Peripheral Networks: Canon-Formation in the Nineteenth-Century Reception of Regionalist Writers

ABSTRACT

In many histories of nineteenth-century literature of the Low Countries, only a handful of authors are associated with emerging genres of regional fiction such as the village tale. In view of the immense popularity of regional literature, these histories overlook a large part of the literary field. This article aims to provide complementary perspectives by looking at how nineteenth-century Dutch and Flemish regional literature was framed and understood in contemporary reception, by systematically tracing all mentions of other artists and authors in reviews of selected authors. This has resulted in a ‘co-mention network’, revealing associations contemporary reviewers made between Dutch and Flemish regionalist writers and other cultural figures. This network functions as a starting point for investigating the impact of genre, gender and nationality on associations between authors, while also attending to which connections are not made, and which authors are left out of dominant narratives. Network analysis suggests that genre boundaries were fluid, and that numerous transnational authors were associated with Dutch and Flemish regionalists. At times nuancing or amending existing accounts of regionalism in literary histories, this article is also a tentative investigation of processes of canon-formation more broadly.

KEYWORDS:

• Canon-formation
• network analysis
• nineteenth century
• reception
• regional fiction
• regionalism

Introduction

In 1888, Jan ten Brink published the first volume of his history of Dutch literature in the nineteenth century. In this account, Ten Brink closely observes changes in the present-day literary field, trying to place authors who are his near-contemporaries within frameworks of genres and influences. One emerging type of literature that he aims to elucidate is the village tale, or ‘dorpsvertelling.’ Ten Brink believes the ‘nieuwere dorpsvertelling,’ newer village tale, originated with Jacob Vosmaer and Cornelis van Koetsveld.¹ Specifically, he identifies Van Koetsveld’s Schetsen uit de pastorij te Mastland (1843) as the first example of the village tale in the Netherlands. By referring to a ‘newer’ village tale, Ten Brink suggests continuity from previous types of literature in the seventeenth and eighteenth centuries that likewise focused on rural or village life and communities. Nevertheless, he believes the nineteenth-century village tale constitutes a distinct literary type that ‘should be defined and delineated more precisely’.² So far, attempts at defining and tracing genealogies of the village tale in the Low Countries are largely based on the analysis of selected texts. This article supplements existing scholarly accounts of regional fiction by relying on data-driven analysis of the nineteenth-century critical discourse surrounding the village tale and related modes of regional writing. Following Ten Brink, several twentieth-century literary histories trace these early Dutch village tales to later regional fiction, also known as local colour literature, or ‘streekliteratuur,’ in the Low Countries. In the 1920s, J. te Winkel mentions A.C.W. Staring, Van Koetsveld, and J.J. Cremer as early authors of village tales, and links these writers to Domien Sleeckx and Jan Renier Snieders (in Belgium) and authors such as H. Hollidee, Tonnis van Duinen, Pieter Heering, H.E. Beunke, Emile Seipgens and Cornelis van Schaick (in the Netherlands).³ Two decades later, G.P.M. Knuvelder draws a line from the village idylls by Cremer and Hendrik Conscience to more realistic stories by Johanna Courtmans-Berchmans, Sleeckx, Jan Renier Snieders, and August Snieders, and from there, to more innovative naturalist rural tales by Cyriel Buysse and Stijn Streuvels.⁴ More recently, Willem van den Berg and Pieter Couttenier have noted that the dorpsroman quickly became one of the representative genres of nineteenth-century Dutch and Flemish literature.⁵ They follow Ten Brink in naming Van Koetsveld as the originator of the local colour tradition in Dutch literature, and trace the genre from Van Koetsveld to Cremer to Seipgens and D.M. Maaldrink. For Flanders, they mention Conscience, A. Snieders, Sleeckx, and Virginie Loveling as important authors in this genre. Van den Berg and Couttenier also list Gotthelf and Auerbach as frequently-cited foreign writers; Auerbach’s importance as a writer of regional fiction is also emphasized by Ten Brink.

Ultimately, a handful of authors – the vast majority of whom are male – have come to represent the history and canon of regional fiction (understood here as an umbrella term for subtly distinct genres such as village tales and regional rural novels) in the long nineteenth century in some of the key literary histories of the Low Countries. These same authors are also mentioned in the Dutch literary lexicon (Algemeen letterkundig lexicon, 2012). The entry for ‘dorpsnovelle’ defines the village tale as a story describing life in a rural community. Cremer is cited as a notable nineteenth-century author in this genre, followed by Ernest Claes and Eelke de Jong in the twentieth century.⁶ The entry for ‘dorpsroman,’ village novel, mentions Gotthelf, Conscience, Streuvels, Timmermans, Claes, Antoon Coolen and Herman de Man.⁷ ‘Streekliteratuur,’ regional literature, is defined as ‘literature characterised by the description of a rural region and its inhabitants’ with a focus on ‘social relations in the community’, dialect, and folklore.⁸ The entry lists Cremer and Conscience as examples of nineteenth-century idealized and romantic visions of rural life, and names Streuvels, Timmermans, Claes, De Man, Coolen and Anne de Vries as key authors of twentieth-century regionalism. Gotthelf and George Sand are listed as transnational examples. Sand is the only female author linked to either genre in the literary lexicon. The basis for the selection of these central figures is unclear, but the same few names occur across all three definitions – for village tale, village novel and regional literature – as well as across several reference works of literary scholarship about the long nineteenth century.

Notably, the poor reputation of regional fiction⁹ has often contributed to the inclusion or exclusion of specific authors under this label: for instance, while Knuvelder lists Buysse and Streuvels as authors of naturalist village tales, Jacqueline Bel’s literary history of the twentieth century (2018), argues that Streuvels’ De vlaschaard (1907) ‘avoided the accusation of being regional fiction’ and that Buysse – who is linked to the genre by some literary historians – was likewise seen as a figure ‘far removed from regional literature’.¹⁰ Labelling stories of rural life by respected literary authors as ‘regionalist’ can, at times, be contentious, and Dirk de Geest et al. note that the ‘polemical attitude of literary critics’ as well as the genre’s ‘ambivalent profile’ and ‘repression by literary historians’ contribute to the difficulty of studying Dutch and Flemish regional literature (streekliteratuur), and to its status as a type of ‘“non-concept,” [simultaneously] applicable everywhere and nowhere.’¹¹

Defining or classifying regional fiction is therefore problematic, as is identifying a canon of its key works and authors. This article departs from predetermined definitions or characteristics and, instead, aims to create a clearer picture of associations with literary regionalism at the time, thereby challenging and supplementing the existing narrative of regionalism in literary histories. In The Center of the World (2018), June Howard claims that ‘[t]he conventions and themes we associate with any given genre traverse a variety of works rather than constituting a category that contains them,’ and that therefore

it does not make sense to assert a definition of a form, or to ask of any modern fiction, ‘does this belong to that genre? or not?’, as if the literary system were a series of pigeonholes and the problem was to find and properly label the right one. Rather, we need to ask: what is the intellectual history of this term, and what is usage doing with it now? […] We need (to put it another way) to refine a concept, follow it over time and across places and institutions, study its capacity for generating narrative.¹²

