bridging function' never disappeared, the two quotes suggest a gradual development to a more critical stance. This change reflects the magazine's institutional embeddedness and its reaction to a changing zeitgeist, including not only societal developments but also the speedy anglicization of academic human geography in the Low Countries (see e.g. Schuermans et al. 2010). The following section elaborates why AGORA's trajectory is important for understanding the history of geographic thought. We argue that the changes in AGORA's positionality reflect the changing Dutch (and later Flemish) society from the mid-1980s to the present, including the roles human geographers and other social scientists saw for themselves. Nevertheless, history is not just structure, and one should not underestimate the role and impact of key actors who influenced the magazine's content and tone, as well as providing its actual 'bridging' function and the magazine's social reproduction. The paper continues by narrating AGORA's history through a focused reading of its 35-year catalogue, driven by patterns obtained through a network analysis derived from the archive's metadata. The network analysis allows identifying the key individuals who shaped AGORA's trajectory. Moreover, the modularity of the network helps uncover the temporal structure behind the magazine's continuity and change. These findings structure a historical narrative into five epochs. For each, we concisely discuss the development of the magazine in terms of organization, focus and views. The paper ends with summarizing the importance of AGORA for the intellectual and social reproduction of Dutch and Flemish socio-spatial sciences in general, and Dutch-language human geography in particular.

AGORA AS A MARGINAL WINDOW ON THE HISTORY OF DUTCH-LANGUAGE GEOGRAPHY
All knowledge is situated, and this holds especially true for geographical knowledge. How we perceive, act upon and write about the world, depends on who, when and where we are (Rose 1997). Although similar geographical questions arise in different places and moments, how they are engaged with is rooted in particular practices of geography (Buttimer 1983;Van Meeteren & Sidaway 2020). Moreover, how we write geography, what we deem important and worthy of inclusion is crucially dependent on our positionality in these debates (Sheppard 2015;Kulpa & Silva 2016;Oswin 2020). Whether we engage with geographical issues as students, teachers, civil servants, researchers with a blue skies research grant, activists or just as laypersons, changes the perspective. In old-fashioned practices of disciplinary historiography, there was little place for the situated messiness that such a plural history entails (Livingstone 1992). Old-school 'Whiggish' historiographical practices focus on the 'highest level journals', the 'most prestigious institutions' and the 'largest of achievements', which geographers have fairly criticized as being hagiographic, navel-gazing (Keighren et al. 2012) and overlooking the contribution of women and minorities (Jöns et al. 2017). However, writing an alternative history that is sensitive to these issues is easier said than done. Where does one begin? What is one's entry point? How to define selection criteria for what is important and what not?
It is increasingly clear that to write alternative histories of geographic thought, an engagement with the marginal spaces of knowledge (Oswin 2020) and publications outside the sphere of international academic peer review, dubbed the 'underground' (Larkham 1987) or 'grey literature', is imperative (Lorimer & Philo 2009;Norcup 2015). 4 It is in 'unofficial' academic publications, outside the hallowed halls of academic publishing, where alternative perspectives, such as the student perspective (Philo 1998;Bruinsma 2020), on disciplinary development may thrive. Moreover, the underground and grey literature often documents the position of 'marginal or hidden spaces' (Lorimer & Spedding 2002). Such marginal spaces on the intersections between institutions, that might bridge jurisdictions, universities or the ivory tower-practice boundary are often overlooked in historical accounts but can have profound influence on disciplinary development (Van Meeteren in press).
Van Meeteren and Sidaway (2020) argue that it is important to consider a set of interrelated publications such as the AGORA archive as being reflective of both cognitive and sociological structures. AGORA Magazine cognitively reflects the concerns that animate its early career and student editors, and how these change over time (Philo 1998). Sociologically, it is a formation where young geographers mingle, get to know each other and make a magazine together. This shared experience and social network will likely influence people's subsequent careers (Pred 1979).
Once you take marginal spaces of knowledge seriously, you quickly bump into a curation problem. As the grey literature is not considered that important, particularly its 'backstage materials' might not be carefully saved. For instance, Barnes and Sheppard (2019) note how difficult it was because of this curation problem to reconstruct the foundation of radical geography. From this profoundly student-driven radical episode in the history of geography that was guided by a mimeographed underground literature, only fragments of material remain, often overlaid with fallible and faded memories.