In line with this view, this article focuses on how literary regionalism was framed and understood in the long nineteenth century, starting in 1843, with the publication of Van Koetsveld’s Schetsen uit de Pastorij, as the village tale rapidly became more popular in the 1840s.¹³ We focus specifically on nineteenth-century book reviews in newspapers and periodicals, since these texts are valuable primary sources for tracing the literary debate surrounding new forms of writing and emerging literary movements and provide a helpful indication of contemporary views on the literary field.¹⁴

In particular, we consider the way the discourse on regional literature takes shape in the Low Countries through the lens of data-driven network analysis. Which authors were mentioned in reviews of regional literature, for instance because they preceded or inspired other authors or were seen as comparable? To what extent are certain key figures repeatedly associated with regionalism in twentieth-century literary histories – such as Van Koetsveld, Cremer and Streuvels – also recurring figures in nineteenth-century reviews? How do genre characteristics, gender, and nationality affect the likelihood of recurrent mentions in reviews? How does reception-based network analysis confirm, nuance or amend the dominant narrative of regional writing?¹⁵ By answering these questions, this article is ultimately also an investigation of processes of canon-formation, and offers some tentative suggestions for the possible origins of the dominant narrative and the key literary figures that emerge in historical accounts of Dutch and Flemish regional fiction.

Methodological Approach

In order to answer these questions, we propose a data-driven approach that is anchored in Karl Rosengren’s ‘mentions technique’ and the statistical techniques of network analysis.¹⁶ While close-reading based studies of regional fiction have produced a variety of useful insights into the (reception of the) genre, their inevitably confined empirical scope has failed to produce convincing bottom-up accounts of the contemporary discourse in which regional fiction took shape.¹⁷ The reason for this is twofold:

1. In terms of representative sampling, qualitative assessment of authors, texts and reception does not allow for generalizable statements on ‘the intellectual history of this term’ (Howard 2018: quote above). In absence of representative corpora, one simply cannot move from particular case studies to conclusions about a genre or discourse as a whole.

2. Subsequently, qualitative, close reading-based approaches to the history of regional fiction run the risk of replicating the dominant narrative as shaped by literary histories and the most influential literary critics. Opinions of established critics such as E.J. Potgieter and Conrad Busken-Huet often trickle down into literary histories, as their views are frequently cited.¹⁸ As a result, the dominant narrative of trends and genres tends to be a reproduction of the views of only a few prominent individuals, rather than an indication of the wider literary discourse of reception.

In order to create a more representative image of the contemporary discourse on Dutch regional fiction in the long nineteenth century, this article considers all periodicals reviews for works by selected authors that are, at the time of writing, available in two key online databases: the Digitale Bibliotheek voor de Nederlandse Letteren (DBNL) and Delpher. Although not all relevant materials have been digitized, it is safe to say that these databases represent the majority of digitized Dutch reviews in both literary journals and daily newspapers. To specify our time frame even further, we narrowed down our scope to the reception of regional fiction after 1843 (publication date of Van Koetsveld’s Schetsen uit de Pastorij), up to the end of the long nineteenth century (1914) with a small margin for works published and reviewed shortly after. Moreover, we only took reviews into account that have been published during an author’s lifetime or up until 10 years after an author’s death, in order to assure that we are tracking the contemporary Dutch discourse on regional fiction instead of later, literary-historical reflections or canonizations of authors or specific works.

Unavoidably, the list of authors we took as a starting point could not be entirely comprehensive: regional fiction was frequently published in the form of short stories in (regional) periodicals, sometimes anonymously, and short fiction was not always reviewed as extensively as longer works. Moreover, no clearly defined corpus of regional fiction exists, partly because it has not always been taken seriously in scholarship.¹⁹ As a starting point, we have therefore listed those authors who are frequently linked to regional fiction in mainstream literary histories, some of whom are primarily read as regionalist writers (for instance J.J. Cremer, Emile Seipgens, and Felix Timmermans), and who have therefore shaped the dominant narrative of regional writing as it exists in literary histories.²⁰ Broadening the scope of this study beyond a handful of ‘usual suspects’, the list also includes some more marginal regional authors who are mentioned only occasionally or tangentially, or whose names occur in regional literary histories but are excluded from national histories, such as Vrouwe Courtmans and Hendrik Tillema. This also allows for the inclusion of some additional women writers who would otherwise be excluded from this study, such as Truida Kok and Marie Gijsen. By approaching the genre from a ‘bottom-up’ perspective, and considering authors who are often overlooked alongside more canonical names, this article hopes to arrive at a more balanced perspective of the position of regional fiction in the nineteenth-century literary field.

Taking these considerations into account, we drafted a list of 35 authors (see Appendix A) for whom we systematically collected all reviews of their books published in newspapers and periodicals in the period 1843–1920. For the sake of practicality, this list consists only of authors whose names feature in the DBNL database, excluding some minor (dialect) writers whose works were not widely reviewed.²¹ We define ‘reviews’ as book reports published in periodicals falling in one of these four categories:

1. Announcements of books that just have been published or will be published in the near future;

2. Reviews of books, often containing a normative evaluation²²;

3. Essayistic reflections on recently published books, often from a broader perspective;

4. Obituaries that are published just after an author’s death.

As it was our aim to reconstruct the contemporary discourse on regional fiction, we did not take into account writings that are clearly historicizing, such as literary-historical pieces that tend to define literary trends or periods retrospectively. Although obituaries also have a historicizing function (in the form of a reflection on an author’s entire oeuvre), they are arguably still part of a contemporary discourse.

For each of the reviews of books by authors on our list, we manually tracked all mentions of other writers and stored them in a freely accessible database.²³ These mentions are proxies of ways in which associations between authors, texts and genres were established. This database provided the starting point for automatically extracting a ‘co-mention network’ of authors associated with regional fiction, which allowed us to trace how the discourse on Dutch and Flemish regional fiction took shape in contemporary reception. In this co-mention-network, authors are represented as nodes, and edges between these authors are established if they are co-mentioned in a review. The weight of these edges increases when there are more than two co-mentions. To give an example: J.J. Cremer and Berthold Auerbach are co-mentioned five times, sotheir relation (i.e. the edge between the nodes representing these two authors) is stronger than the relation between Cremer and George Sand, who are co-mentioned only once.

This extracted co-mention network serves as a point of departure for the analysis presented in the next section. Taking our cue from network theory, we analyse the structure of this network in terms of centrality and communities of authors associated with regional fiction, and evaluate these statistical observations in light of the dominant narrative on regional fiction as presented in literary histories from the nineteenth century onwards.²⁴

Results and Discussion

Our approach resulted in a list of 342 authors who are associated with (authors of) regional fiction.²⁵ These are either authors from the Dutch language area or authors from other countries with whom they are compared. Scanning the reviews of the authors on our initial list for mentions of other authors resulted in a list of 684 co-mentions, with a higher weight when two authors are co-mentioned in more than one review.²⁶

Network theory provides a range of statistical metrics to assess the centrality of nodes in a network. For present purposes, both indegree and outdegree centrality are most insightful. The indegree and outdegree centrality of authors in the co-mention network respectively represent the number of in- and outgoing links. More specifically, an author with high indegree centrality is often mentioned by reviewers, whereas an author with high outdegree centrality is often compared to other authors in reviews of their work. Scores on each of these metrics are thus proxies of the importance an author has in the contemporary local colour discourse.

Figure 1 and 2 display co-mentions networks in which node size respectively represents indegree and outdegree centrality. The most important authors in the discourse are those whose names are largest in the network visualizations. For sake of readability, a limited number of authors (weighted indegree centrality ≥ 3) is shown.

Figure 1. Co-mention networks of authors in the local colour discourse. Node size indicates weighted indegree centrality, node colour indicates gender (blue=male, red=female). Edge size and colour indicate weight. For sake of readability, only those authors with a weighted indegree centrality of 3 or more are shown (N=73).

Figure 2. Co-mention networks of authors in the local colour discourse. Node size indicates weighted outdegree centrality, node color indicates gender (blue= male, red=female). Edge size and color indicate weight. For sake of readability, only those authors with a weighted outdegree centrality of 1 or more are shown (N=30).

The first figure shows a range of authors with a variety of nationalities whereas the latter exclusively contains authors from the Dutch language area. This is a logical consequence of our research set-up: Figure 2 only shows authors with an outdegree centrality of 1, or, in other words, those who have one or more outgoing links. As such, this visualization only reflects those in our preliminary list of 35 Belgian and Dutch authors whose reviews contain at least one reference to another author. Walter Scott, for instance, is mentioned in a variety of reviews of authors on that list and thus has a relatively high indegree centrality, but he has as an outdegree centrality of 0 for the simple reason that he was not on the list that we took as a starting point for our research. As we did not systematically search for reviews of those authors who were later added to the list (such as Walter Scott), Figure 2 only represents the relative importance in the Dutch local colour discourse of authors on our preliminary list.

Conversely, Figure 1 shows the transnational ties within the contemporary review discourse on local colour fiction. While there are various Dutch and Belgian authors who are relatively central in both visualizations (e.g. Streuvels, Buysse, Cremer), Figure 1 indicates the extent to which the Dutch and Belgian authors on our preliminary list are compared to authors from other language areas, such as Dickens, Zola, Scott, and Flaubert. Noteworthy are those authors on our preliminary list who are more central in Figure 1 than in Figure 2. While, for instance, Hendrik Conscience has a relatively high outdegree centrality of 31 (which makes him the 12th most central author in Figure 2), his indegree centrality of 56 upvotes him to the most central position in Figure 1. Such cases are interesting because they show that some authors are mentioned more frequently in reviews about other authors than they are compared to authors in reviews about their own work. In the case of Conscience, this may be related to his important position in Flemish letters: famously referred to as the ‘man who taught his people to read’,²⁷ he was a popular and widely known Flemish folk author and the first to write literature in Flemish. Reviewers therefore seem to use him as anchor point in the Dutch-language discourse on regionalism, particularly in Flanders (hence his extremely high indegree centrality). However, they apparently feel less inclined to compare his works to others (hence his relatively lower outdegree centrality).

Before we discuss how Figures 1 and 2 relate to the dominant discourse on local colour fiction as expressed in literary histories of the Low Countries, we should first zoom out and look at the structure of the co-mention network we extracted from the reviews. Whereas the previous two figures filtered out a range of authors with low centrality scores, Figure 3 demonstrates what the network of all 342 authors associated with local colour fiction looks like on an abstract level. In this visualization, node size represents weighted degree centrality, which takes together both indegree and outdegree centrality and thus does not distinguish between in- and outgoing links. For the sake of both argument and readability, no names of the authors are shown.

Figure 3. Co-mention networks of authors in the local colour discourse. Node size indicates weighted degree centrality, node colour indicates gender (blue=male, red=female). Edge size and colour indicate weight. For sake of readability, no node labels are shown (N=342).

Based on visual assessment only, the periphery of the network takes up more space than its centre, which suggests that a considerable number of authors associated with local colour fiction are mentioned sparsely. Furthermore, there is a limited range of very strong ties between authors: only the centre of Figure 3 shows nodes that are strongly connected, whereas the thin lines between most other nodes suggest sparse density, which is confirmed by the very low graph density of 0.006 (on a scale between 0.00 and 1.00). Looking at the data on which this visualization is based further confirms this image: high edge weights between authors are uncommon, the vast majority of co-mentions are only made in one particular review. This suggests that the contemporary discourse surrounding regional literature was highly fragmented, possibly because regionalism was not recognized as a distinct literary genre.²⁸ A similar network analysis for other literary genres or movements, such as Anglophone modernism, might be expected to reveal a network with more strong links and clusters of authors frequently mentioned together, because a more coherent group of authors tends to be associated with this movement.²⁹ The scattered network might also help explain the lack of consensus on the parameters of regionalist fiction or its central figures in literary history³⁰: contemporary nineteenth-century reviewers make numerous connections between regionalist writers and other authors, but strong links between regionalists on our initial list are limited, with a few notable exceptions. An obvious interpretation is that the local colour discourse is centred around a few key figures, although it should be noted that each of these 342 individual authors jointly contribute to the discourse around the genre. Notably, several of these key figures overlap with those who continue to be associated with regional literature: Figure 1 shows that Conscience, Streuvels, Buysse, Cremer, V. Loveling and Auerbach – recurring names in literary histories and the lexicon – are important in the network. Our analysis thus shows that when it comes to identifying the central actors of literary regionalism, there is some overlap between nineteenth-century reviews and later literary histories. The following sections consider the role of genre, gender, and nationality in forming a ‘canon’ of regionalist literature, offering possible explanations for this study’s quantitative findings.

Regionalism as Genre

An analysis of the data provokes the following question: what is the role of regionalism as genre in the network that emerges from reviews? In other words, to what extent do reviewers make comparisons on the grounds of characteristics associated with regional fiction in literary histories and lexical works, such as rural setting, use of dialect, small communities, or folklore? A preliminary answer already emerges from Figure 1, which reveals the authors with the highest in-degree centrality. To an extent, the central authors in the reception network overlap with the regionalist authors mentioned in key literary histories: Conscience, Cremer, Streuvels, Buysse and Virginie Loveling are prominent in both, and Auerbach – mentioned by several literary historians as an important foreign influence – is also mentioned comparatively frequently, as is George Sand. As Figure 1 shows, Conscience’s role in the network is particularly important: he is the most frequently mentioned author. As previously explained, this is in line with his reputation in Flemish literature more generally, but it also confirms his importance to later regionalist writers: several literary histories draw a line from Conscience to authors like Sleeckx, Snieders and Streuvels, and contemporary reviewers also make these connections. Conscience is also mentioned in multiple reviews for Cremer, Reimond Stijns, Virginie and Rosalie Loveling, and Johanna Courtmans-Berchmans.