Retrieving these fragments, and piecing them together to build a coherent narrative of the past, is the work of archiving (Withers 2002;Ogborn 2011). But archives come in various degrees of orderliness, where only certain sources might have been preserved, telling only a partial story (Lorimer & Philo 2009). Although a full set of AGORA issues was preserved, prior to digitization it was largely confined to a cardboard box in an unremarkable storage room at Utrecht University. Here, a volunteer with the role of 'secretary' (usually an Utrecht University student) would dutifully add the new issue to the box every few months. Most current editors have little idea about AGORA's history, and after retiring from the editorial board, one seldom looks back. Apart from select occasions, usually jubilee editions, when former editors were asked for reflections (Kromhout 2009;Verwaaijen 2009;Pelzer et al. 2014;De Beer & Koelemaij 2019), the history of AGORA was a fragmented oral history of which nobody had a full picture. Apart from this box of old issues and piecemeal sharing of oral history, very little secondary documentation about the journal has been preserved. We know little, apart from cautious tales in old editorials, about how the number of copies printed evolved over the years, or how the financial record was in the past.
In the last 2 years, to celebrate the 35th anniversary, the archives were digitized by a dedicated volunteer group. Digitizing assembled and reshuffled the archive in profound ways. The newly digitized box, as a site of memory, travels much easier (Ogborn 2011), engendering new conversations between editors and readers alike. Digitizing the archive offers renewed opportunity to engage with older AGORA issues, allowing for comparison between how topics were treated in the past and the present. Moreover, digitization offers the opportunity to create archival metadata. Such metadata can tell us something about the digitized box's contents as a whole and the relations that its elements (i.e. individual issues) piece together. By conjoining information about authors that contributed to AGORA over the years, new knowledge about the magazine's history can be parsed.
Digitization allows more than merely describing the activity of authors through their contributions or the changing themes of individual issues. From a macro perspective, the relations between co-contributors form a larger network structure that can be used as a guide to read the magazine's history. What are the distinctive groups of people associated with particular phases in the journal's evolution? What kind of social structures were present that allowed the magazine to persist for 35 years? Who were the key players in those structures, and to what extent does the presence of these players tell us something about the position of AGORA in the Dutch (and later also Flemish) institutional field of the social-spatial disciplines? What does this tell us about which and how knowledges circulate? We outline those network-analytical methods and findings in the next section, before corroborating and contextualizing the resulting historical narrative through a reading of the archive, using the metadata as a roadmap. As there is no neutral place from where to conduct research, writing history requires being explicit about our positionality (Hawkins et al. 2014). We are four early-career researchers who studied (human) geography in the Low Countries: two of us at the University of Amsterdam, one at Utrecht University and one at Ghent University, whereas our doctoral research took place in Flanders (at the universities of Ghent and Leuven). We became actively involved in the editorial board of AGORA between 2009 and 2013, and we have met through the AGORA-network. As such, besides the aforementioned digitized archive and metadata, we also contribute our collective memories to this research.
Like with any map, this roadmap in conjunction with our positionality illuminates certain aspects of AGORA's history at the expense of others. For instance, we read AGORA's history through our experience of the editorial work in the present or recent past. As we juxtapose our personal experiences in the present with the published word of earlier generations some distortions could occur. Moreover, focusing on authorship metadata means that the work and influence of AGORA volunteers on the production, editorial and secretarial side is inevitably underexposed.

USING NETWORK-ANALYTICAL METHODS TO GENERATE AGORA'S HISTORICAL STRUCTURE
Visualizing the network -To generate metadata about AGORA's history, we parsed the digital archive into an affiliation network. Affiliation networks map actors that have co-participated in events (Borgatti & Halgin 2011); in this case, the event of having published in a particular AGORA issue. A link between authors is established when they published in the same issue. We dichotomized the links, and resultantly, frequencies of co-participation are not taken into account. This simplifies the interpretation of centrality measures (idem: 424) while slightly damping the relative influence of already central actors. In the resulting network, nodes represent authors contributing to AGORA at least once between 1984 and 2019. Links indicate co-presence in an issue. Co-presence means one cannot presume that there is a social link between authors, although in practice this will often have been the case between contributing editors. As AGORA's structure is based on thematic issues with a dedicated group of editors responsible for each issue, through writing articles and inviting authors to contribute, copresence in a certain issue does depict a certain level of (social) relatedness. Due to this structure, it is the editors that tend to be the more central nodes in the network as a whole. Thus the network provides an indirect and partial picture of the social structure around AGORA. We space the visualization with the Force-Atlas2 algorithm (Jacomy et al. 2014) that pushes unlinked nodes away from each other while pulling linked nodes together resulting in a curved shaped geometry ( Figure 1) consisting of 1482 nodes (authors) connected by 18,644 edges (affiliations).
The majority of authors (1189 out of 1482) appeared in one issue. These authors are therefore only connected to others in the same Figure 1. AGORA Magazine's affiliation network  issue, ranging from 3 to 30. This type of author is found in the many different small clusters of nodes close to each other but further away from other nodes. These clusters represent the different issues of AGORA. Authors appearing in several issues (293 out of 1482) are connected to all the authors in the various issues of AGORA with whom they co-appeared. They are the connectors between the small clusters in Figure 1. The Figure's structure allows telling a tale of temporality and centrality regarding AGORA's history.
Understanding the network's temporality -The colours in Figure 1 represent different author clusters, which have a high number of shared connections calculated by the Louvain clustering algorithm (Blondel et al. 2008). The algorithm produces a modularity score of 0.607 and generates 12 modules, of which five contain more than 100 nodes, four modules which are not connected to the main network at all and three modules which are connected but have a limited number of nodes. As the AGORA network represents a structure that gradually grew through time, it is unsurprising that the network's geometry and modularity reveal an epochal structure narrating the magazine's evolution.
The five main modules depict different eras, clockwise from the earliest module (purple) in the bottom right to the most recent module (yellow) top right. The modules' names are derived from the main developments in AGORA's history narrated below. Each module comprises a core group of editors contributing to a relatively large number of issues. Figure 1 shows that the magazine evolves through disruptions with a new group of editors joining every 5 to 10 years. This pattern is most striking between the two early modules, purple (STRABO) and magenta (SISWO), on one side, and the three more recent modules (green, blue and yellow) on the other side. This disjuncture is caused by very few connections between the authors in the second (magenta) and third (green) era.
Interpreting centrality measurements -Social network analysis provides a toolkit to distinguish particular types of authors based on their centrality. There are different centrality measures to depict the 'local importance' of a node irrespective of the whole network, or to position a node in the network as a whole (Freeman 1979). Here we use degree centrality to indicate the former role and betweenness centrality to indicate the latter. When appearing in more issues, an author connects to more other authors and thus the author's centrality increases. Figure 1 shows centralities in two ways: by visualizing connecting authors in central positions, and by scaling the node size by the number of others a node has a direct connection with, i.e. 'degree centrality' (Freeman 1979). The authors with the highest degree centralities are editors who have engaged with the magazine during several years.
Authors are also ranked by the type of connections. When authors facilitate connections to otherwise unconnected authors, they obtain a higher betweenness centrality (Freeman 1979). Authors who have only contributed to a single issue, and are therefore only directly connected to others in that issue, have a betweenness centrality of zero. Authors who feature in multiple editions facilitate connections between editions and therefore score a higher betweenness centrality. This score increases when the author is a bridge to relatively less connected sections of the network. They particularly increase in this network when authors are active in different epochs. Betweenness centrality is also used to filter the labels in the network. Authors with a betweenness score above 3.500 are labelled in Figure 1. Figure 2 plots all authors' degree (X-axis) and betweenness (Y-axis) centralities. Node colours indicate the temporal modules. Figure 2 reveals four types of authors. The first type (type I) are authors who contributed to AGORA once or twice and have a low score on both centrality measures. The type I quadrant contains the majority of contributing authors, and all temporal modules are represented. Type II authors have a relatively higher degree than betweenness centrality. They are connected to many other authors, but these connections facilitate fewer links to other parts of the network. Authors found in type II-a are either editors or more senior authors who appear in the magazine multiple times but tend to have contributed to one era only. Authors in type II-b are editors of AGORA in the most recent temporal module and have contributed to many different issues. Their relatively higher degree centrality compared with their betweenness centrality is because as a more recent author, one is less likely to make 'backward' connections to people involved in previous eras. Hence, they score lower on betweenness than previous editors with similar leading editorial roles.
Type III authors have a relatively high betweenness centrality compared with their degree centrality. This means they bridge modules but do so with relatively few publications. These authors appear in a diverse set of AGORA issues over a longer time span. The majority of the authors of type III have or used to have senior academic positions at Dutch and Belgian universities or are well-known public figures in Dutch-language socio-spatial policy debates. Most authors of this type have never been part of the editorial team of the magazine but have been invited to write contributions at different moments in their careers. This group of authors tends to be positioned in the centre of Figure 1, outside the curved geometry.
The last group of authors (type IV) score high on both betweenness and degree centrality. These authors have been active over a longer period and feature in more AGORA issues compared with type III authors. A number of former leading editors are in this group, several of whom eventually established an academic career at a Dutch or Belgian university. The green and magenta temporal modules are overrepresented in this group, which is partly caused by the limited number of ties between the earlier and more recent periods. Type IV authors are very likely to have played an important role in rejuvenating the journal as holdovers from a previous generation. Table 1 investigates prominent authors from Figure 2 that are located above the dotted regression line and have a betweenness centrality above 20.000. Salient features of Table 1 are the high number of career academics on this list and the fact that there has been a good deal of mobility of these academics between Dutch and Flemish universities. However, several consultancy and government careers have been launched after AGORA editorships as well. Another salient feature of Table 1 is the sheer gender imbalance of 'visibly influential AGORA contributors' 5 with only one woman (Lia de Lange) featuring in the list.