However, some of the authors that show up prominently in reviews are not associated with regional literature in literary survey works. Emile Zola, Charles Dickens, Herman Teirlinck and Guido Gezelle all feature in the top 15 of highest ranked authors on indegree centrality (see Table 1). Gezelle was Stijn Streuvels’s uncle and the two are frequently linked (weight = 9), which is perhaps unsurprising due to their shared connection to West-Flemish writing and their familial connection: compare, for instance, sisters Rosalie and Virginie Loveling, who are also frequently mentioned together (combined indegree/outdegree weight = 18). Gezelle wrote rural poetry in West-Flemish dialect and is primarily linked to Flemish authors (Streuvels, Timmermans, Loveling, Buysse), but also to Bernard van Meurs, who wrote dialect poetry in Dutch, which explains why his name occurs frequently in reviews. Herman Teirlinck’s central position is also not altogether surprising: several reviewers link him to Streuvels or Buysse, and all three authors were associated with the Flemish literary and cultural magazine Van Nu en Straks. Knuvelder’s literary history also repeatedly mentions Teirlinck and Streuvels together in this context, but does not identify Teirlinck with literary forms of regionalism; Bel likewise links Teirlinck to Streuvels through their work for Van Nu en Straks, but also briefly mentions some of Teirlinck’s rural novels.

Table 1. The top 15 highest ranked authors, based on weighted indegree centrality.

Label	Interval	Gender	Origin	Weighted In-Degree
Hendrik Conscience		M	Belgian	56.0
Stijn Streuvels		M	Belgian	48.0
Emile Zola		M	French	32.0
Virginie Loveling		F	Belgian	27.0
Cyriël Buysse		M	Belgian	27.0
Charles Dickens		M	English	26.0
J.J. Cremer		M	Dutch	25.0
Herman Teirlinck		M	Belgian	24.0
Rosalie Loveling		F	Belgian	19.0
Guido Gezelle		M	Belgian	18.0
Nicolaas Beets		M	Dutch	16.0
Guy de Maupassant		M	French	13.0
Israël Querido		M	Dutch	13.0
Berthold Auerbach		M	German	13.0
Jacob van Lennep		M	Dutch	13.0

The central positions of Dickens and Zola are also easy to explain since both authors were widely read and well-known in the Netherlands and Flanders, which means reviewers could assume that readers would be familiar with their work.³¹ In general, an author’s contemporary fame seems to play into their prominence in the review discourse: many central figures in the network continue to be canonical authors today. This type of data analysis might therefore also contribute insights into dynamics of canonization more generally. Moreover, realism and regionalism are often strongly connected, and regional writing was also described as a ‘healthy realism’.³² Meanwhile, Flemish authors like Streuvels and Buysse are sometimes linked to naturalism, perhaps explaining why some naturalist writers rank highly in terms of centrality: both Zola and Guy de Maupassant are well-known representatives of the French naturalist school and both are repeatedly mentioned in reviews for Streuvels and Buysse. The strong connection between these two authors and (French) naturalists might be part of why some literary historians suggest that Buysse and Streuvels themselves are not, in fact, regional authors: this assumption makes sense if regionalism is equated with sentimentalism.³³ However, Streuvels is also repeatedly linked to Conscience (weight = 6) and Virginie Loveling (weight = 4), and he is mentioned in several reviews about Felix Timmermans (weight = 4) and Marie Gijsen (weight = 4). Buysse is mentioned in reviews for Virginie Loveling (weight = 3) and H.E. Beunke (weight = 1). Both authors are, thus, firmly linked to both Dutch and Flemish regional writers, as well as naturalist and realist writers and even several painters, such as Emile Claus and Jean-François Millet: this intermedial connection between literary regionalism and visual arts does not always receive much attention.³⁴ Our data-driven approach to the review discourse is useful for revealing these kinds of patterns of association, which qualitative analysis of individual reviews might overlook.

Slightly less central, but still comparatively high-ranking, are several authors of historical novels, such as Jacob van Lennep, Walter Scott, and Geertruida Bosboom-Toussaint. The historical novels of these authors were well-known and widely-read in the Netherlands. Historical novels share several features with contemporary regional fictions, perhaps most importantly the interest in representations of couleur locale, and Scott’s works are sometimes mentioned as precursors to contemporary regionalism.³⁵ Perhaps equally important is the fact that several regionalists on our initial list, like Conscience and Cremer, wrote historical novels themselves before shifting their attention to contemporary rural life. The genres – insofar as they can be treated as fully distinct – have many traits in common, such as detailed descriptions of setting or dress, and use of language specific to time and place. The central position occupied by historical novelists shows that contemporary reviewers repeatedly linked the two.

However, ultimately, several authors with high in-degree centrality are not mentioned in connection with regionalism in literary histories at all, nor are they linked to the genealogy of the village tale. Their central position might suggest that regionalism or ruralism were not necessarily seen as the most salient features of (some) regionalist works, and that regionalist writing was reviewed in relation to other genres and a wider transnational literary field. Of course, network analysis paints only a partial picture of the discourse, and zooming in on individual reviews reveals that there is the occasional reference to genre: J.H.C. Heijse writes in De Gids in 1872 that ‘stories of folk life’ are in vogue, and reviewer H. van Loon likewise writes that ‘the “peasant novel” seems to be in fashion’.³⁶ Qualitative analysis of individual reviews therefore nuances or explains some of the data-driven results from our network analysis: in this case, it shows that regional stories are linked to broader trends in some reviews that do not mention specific authors or examples. Even though the role of ‘regionalism as genre’ has a limited impact on centrality scores, several reviewers are explicitly aware of the emergence of rural village tales as a (transnational) trend.³⁷ However, data-driven analysis shows that there are few attempts at establishing a canon of associated authors in reviews. This might explain, at least in part, why the network is far from self-contained, and instead consists of a sprawling constellation of authors with numerous marginal figures: the mode and median weight are 1, meaning the vast majority of co-mentions occur in only 1 review.

Women Writers

Another question raised by the data visualization is the impact of gender. Assessing the colour of the nodes in Figure 3 shows clearly that male authors are most visible in the discourse. Almost 90% of authors associated with the genre are male (see Figure 4). In terms of visibility, our discourse analysis suggests that female writers are peripheral figures, which seems to be in line with the dominant discourse on local colour fiction as expressed in Dutch literary histories.³⁸ However, a closer look at the network centrality of both male and female authors in the discourse creates a different impression. Table 2 shows that the difference between the mean indegree centrality of male and female authors is small. Although male authors’ visibility is much higher in the local colour discourse, they are almost as important as female authors in terms of indegree centrality. This is due to a few highly central female authors, such as the sisters Virginie and Rosalie Loveling (see Table 1 for the top 15 highest ranked authors on indegree centrality). Although not ranked in the top 15, Bosboom-Toussaint, George Eliot and George Sand also have comparatively high centrality scores. A more specific statistical test confirms this image. A linear regression was calculated to predict the indegree centrality of authors based on their gender. This test did not reach statistical significance (F(1, 341) = 0.080, p = 0.778, R² = 0.0002), which means that gender is not a significant predictor of importance of authors in the local colour discourse in terms of indegree centrality (see Appendix B for all relevant statistics).

Figure 4. Relative frequency distributions of gender among authors associated with local colour discourse (N =342).

Table 2. Mean weighted indegree for authors associated with local colour fiction (N = 342) separated by gender.

Weighted Indegree Centrality
Gender	N	Mean	Std. Deviation
Male	301	3,00	5,699
Female	41	2,73	5,119
Total	342	2,96	5,626

However, our initial list is largely based on literary histories, and therefore on a dominant and historicizing discourse. Very few women are linked to regional fiction in these works: here, too, Virginie Loveling is an exception. This might mean that for women writers to be included, their role in the contemporary discourse needed to be significant: more marginal women writers might simply not have drawn the attention of literary historians. To assess the role of gender more accurately, we have included some women writers of regional fiction that are often excluded from literary histories, such as Truida Kok, Marie Gijsen, and Johanna Courtmans-Berchmans.³⁹ The indegree centrality of these authors is low: Courtmans-Berchmans is mentioned three times, despite reviews of Courtmans-Berchmans’s own work noting her fame and popularity amongst contemporary readers.⁴⁰ Kok and Gijsen are not mentioned at all, despite writing village tales (about life in ‘t Gooi and in Brabant, respectively), and despite thematic similarities to authors like Cremer. While statistical analysis suggests that there is no gender bias in the bottom-up discourse surrounding local colour fiction – gender is not a statistically significant predictor of centrality – quantitative analysis nuances this view. The numbers are possibly skewed by both the low number of women on our initial list for analysis and the high centrality of both Loveling sisters.

Something similar holds for outdegree centrality. Remarkably, the mean outdegree of female authors is higher than it is for male authors (see Table 3), which can be explained by the few, relatively central female authors who were on our initial list (see Table 4 for the top 15 highest ranked authors on outdegree centrality). A linear regression analysis, however, also did not reach significance for this centrality metric (F(1, 341) = 2.930, p = 0.088, R² = 0.009), which indicates that gender is not a significant predictor of importance of authors in the local colour discourse in terms of outdegree centrality (see Appendix B for all relevant statistics). Once again, a slightly different picture emerges from qualitative analysis. For instance, closer examination of the collected reviews reveals that Kok, despite writing regionalist fiction, is not linked to literary regionalism by reviewers: the majority of reviews read her work as youth literature for girls, perhaps due to the author’s own gender.⁴¹ Moreover, it is important to note that while Courtmans-Berchmans is often linked to (male) regional writers in reviews of her work, especially Conscience, she is hardly ever mentioned in reviews about others, with the exception of the Loveling sisters and one review about August Snieders. The fact that Courtmans is almost exclusively mentioned in reviews about other women writers, while being all but ignored by reviewers of male regionalist authors, is telling. This suggests that both literary histories and nineteenth-century reviews might be susceptible to overlooking or marginalizing the contributions of women writers, or to highlighting their gender over contributions to a specific genre, even though statistical analysis implies that gender has no impact on centrality in the network.

Table 3. Mean weighted outdegree for authors associated with local colour fiction (N = 342) separated by gender.

Weighted Outdegree Centrality
Gender	N	Mean	Std. Deviation
Male	301	2,48	13,114
Female	41	6,54	20,834
Total	342	2,96	14,282

Table 4. The top 15 highest ranked authors, based on weighted outdegree centrality.

Label	Interval	Gender	Origin	Weighted In-Degree
Stijn Streuvels		M	Belgian	145.0
Cyriël Buysse		M	Belgian	108.0
Virginie Loveling		F	Belgian	96.0
Rosalie Loveling		F	Belgian	86.0
J.J. Cremer		M	Dutch	74.0
Lode Baekelmans		M	Belgian	72.0
Reimond Stijns		M	Belgian	44.0
Felix Timmermans		M	Belgian	40.0
Marie Gijsen		F	Dutch	39.0
Josef Cohen		M	Dutch	39.0
August Snieders		M	Belgian	35.0
Hendrik Concience		M	Belgian	31.0
H.E. Beunke		M	Dutch	28.0
Cornelis van Koetsveld		M	Dutch	26.0
Margo Scharten-Antink		F	Dutch	26.0

Transnational Connections

Finally, the visualization in Figure 1 shows numerous transnational connections: in fact, although the majority − 60,6% – of authors in the network is either Dutch or Belgian, foreign authors are important in the network both in terms of the presence of these authors in the network (39,4%) and centrality. Zola, Dickens, de Maupassant and Auerbach are amongst the 15 authors with the highest indegree scores (see Table 1). Figure 5 shows the relative frequency distributions of nationality of authors associated with the genre. Not surprisingly, the Dutch and Belgian nationalities are most prominently present. This is followed by French, German, English and American, and a range of 14 other less frequently mentioned nationalities. These include well-known canonical authors such as Gustave Flaubert, Mark Twain, Edgar Allan Poe and even William Shakespeare, as well as painters such as Gustave Courbet and Jean-François Millet. Again, the distribution of nationalities is not entirely surprising: the majority of works published in the Netherlands in the nineteenth-century were translations (explaining why foreign authors make up a large percentage of co-mentions), and translations from the English and German were most popular. Between 1845–1899, the market share of first-edition novels translated from the English language ranged from 21% to 39%, while novels from the German made up 14% to 30% of the literary production in these years. Translations from the French made up 11% to 19%.⁴² For other languages, this percentage was usually between 5–10%, with the exception of 1845–1849, due to a sudden spike in the popularity of Swedish novels.⁴³ The fact that French authors make up 12,9% of the nodes in the network – more than any other group besides Dutch and Belgian – is largely attributable to the numerous connections between Belgian and French authors, which make up 29% of all co-mentions.

Figure 5. Relative frequency distributions of nationality among authors associated with local colour discourse (N =342).