These centrality inferences based on AGORA's metadata provide a temporal and social map to read the magazine's archive. It helps indicate the historically salient needles in the 1000s of pages of archival haystack. The network indicates abrupt ruptures and bridges in the journal's trajectory and editorial team. We now continue to follow AGORA's history guided by this map.

AGORA'S FIVE LIVES SO FAR
The network analysis identified five different AGORA-eras based on pivot points in the network structure indicating changes in the journal.
We used this 'map' of the network to focus our reading through the AGORA archive to write a history of the magazine, by going back and forth between network analysis and reading, as well as between data and interpretation. As will become apparent from this cyclical analysis, there is a congruence between the development of the journal and the developments within Dutch-speaking human geography over the last 25 years.
The STRABO era (1984-1994) -In 1984, AGORA was founded by recent human geography and spatial planning graduates from Amsterdam's Vrije Universiteit (VU), who had established 'STRABO' 2 years earlier: a non-profit consultancy firm focusing on spatial planning and the built environment. AGORA's founding occurred at a dire moment in Dutch geography. In AGORA's founding year, the VU's geography department was dissolved (de Pater & de Smidt 1989, p. 352). As several staff members and students from the VU were integrated at the University of Amsterdam (UvA), graduates from the UvA soon joined AGORA's Amsterdambased voluntary editorial team. In these years, the job market for geography graduates had largely collapsed in the ongoing reverberations of the 1970s economic crisis. Previously, Dutch geography departments educated professionals for applied research, serving the welfare state. By the mid-1980s, as the state retreated and academic geography shrunk with it, young geographers reoriented their career ambitions towards the private sector (De Pater 2001, p. 182). Verwaaijen (2009) recalls how gloomy the prospects on the job market were during those years and how AGORA was first and foremost an opportunity for young professionals to present themselves to prospective employers.
Contrary to more traditional Dutch academic journals (such as TESG and the Dutchlanguage Geografisch Tijdschrift), the magazine tried to appeal to a broad audience bridging the worlds of academia, private sector and government through short articles, provocative titles, many illustrations, pictures and interviews. Each issue devotes multiple pages to an overview of recent students' theses. In those early years, reviews were strongly oriented to policy reports and other grey literature, as well as upcoming events that were announced under the heading 'Agoranomi'. As Verwaaijen (2009, p. 4) recalls: 'the platform function of AGORA in the pre-internet era should not be underestimated'. The magazine positioned itself as part of an institutional infrastructure where young graduates could stay in touch with the latest disciplinary and professional developments and career opportunities. The first issues focus on topics illustrative of the mid-1980s geography and spatial planning debates. An overarching theme is how to deal with fallout of the 1970s economic crisis and structural economic change. Was it better to double down on the 1970s approach of government-led urban renewal or with a new, fresh, market-led kind of urban development? (Van der Cammen & de Klerk 2003; Poorthuis & van Meeteren 2019). Recurring topics include not only decentralization of decisionmaking, privatization of government services, the worrisome disappearing of inner-city jobs, telecommunication and the emergence of new office locations, the decay of the postwar modernist estates, the increase of singleperson households through individualization and the ageing population but also the emergence of the first 'yuppies' and early practices of city marketing. A 1986 editorial (AGORA 1986-4, p. 2) shows AGORA's orientation towards economic geography: 'It is pleasing to note that a change occurs in terms of students' research. There is increasing attention for economic topics, which implies more choice for us regarding theses that deal with the spatial impact of economic and technological changes'.
From 1990 onwards, AGORA issues have a thematic focus where several articles address the same topic. This is complemented by several 'varied' articles and recurring features. This structure persists until today. During the STRABO-period, themes varied from a focus on particular Dutch cities or regions, to highlighting more underexposed topics (e.g. 'Garbage' 6 1993-1) and more commercially oriented ones (such as 'Marketing of housing ' (1992-5) and 'Retail' (1994-5)). This 'commercial' category, as Verwaaijen (2009) recalls, was strongly encouraged by STRABO's management. In return to STRABO's support, AGORA had to annually publish at least one 'commercial' theme issue, to obtain revenue from advertisements. Nevertheless, Verwaaijen (2009) notes that the editorial independence of AGORA was 'reasonably guaranteed'.
The early issues indicate how the STRABO interdependence worked out in practice. Although the STRABO influence shines through within outspoken policy-advice statements, there is no uniform discourse that contributing authors -often recently graduated students -adopt. AGORA, while under the umbrella of a private organization, was an open discussion platform expressing diverse viewpoints. Despite being a vibrant forum for socio-spatial academics and professionals, AGORA was unprofitable for STRABO, which had gradually transformed into a more 'professional' for-profit consultancy firm. From its changed business perspective, AGORA increasingly became an irresponsible burden. The magazine had already distanced itself from STRABO to a certain extent when Ed Slotboom became the first non-STRABO editor-in-chief (Verwaaijen 2009). In 1995, the Royal Dutch Geographical Society (KNAG) and the Interuniversity Institute for Social Science Research (SISWO) took over the financially struggling magazine.