Going beyond importance in terms of frequency of occurrence, Figure 6 shows the mean weighted indegree centrality of authors separated by nationality. Scottish authors rank highest here, which can be explained by the fact there are only two Scottish authors in the dataset (Walter Scott (indegree: 9), Robert Louis Stevenson (indegree: 1)) and their average indegree centrality is 5. Conversely, there are more Dutch and Belgian authors in the dataset, respectively 124 and 83, most of whom have low indegree scores around 1: only few of them have high indegree scores (see Table 1). Interestingly, Belgian authors rank higher than Dutch authors in terms of both mean weighted indegree and outdegree (Figure 7). This might be partly explained by the higher number of Dutch authors: there are simply more marginal authors on the list affecting the mean. In other words, because there are more Dutch authors on the list, and because the vast majority of all authors have a centrality score of 1, a higher number of authors in a specific category (in this case, authors with Dutch nationality) makes it statistically probable that the mean in this category is lower. Moreover, some Flemish authors, such as Virginie and Rosalie Loveling and Streuvels and Buysse, are strongly interconnected. Conscience’s high centrality score, too, depends largely on Flemish connections: he is mentioned in reviews for Seipgens and Cremer, but does not occur in other reviews for Dutch writers. His central position in the network is therefore based almost entirely on his influence on fellow Flemish writers. There is thus a discrepancy between importance in terms of simple frequency of mentions of authors and importance in terms of network centrality. Both are indications of the influence authors, individually or collectively, have had in shaping the discourse on local colour fiction.

Figure 6. Mean weighted indegree for authors associated with local colour fiction (N = 342) separated by nationality.

Figure 7. Mean weighted outdegree for authors associated with local colour fiction (N = 342) separated by nationality.

Nationality also seems to impact the network in other ways, which might be demonstrated by the relatively low weighted indegree centrality score of Cornelis van Koetsveld. Despite the fact that Van Koetsveld remained undeniably popular throughout the nineteenth century – his Schetsen uit de Pastorij te Mastland (1843) was reprinted 15 times before 1918 – he is not mentioned nearly as often as Cremer and Conscience by reviewers. Several literary historians mention this work as central to the genealogy of regional fiction, identifying it as one of the earliest village tales in the Netherlands: in this respect he seems to occupy a similar position to Conscience. However, he is not nearly as influential in the contemporary reviews of other texts. Notably, Van Koetsveld is exclusively mentioned in reviews for Dutch authors, whereas Cremer is also mentioned in reviews for Conscience, Sleeckx, A. Snieders, V. and R. Loveling, Streuvels, and Buysse. Van Koetsveld might simply not have been considered to be as relevant for his Flemish contemporaries. These kinds of trends and gaps in the review discourse would be hard to detect with qualitative analysis only.

Overall, the network of co-mentions also demonstrates the importance of looking beyond the confines of national or linguistic boundaries in discussing regional fiction. As Howard has pointed out, ‘the local and the global are by no means distant ends of a continuum’: regional fiction is always relational, already linked to both the national and impacted by transnational connections.⁴⁴ This has not always been acknowledged. For instance, the Algemeen Letterkundig Lexicon suggests that some works match the characteristics of regional literature but are not considered as such, because the subject matter ‘transcends’ the local.⁴⁵ This distinction between ‘merely-regional’ and ‘also-regional’ is highly artificial, demonstrated not only by the fact that several foreign regionalist writers (Auerbach, Jeremias Gotthelf, Fritz Reuter, George Sand and others) show up in Dutch and Flemish reviews, and their relevance and readerships thus clearly transcended the locale of their fiction, but also by the sheer volume of references to other countries and other genres in the discourse around regional texts.

Conclusion

Ultimately, these co-mention networks also provide interesting insights into the processes of canon-formation. Although literary histories necessarily paint a more limited and focused picture of regionalist writing than the sprawling network of associations that emerges from contemporary reviews, it is telling how many of the most prominent authors associated with regionalism reveal overlaps between the two discourses. Moreover, although the reception-based network of co-mentions shows many names which are not associated with (literary) regionalism, the authors whose names occur most frequently – including Zola and Dickens, who both feature in the top 10 – occupy central positions in European literary history. Several household names, such as Shakespeare, Goethe, Victor Hugo and Edgar Allan Poe (Figure 1), are mentioned repeatedly. As the introduction points out, regionalist fiction is marginalized in many literary histories: it is, in that sense, peripheral. However, although the network takes authors associated with literary regionalism as its starting point, it is, in some ways, less a reflection of this genre, and rather a reflection of canonicity more generally.

The variety of associations − 342 names occurring across reviews for only 35 authors – shows how easily actors in the cultural field are linked across borders, genres and media. Foreign authors make up nearly half the network and therefore have an enormous impact on the understanding and framing of regionalism as a (developing) genre. The network also reveals the fluidity of genre, suggesting that use of dialect, elements of folklore, or (regional, rural) setting are not necessarily the most salient features for comparison in reviews. In literary histories and lexical works, these are precisely the grounds for grouping regionalist authors together, but this understanding seems to be imposed retroactively: for contemporary readers and reviewers, other connections, and links to canonical names, might be more important. Other than literary histories, the co-mention network also suggests the importance of looking beyond literature as a medium, as several connections are made between literary works and the visual arts. This is a testament to the importance of considering literature as embedded in a wider cultural field that includes other media.

The analysis above also points to the importance of investigating literary canons critically, attending to who might be excluded, and why. A popular author like Courtmans-Berchmans does not occupy the same position as her contemporary Conscience, either in literary histories or reviews. Unlike Conscience, she is almost exclusively mentioned in reviews about the work of other women writers: it is hard to imagine that gender does not play a role here. Some other women writers, such as Truida Kok, are overlooked entirely. Interestingly, however, gender does not predict centrality in the co-mention network. The few women writers who recur in literary histories are comparatively central in reviews, perhaps suggesting that in order for a woman to be included in the literary canon, she has to reach a certain level of fame and recognizability amongst contemporaries.

These findings require the combination of data-driven with qualitative analysis. The analysis of individual reviews would not reveal that Van Koetsveld is mentioned in reviews about fellow Dutch authors but not in reviews about Flemish writers, nor would it reveal that Courtmans-Berchmans is mentioned mostly in reviews about other women: empirically informed analysis, in this case, benefits an understanding of the impact of characteristics such as genre, gender and nationality. At the same time, to make sense of statistical findings, it is important to return to individual reviews: the fact that an author’s gender does not statistically predict centrality in the network should not be taken to mean that it has no impact on the way authors are compared in individual instances.

Many of our conclusions are tentative. This article is, in many ways, exploratory, offering some initial insights into the dynamics of canon-formation and the nineteenth-century framing of regionalist fiction. It raises many questions, not only about the way similar methods would unfold for other genres, but also about the impact of other characteristics. For instance, what would a network analysis reveal if authors were tagged not only by nationality and gender, but also by religious affiliation? Moreover, more detailed close readings of reviews might nuance these insights: as briefly demonstrated above, some reviews mention genre trends or similarities to other national literatures without mentioning specific authors: co-mention networks alone would not reveal these kinds of observations. To enable other researchers to continue investigations along these lines, we have made the data (reviews, nodes, and edges) collected for this study available in an open-access database.⁴⁶

Acknowledgments

The authors would like to thank Ninge Engelen and Nils Lommerde for their invaluable contribution to the data collection for this article.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data that support the findings in this study are openly available at https://github.com/roelsmeets/peripheral-networks.