The SISWO era (1995-2000) -KNAG and SISWO's adoption of AGORA came as a relief for the editorial board. The last issue of 1994 stated that the future of AGORA was highly insecure. The subsequent volume (1995-1), which introduced a renewed layout (Figure 3), included an official statement from the two new publishers. They stress that STRABO's withdrawal 'had nothing to do with the relevance and quality of the magazine, which had filled a gap in the provision of information on spatial planning and the built environment for over a decade'. However, the new publishers are reticent and state that '1995 can be seen as a phase in which it remains to be seen whether this new exploitation formula will be prone to continuation'. The editorial board subsequently stressed that 'we should shift to business as usual as soon as possible'. Behind the scenes, SISWO director Marten Bierman saw in AGORA the potential collective discussion platform in Dutch spatial planning that he thought the profession needed. 7 Although AGORA survived, the volatility of the magazine during these turbulent years is reflected by the rapid turnover of editors-inchief. Within 5 years, there were at least seven different ones, and AGORA's publication was sometimes delayed (1995-3). By this time, all editors were UvA-students and graduates, who were still motivated by the desire to contribute to the field and gain some relevant working experience to enhance their CVs (Bontje 2019).
Thematically, AGORA maintained a spatial perspective, whereas the themes varied between 'hot topics' to more original features such as 'The new way of working' (1996-1), 'Multinational catering industry' (1997-1), 'Small-scale aviation' (1997-3), 'Criminality'  and 'Religion' (1998-4). AGORA increasingly distanced itself from its semicommercial origins: shopping malls, real estate and retail blurred into the background, although they continued to receive attention outside the thematic sections. Although the magazine was no longer part of a consultancy firm, articles still tended to culminate in policy recommendations.
What also remained were thematic issues that specifically focused on one particular region in the Netherlands, such as ' South-Limburg' (1995-2), 'Node Arnhem-Nijmegen'  and 'Zeeland' (1999-4). AGORA kept its recurring feature sections, including the one page 'vision', often written by prominent, commonly White, male and senior voices, as well as the Agoranomi and the theses overviews, although these got less substantive in terms of length and frequency. Although most articles in this period were written by the editors, they also clearly approached other recent graduates to contribute. Additionally, articles were regularly written by more senior academics, civil servants or private sector professionals.
Although the magazine survived the rather threatening SISWO and KNAG communiqué for a number of years, the editorial board continuously feared AGORA's dissolution (Bontje 2019). Especially after a new generation of editors took over in the late 1990s, the relation between AGORA and SISWO, which was struggling itself due to ongoing budget cuts, became increasingly difficult. SISWO was an organization that had always promoted interuniversity and interdisciplinary dialogues and Dutch-language publications. It was founded as interface between the Dutch social sciences and social science professionals and civil servants (SISWO 1985). As such, publishing AGORA fitted their mission statement. SISWO took care of the layout, website and the secretariat of the magazine, but there was little engagement beyond that, and increasingly AGORA was considered an irresponsible cost item (Kromhout 2009). As SISWO's funding declined in the late 1990s although universities decreasingly valued cooperation within the Netherlands, SISWO started shedding publications before eventually ceasing to be a publisher entirely in the mid-2000s (Wester 2006). The end of the SISWO era marked the slow but steady transformation of academic focus towards the international (Anglophone) publication realm at the detriment of Dutch language academic publishing. AGORA had to reinvent itself in that changing conjuncture. Going Indie (2001-2006) -SISWO eventually requests AGORA to become financially selfsustaining. As a response, the editorial board convened a conference to become more visible and attract new readers. The conference, held in June 2000 in Amsterdam, focused on 'Neighbours and Neighbourhoods' ('Buren en Buurten'), turned out to be very successful in terms of attendance, content, visibility and thus forged new networks. A new generation joined the editorial board, and for the first time in AGORA's history, this included scholars from Flanders, the Dutch-speaking part of Belgium (Kromhout 2009;Loopmans 2019). Despite the conference being successful and leading to new subscriptions, the partnership with SISWO ends. The editorial in 2000-5 announced AGORA's next issue, although warning that even the editorial board was not sure if it would appear, as they had yet to find a new publisher. Even though not the initial goal, the Neighbours and Neighbourhoods conference generated the momentum to become an independent non-profit publication. An ad hoc funding application was granted, whereas more structural financial resources were secured through bulk subscriptions -where issues are procured at a large volume for distribution to students -by the universities of Amsterdam (UvA), Utrecht (UU) and Leuven (KUL), and through an agreement with the Royal Dutch Geographic Society (KNAG) (AGORA 2001-1, p. 2;Kromhout 2009). The editorial office moved to Utrecht University, and the layout of the journal was now in the hands of the UU's Cartographic Laboratory (KartLab). AGORA became strongly anchored at universities: the bulk subscriptions ensured student readership, new editors were students or PhD candidates rather than consultants at private companies, and the production process was institutionalized at UU. This rather abrupt change towards editorial independence and university interdependence shapes AGORA's trajectory going forward.
A first change is that AGORA takes a more critical stance. The editorial board could do so after it had found financial resources beyond commercial advertising. Key competitors in the Dutch-language publishing space, such as Geografie, Rooilijn and Stedenbouw & Ruimtelijke Ordening allegedly had a less critical viewpoint (Kromhout 2009), allowing AGORA to occupy the critical niche within Dutch-language socio-spatial publications. The focus of AGORA 3.0 shifted towards topics such as 'Sanctuary spaces ' (2001-5), 'Urban poverty ' (2003-1) and 'Postmodernity' (2002-4), whereas the relationship between research and policy was critically assessed in 2005-1. These topics replaced the previous, almost taken-for-granted focus on economyrelated issues such as shopping malls, and 'hobby-themes' (Kromhout 2009, p. 8) such as 'small-scale aviation ' (1997-3).
AGORA also gradually sheds its 'message board' function. From 2001 to 2011, the student theses overview is replaced by fewer but more extensive thesis reviews. The section Agoranomi, the notification board, disappears in 2003-5 and is replaced by a book review. The editorial board clearly desires more dialogue between the journal and its diverse readership by asking for letters and commentaries.
A third shift is the geographical and disciplinary diversification. Whereas before 2000, content was almost exclusively targeting the Netherlands, written by Dutch authors, indie-AGORA extends into Belgium in conjunction with its Flemish editors. Additionally, AGORA increasingly includes international case studies (such as Berlin, London, Portugal and Hannover), often written by international authors whose work is being translated into Dutch by the editorial board. AGORA's initial focus on applied geography and spatial planning, as exemplified in the subtitle 'meeting platform for science and society in the field of spatial planning', shifts towards a more diverse disciplinary focus, with contributions at the interface of human geography, sociology, anthropology, political sciences, urban design and spatial planning. AGORA's subtitle is renamed into a 'journal for socio-spatial issues'.
Multinational era (2007-2013) -In the following years, AGORA steadily grew into a critical, Dutch-speaking, popularizing journal in the lap of universities. This position is retained as a new generation of scholars takes over. Although it was sometimes difficult for the editorial board to simultaneously take care of content, promotion, a website and the magazine's finances, they did manage to publish five issues a year, even though publication was sometimes delayed. The editorial line of critical and societally relevant articles was kept. During the late-2000s, the share of Flemings in the editorial board increased, mainly coming from KU Leuven and Ghent University, with occasional guest contributions from Brussels (VUB) and Antwerp (UA). Dutch editors were still mainly situated in Amsterdam and Utrecht University, with occasional others from elsewhere or outside academia. Thematic issues are commonly alternately edited by an entirely Flemish ('Time' (2010-3), 'Nuisance' , 'Fear' (2009-4)) or entirely Dutch ('Cycling' (2010-4), 'Reconstruction' (2012-2)) editorial team, making it a multinational journal. As a result of this division, many Flemish contributions and contributors from this era form a separate cluster indicated in light blue in the upper left corner of Figure 1.
Although AGORA largely remains the same content-wise, the context in which it operates changes significantly. In the maelstrom of globalization, universities transnationalize in such a way that English is increasingly considered the lingua franca of scholarly communication (Mamadouh 2018;van Meeteren 2019). Within the European Union, this is shaped through the Bologna process (1999), regulating the comparability and standardization of Higher Education. In Flanders, this leads to the tax exemption of doctoral bursaries to deliver more international research output (PhD degrees and publications), whereas in the Netherlands educational programs were transformed into broader English bachelor and master programs to attract more students. Concomitantly, governments increasingly provide funding based on publication and citation outputs, especially those in high-impact English-language journals (Halffman & Radder 2015;Bogaert et al. 2017). Until then, writing and publishing in Dutch, whether in more academic or more popular journals, was considered a valuable way of disseminating academic knowledges and was therefore quite common for early career as well as established scholars. With the new financial allocation models; however, even tenured staff was pushed to shift publication strategy from writing for domestic geographic journals and audiences towards English Web of Science publications (Schuermans et al. 2010).