Additional information

Funding

We would like to thank the Dutch Research Council (NWO) . The research for this article was part of the NWO VICI project ‘Redefining the Region: The Transnational Dimensions of Local Colour’ (VI.C.181.026).

Notes on contributors

Anneloek Scholten

Anneloek Scholten (1995) is a PhD candidate at Radboud University. Her dissertation considers the transnational dimensions of Dutch and Flemish regional literature in the long nineteenth century. Particularly, it focuses on the way the genre represents the relationship between the local, national and global, and the circulation of regional fiction across borders. Her research is part of the NWO-funded Vici project ‘Redefining the Region: The Transnational Dimensions of Local Colour.’

Roel Smeets

Roel Smeets (1991) is Assistant Professor of Modern Literature and Digital Culture at Radboud University. In his work, he uses data-driven methods to study literary representations.

Notes

1. Ten Brink, Geschiedenis der Noord-Nederlandse Letteren, 258-259.

2. ‘Het begrip der dorpsvertelling dient nauwkeuriger begrensd te worden.’ Ten Brink, Geschiedenis Der Noord-Nederlandse Letteren, 259.

3. Te Winkel, De ontwikkelingsgang der Nederlandsche letterkunde, 501–506; 328. Te Winkel’s overview of authors per region is comparatively extensive.

4. Knuvelder, Handboek tot de Geschiedenis der Nederlandse Letterkunde, 431, 566, 577–579.

5. Van den Berg and Couttenier, Alles is taal geworden, 399.

6. ‘dorpsnovelle,’ in Algemeen letterkundig lexicon (DBNL 2012). https://www.dbnl.org/tekst/dela012alge01_01/dela012alge01_01_00781.php

7. ‘dorpsroman,’ in Algemeen letterkundig lexicon (DBNL 2012). https://www.dbnl.org/tekst/dela012alge01_01/dela012alge01_01_00782.php

8. ‘streekliteratuur,’ in Algemeen letterkundig lexicon’ (DBNL 2012). https://www.dbnl.org/tekst/dela012alge01_01/dela012alge01_01_03045.php.

9. Discussed at length in De Geest et al., ‘The Case of Regional Literature’. The genre is often accused of being sentimental or simplistic.

10. Bel, Bloed en Rozen, 107, 356 (emphasis added).

11. De Geest et al., ‘The Case of Regional Literature’, 91, 96.

12. Howard, The Center of the World, 46.

13. Regional fiction is understood here simply as a mode of literature focusing on representations of regions that are represented as somehow distinct, for instance by being contrasted with national or metropolitan norms.

14. The importance of journalist reviewers in shaping a ‘literary frame of reference’ and canon is attested to in Rosengren, ‘Literary Criticism’.

15. By ‘dominant narrative’, we mean the narrative that has trickled down into mainstream literary history. Published reviews are also part of a (relatively) dominant discourse on contemporary culture – we do not have access to the perspective of most nineteenth-century readers, and many reviews on regional literature were written from urban centres – but they nevertheless offer a more varied, less uniform perspective on the literary field.

16. ‘Mentions’ are associations made by reviewers, and an indication of their literary frame of reference, indicating those authors that embody various literary traditions. Rosengren relies on ‘mentions technique’ to compare the literary frame of reference of journalistic reviewers to that in essays and academic histories. Rosengren, ‘Literary Criticism’.

17. For instance, Oosterholt usefully describes contemporary understandings of ‘popular’ literature and the distinction between ‘highbrow’ and ‘lowbrow’ as it emerges in the reception of village tales, but he bases his analysis of reviews by only three leading literary critics: Kneppelhout, Potgieter, and Busken-Huet. Oosterholt, ‘Populaire literatuur?’.

18. This is also noted by Honings and Jensen, Romantici en revolutionairen, 321.

19. De Geest et al., ‘The Case of Regional Literature’, 96.

20. Keywords suggesting that an author is read as regionalist include ‘village tale’, ‘village author’, ‘heimat’, or ‘regional literature’, as well as mentions of specific (rural) regions.

21. Authors whose names do not occur in DBNL were presumably rarely reviewed, justifying their exclusion from our search in Delpher.

22. The code of conduct for nineteenth-century periodicals dictated that they had to review all the works they received, and they could not review works that had not been sent to them. Therefore, whether or not a specific magazine reviewed a book does not by itself reveal much about the book’s reception or critical appreciation: this depended entirely on whether the publisher decided to send a copy for review. See Streng, De roman in negentiende eeuw, 320–321. Due to the high number of books sent to many periodicals, reviewers sometimes opted for combined reviews of several books, or resorted to brief announcements that merely noted the existence of a recent publication. Longer reflections on books or literary developments, on the other hand, were often initiated by literary critics or editors. The combination of books in combined reviews does not necessarily imply that books were thought of as similar, and happened primarily because of limited time or space to review large numbers of books received. As such, co-appearances in combined reviews are not considered co-mentions.

23. Scholten, Smeets, Engelen, Lommerde 2021. Published online: https://github.com/roelsmeets/peripheral-networks

24. Literary histories themselves also show shifts in the way they frame regional fiction from the nineteenth century to the present, but to describe these in detail would require a separate study.

25. For an overview of this list, see nodes.csv on our freely accessible data repository.

https://github.com/roelsmeets/peripheral-networks

26. For an overview of this list, see edges.csv on our freely accessible data repository https://github.com/roelsmeets/peripheral-networks

27. ‘De man die zijn volk leerde lezen’. For the origin of this widely-repeated catchphrase, see Schmook, ‘Wie heeft de slagzin “hij leerde zijn volk lezen” bedacht?’.

28. For an in-depth discussion of nineteenth-century understandings of genre in the Netherlands, see Streng, De roman in de negentiende eeuw, 67–121.

29. Modernist cultural production is, in many ways, characterized by networks and collaborations. See for instance Hannah, Networks of Modernism.

30. Only a few names recur repeatedly while many others are mentioned infrequently; and for some authors – like Streuvels or Buysse – literary historians do not always agree on whether they should be classed as regional.

31. Streng, De roman in de negentiende eeuw, 37, 92.

32. Campbell, ‘Realism and Regionalism’. For rural realism as ‘healthy’, see e.g. Belpaire, Het landleven in de letterkunde, 3, 42. For the wider association between rural culture and health, see Sintobin, ‘‘Schamel stuk mens?’, 311–12.

33. Bel, Bloed en rozen, 107, 356. The definition of ‘streekliteratuur’ (regions literature) in Het Algemeen Letterkundig Lexicon also suggests that some works that meet the definition’s criteria are not seen as exclusively regional because its problems and plot ‘exceed the typically regional’ (‘die het typisch streekgebondene overstijgt’), https://www.dbnl.org/tekst/dela012alge01_01/dela012alge01_01_03045.php. International scholarship suggests that this type of devaluation of regionalism as sentimental is a transnational trend. See for instance Dike, ‘Notes on Local Color,’ 81.