This trend presents a dilemma for AGORA. Contributions are mainly provided by (early career) scholars, who have less to gain in terms of quantifiable research output from writing for the journal. Starting in this period, PhD researchers who wanted to write for AGORA were sometimes actively discouraged by their supervisors to do so and were told to 'aim higher' in the publication hierarchy. The journal needed authors more than vice versa (Pelzer et al. 2014). Yet AGORA's readership remained Dutch-speaking: not only (undergraduate) university students but also consultancy firms and geography teachers. This audience values shorter, more accessible pieces that do not assume prior knowledge and language of academic debates. This increasing tension between academic and non-academic knowledges underlines the importance of AGORA's objectives, namely the dissemination of specialized academic knowledges to a wider audience. It does so by filling an awkward niche between Anglophone academic journals, and local, more policyoriented debates.
From 2008 onwards, the editorial board pushed to make AGORA even more accessible. It introduced a new artistic layout that was pursued independently of Utrecht University to be 'better aligned with the times' (AGORA 2008-1, p. 3) (see Figure 3). With newfound artistic freedom came new editorial responsibilities to maintain production. Gradually, new one-page sections (such as 'Free Space' and 'Classic') were introduced, as well as a more personal editorial by the editor-in-chief, and officially became a 'magazine' rather than a journal. As the then editor-in-chief put it: 'Rather coffee table than scientific library' (Pelzer et al. 2014, p. 56). Altogether, AGORA gradually drifted away from university departments' priorities, which became less inclined to spend money on bulk subscriptions. This divergence jeopardized the financial basis on which AGORA had leaned in the prior decade. The eventual cancellation of the University of Amsterdam's bulk subscription, once the home base of AGORA, when the magazine was no longer considered a budget priority in the context of a changing internationalizing curriculum, is symbolic in many ways.
Transnational connections (2014-2019 and onwards) -Dutch-Flemish relations in human geography and related disciplines intensified in recent years, for which AGORA bears some responsibility. The social linkages fostered in AGORA's editorial board meetings generated awareness of PhD and job opportunities across the Dutch-Belgian border. In the aftermath of the 2010 Eurozone crisis, opportunities for early career academics were limited, especially in the Netherlands. This explains why a number of Dutch AGORA editors ended up obtaining a Belgian PhD position, being hired by professors who had their own early-career histories as AGORA editors. 8 Intensified Dutch-Flemish collaborations forged a transnational community. Since the editorial board now reached all the way from Ghent to Groningen, the editorin-chief became a duo position in 2015, institutionalizing transnational collaboration throughout the magazine. By now, Utrecht established itself as the main AGORA-hub in the Netherlands, whereas the centre of gravity of Belgian editors fluctuates between student contingents from the universities of Ghent, Leuven, Brussels and Antwerp.
Topic-wise, urban studies are well represented. Issues about 'The post-socialist city' (2015-3), 'Smart city' (2017-1), or 'Urban metabolism' (2018-1) reflect this trend. AGORA's contributors originate more than ever before from a variety of academic backgrounds. Although the editorial board mainly consists of human geographers, urban planners, architects and urban sociologists, scholars from different fields are frequently approached to share their view on certain spatial issues. Although some chosen themes reflect personal editorial interests (e.g. 2015-2 'Space and history', 2018-2 'Food cultures', or 2019-1 'Music'), others rather contain output on behalf of a particular research group or project (e.g. 2015-4: 'The real estate/financial-complex' from Leuven, 2016-1: '(Un)Welcome' from Nijmegen, 2016-3 'Healthy city' from Utrecht, or 2017-4: 'Urban diversity' from the collaborate DIVERCITIES project on behalf of the universities of Amsterdam and Antwerp). Through opening AGORA's platform to communicate findings of academic research projects to a larger layperson audience, new relationships with the academic community are forged.
The critical attitude that AGORA proudly adopted a few years earlier still exists to a certain extent. Many contributions in the journal deal with issues like activism, inequalities, social justice and exclusion, climate change and environmental degradation. This has become, however, less of a contrast with general viewpoints of Dutch and Flemish socio-spatial academic interests. Although these became more critical in recent years, AGORA's former critical niche actually became more mainstream.
Although AGORA still aims to fulfil a bridging function between research and practice, and features a fair number of contributions from non-academic authors, it is striking that the vast majority of authors do have an academic affiliation. This includes the short one-page opinion pieces, whereas the books reviewed tend to be international scholarly publications. Furthermore, AGORA continued to develop an increasingly global orientation, thus being relatively less focused on socio-spatial issues in the Netherlands and Flanders itself. This is perhaps reflective of the tighter integration with academic research agendas. AGORA-affiliated earlycareer academics with teaching obligations furthermore tend to use it in their undergraduate courses. In other words, AGORA gradually transformed into a magazine 'by academics, for academics' while losing some of the feeling with the applied world it used to have. Besides content, AGORA is also illustrative for broader trends within Dutch-speaking Belgian and Dutch geography departments. The magazine has not always been successful in challenging and contesting the reproduction of a White, male and colonial bias that seems to endure within geography as a discipline (Bono et al. 2019;Oswin 2020) in terms of topics and who embodies it, despite continued efforts by many individuals, including within the editorial board.
AGORA's increasing academization engenders a variety of problems. Not only have universities become even more international, meaning that Dutch-speaking scholars (and therefore potential authors and editors) have become a minority, the increased anglicization of higher education also leads to more and more English-language university programs, which jeopardizes AGORA's potential readership. Universities nowadays prioritize spending money on access to international journals by for-profit publishers, something that is threatening AGORA's financial position. The broad orientation of the magazine makes it impossible to attract substantive advertisement income, whereas at the same time online blogs and social media compete for AGORA's readership. Yet despite, or maybe because of these circumstances, there is a decreasing number of alternatives where one can get a Dutch-language overview of contemporary scholarship on socio-spatial issues within the Low Countries in an easy and accessible way.

CONCLUSION
AGORA's 35 years of archive tells a multitude of stories. What is published in AGORA reflects the geographical worldview and concerns of the magazine's readers and makers. It paints a student and early-career perspective on debates and developments in Dutch language geography. But AGORA's tumultuous history also tells a story of the changing role of geography and planning as academic disciplines in the Low Countries. Between 1985 and 2020, Dutch and Belgian universities have changed enormously in their societal role, and AGORA has repeatedly reshaped itself to fill niches underserved by universities. Academic globalization made universities increasingly compare themselves with international rather than national counterparts, including the quasi-disappearance of Dutch as a language of education and publication. As a reaction, AGORA scaled up as well, from an Amsterdam-centred journal to one that is co-created by editors from most Dutchlanguage universities in the Netherlands and Belgium that offer education in geography and related socio-spatial disciplines.
Nevertheless, browsing 35 years of AGORA reveals that the topical focus of Dutchlanguage urban planners and human geographers is remarkably consistent, although the ways in which certain topics are being addressed has changed. AGORA slowly drifted away from the agenda of the academic institutions, partly retaining the local focus that universities lost. As such, AGORA stepped in a niche of local relevance in academic knowledge beyond Web of Science that Anglophone scholarship in the Low Countries might overlook (Aalbers & Rossi 2007;Schuermans et al. 2010;Houssay-Holzschuch 2020). However, there has been a necessity to align with the agenda of academic scholarship to some degree. As AGORA's contemporary editors' daily work is predominantly in international academic environments, the topical and theoretical focus does seep through its Dutch translation in AGORA Magazine.
Apart from being a custodian of local traditions of scholarship, AGORA also made significant contributions to the social reproduction of the academic field. AGORA played an active contribution in building a Dutch-language platform on a transnational scale. AGORA as an institution sustained a community of young scholars across universities. Within this scale and community, both academic discourse and social networks among early-career academics are reproduced. In successive generations of AGORA editors, the magazine functioned as a stepping stone for an international academic career. Through the AGORA alumni network, connections between aspiring scholars and established professors and professionals are formed, PhD positions are advertised and filled, and talent mobility, even though with a White, male bias, is facilitated across the boundaries of Dutch-language geography departments. These cognitive and sociological roles that AGORA performs in the reproduction of Dutch-language human geography underline the importance of minor league institutions as breeding grounds of aspiring scholars to support a healthy major league academic discipline in the long term. This importance is not just because an institution such as AGORA Magazine has a low barrier of entry, anyone willing can participate but also because it facilitates communication between universities that otherwise might not occur. Surveying the history of AGORA Magazine, we find that its role in forming early-career academic communities and networks underlines that AGORA as an institution serves a goal to the socio-spatial scholarly community beyond publishing articles.
Our reconstruction of AGORA's history guided by a network-analytical 'map' of cocontributing authors was enriched by our personal experiences contributing to AGORA over the past decade, yet we are aware that data-driven history writing has its blind spots. Our narrative would be different if we had relied more strongly on oral histories with former editors and other people involved behind the scenes. Nevertheless, given the availability of the digital archive and its metadata, we were able to narrate an overarching picture of the magazine: it allowed uncovering a historical structure in what was before a dusty box of old paper issues. Moreover, the digital approach made writing this history feasible without proper funding in a way that would have been difficult using more time-intensive methods.
Shedding light on AGORA's history urges us to wonder about the magazine's future. AGORA Magazine is not immune to the enduring crisis in publishing, and production costs rise from year to year, whereas the willingness to pay from individuals and universities for the magazine continues to flounder. The history of AGORA shows how it so far has always risen to the occasion of crisis to reposition itself in the disciplinary gaps left by others, and those gaps only seem to grow deeper as universities decouple from their local contexts. We hope that the archive and the histories it contains will offer much inspiration for many generations of creative young geographers to reshape AGORA in their image.