34. The connection between Streuvels and the visual arts is the subject of Sintobin et al (eds), Voor altijd onder de ogen, and Cremer’s paintings are discussed in Loos, ‘Jacob Cremer (1827–1880)’, but in literary surveys the importance of visual arts often remains unaddressed.

35. By for instance Ten Brink, Geschiedenis der Noord-Nederlandse Letteren, 259.

36. Heijse, ‘‘t Fingertje naast den duum,’ 366; Van Loon, ‘Het vervloekte geslacht,’ 122.

37. Van Loon’s review for Cohen explicitly mentions this trend in other countries, especially in Germany and Scandinavia.

38. As the introduction demonstrates, women writers are rarely linked to regionalist writing in several important literary histories and in the Algemeen Letterkundig Lexicon.

39. Te Winkel mentions Courtmans-Berchmans as Conscience’s contemporary in De ontwikkelingsgang der Nederlandsche letterkunde (326), and Bel mentions both Gijsen and Courtmans-Berchmans in Bloed en Rozen, but here, Courtmans-Berchmans is mentioned only briefly as an example of a well-known Flemish female writer without reference to her (regionalist) work, thereby implicitly suggesting that gender is seen as the most salient feature of her authorship (137).

40. An obituary in Het nieuws van den dag describes her as one of the most popular writers in Flanders, occupying a primary position in popular libraries alongside Sleeckx and Conscience. Anonymous, ‘Uit Zuid-Nederland’.

41. See the following reviews: Anonymous, ‘Leestafel’; Anonymous, ‘Boeken’; Anonymous, ‘Boeken Enz’; Anonymous, ‘Kinderboeken’; Anonymous, ‘Onze boekentafel’; Anonymous. ‘Kunst en letteren’; T. van den Blink, ‘Mondeling en schriftelijk stellen’; G., [review], N. van Hichtum, ‘Welke boeken’; J.Z.-B., ‘Lectuur’; and S. Rombouts, ‘Het klasse-leesboek.’

42. Streng, De roman in de negentiende eeuw, 427 (table 15). Streng’s data also suggests that novels from the English might have been reviewed more frequently, 454 (table 36).

43. Streng, De roman in de negentiende eeuw, 79.

44. Howard, The Center of the World, 42.

45. ‘Er zijn allerlei grensgevallen in de literatuur aanwijsbaar die weliswaar de genoemde kenmerken bezitten, maar desondanks niet of niet uitsluitend tot de streekliteratuur gerekend worden. Vaak is er dan sprake van een problematiek die het typisch streekgebondene overstijgt.’ Algemeen letterkundig lexicon, s.v. ‘streekliteratuur.’

46. Scholten, Smeets, Engelen, Lommerde 2021. Published online: https://github.com/roelsmeets/peripheral-networks

Appendix A:

List of authors

Cornelis van Koetsveld

Cornelis van Schaick

J.J. Cremer

S. Sr. Coronel

H.E. Beunke

Emile Seipgens

W.

W. van Palmar

Pieter Heering

Tonnis van Duinen

Gerard Keller

H.

H. Hollidee

Hendrik Tillema

Alfons Olterdissen

Idsardi

Josef Cohen

H.

H.J. Maas

Marie Gijsen

Bertha Elisabeth van Osselen-van Delden

Truida Kok

Johanna Courtmans-Berchmans

Margo Scharten-Antink

Hendrik Conscience

Virginie Loveling

Rosalie Loveling

Reimond Stijns

August Snieders

Lode Baekelmans

Pieter Ecrevisse

Gustaaf Segers

Anton Bergmann

Domien Sleeckx

Edward Vermeulen

Cyriel Buysse

Stijn Streuvels

Felix Timmermans

Appendix B:

Statistics

Table

Model Summary
Model	R	R Square	Adjusted R Square	Std. Error of the Estimate
1	,015^a	,000	-,003	5,634
^aPredictors: (Constant), gender

Table

ANOVA^a
Model		Sum of Squares	df	Mean Square	F	Sig.
1	Regression	2,533	1	2,533	,080	,778^b
	Residual	10791,045	340	31,738
	Total	10793,579	341
^aDependent Variable: weighted indegree
^bPredictors: (Constant), gender

Table

Coefficients^a
		Unstandardized Coefficients		Standardized Coefficients
Model		B	Std. Error	Beta	t	Sig.
1	(Constant)	2,997	,325		9,228	,000
	gender	-,265	,938	-,015	-,283	,778
^aDependent Variable: weighted indegree

Table

Model Summary
Model	R	R Square	Adjusted R Square	Std. Error of the Estimate
1	,015^a	,000	-,003	5,634
^aDependent Variable: weighted indegree

Table

ANOVA^a
Model		Sum of Squares	df	Mean Square	F	Sig.
1	Regression	2,533	1	2,533	,080	,778^b
	Residual	10791,045	340	31,738
	Total	10793,579	341

^aDependent Variable: weighted indegree

^bPredictors: (Constant), gender

Table

Coefficients^a
		Unstandardized Coefficients		Standardized Coefficients
Model		B	Std. Error	Beta	t	Sig.
1	(Constant)	2,997	,325		9,228	,000
	gender	-,265	,938	-,015	-,283	,778
^aDependent Variable: weighted indegree

Table

Weighted Indegree Centrality
nationality	N	Mean	Std. Deviation
American	10	1,30	,675
Austrian	3	1,67	,577
Belgian	83	4,72	9,192
Danish	2	3,50	3,536
Dutch	124	2,28	3,334
English	19	3,32	5,945
French	44	3,00	5,140
French / Italian	1	1,00	.
German	26	2,46	2,846
Greek	2	1,00	,000
Hungarian	1	1,00	.
Irish	3	1,33	,577
Italian	5	1,00	,707
Norwegian	3	1,67	1,528
Pomeranian	1	1,00	.
Russian	5	2,00	2,236
Scottish	2	5,00	5,657
Spanish	4	1,25	,500
Swedish	2	3,00	2,828
Swiss	2	2,50	2,121
Total	342	2,96	5,626

Table

Weighted Outdegree Centrality
nationality	N	Mean	Std. Deviation
American	10	,00	,000
Austrian	3	,00	,000
Belgian	83	8,59	25,938
Danish	2	,00	,000
Dutch	124	2,43	9,255
English	19	,00	,000
French	44	,00	,000
French / Italian	1	,00	.
German	26	,00	,000
Greek	2	,00	,000
Hungarian	1	,00	.
Irish	3	,00	,000
Italian	5	,00	,000
Norwegian	3	,00	,000
Pomeranian	1	,00	.
Russian	5	,00	,000
Scottish	2	,00	,000
Spanish	4	,00	,000
Swedish	2	,00	,000
Swiss	2	,00	,000
Total	342	2,96	14,282