Mapping Yeşilçam: A relational approach to the Turkish film industry

Turkey’s Yeşilçam film industry produced more than 5500 films during its 40-year lifetime. The industry had a unique narrative approach, shaped around its economic model, Turkey’s ambivalent connection with modernization and the country’s domestic culture. Yet, particular characteristic qualities of the industry remained rather limited until the last decade, in which vast databases emerged as a consequence of the digital turn. In this study, we develop a relational approach and conduct network analysis, with the aim to better understand the patterns of Yeşilçam’s constitution. Our findings suggest that, rather than a homogenous industry, as often considered by film scholarship, Yeşilçam was divided into two main clusters with significantly different professional, narrative and financial dynamics.


197
In the following part of the paper we reflect on the Yeşilçam industry, tracing its history to provide background information, then elaborate on our methodology and findings. Later, in the conclusion part, we contemplate our findings, putting them in the broader field of cinema history.

Tracking down Yeşilçam in history
Named after a street in Beyoğlu district of Istanbul where the majority of the production companies were located, Yeşilçam had significant international and transnational connections.
However, the majority of the films were produced in Turkey, in Turkish, and for the national audience. The word Yeşilçam does not only refer to the name of this domestic film industry, but also identifies the films it produced, the films' narrative structure, and the related movie-going experiences of the audience (Akbulut 2012;Arslan 2011;Dönmez-Colin 2014;Erdoğan and Göktürk 2001;Kaya Mutlu 2010;Özön 2010;Refiğ 2009;Scognamillo 2003;Yıldırım 2016).
As Akser claims the naming of Yeşilçam "was analogous to the naming of Hollywood" (Akser 2018:155) where the sacred color of "yeşil" (green) stands for "holly" and "çam" (pine tree) stands for "wood." There is no consensus among film historians on the time span covering the Yeşilçam period. Erdoğan and Göktürk (2001:535) define Yeşilçam as a period from the mid-1960s and lasting until the mid-1970s, andSaydam (2020:401) similarly contends that the Yeşilçam existed in the 1960s and in the first half of the 1970s. On the other hand, Arslan (2011) and Yıldırım (2016:24-26) define Yeşilçam as a period of ca. 40 years, from the late 1940s until the end of the 1980s. However, they break this period into sub-periods based on the internal dynamics of the industry. For instance, Arslan (2011) points out three sub-periods, a sequence of the early Yeşilçam period in the late 1940s, the high period in the 1960s and 70s, and the late period in the 1980s.
During a period of 40 years, Yeşilçam industry produced more than 5500 films. 2 That being said, Yeşilçam was not, in fact, a large-scale industry with strong income sources; on the contrary, the industry was only able to reach such production capacity thanks to its fast, highly standardized, and very practical mechanisms of finance and production. Popular idioms coined by the Yeşilçam professionals included "garment film" (TR konfeksiyon film) (Erdoğan 1998a:261;Özgüç 2012:20) and "one minibus, one film" (TR bir dolmuş bir film) (Şoray 2017:99) refer to this mode of production and finance mechanisms of the industry. 3 Yeşilçam's production mode and financial mechanism, which helped sustain the industry's high production capacity (at least in quantity), needs to be elucidated. Firstly, there was a financial connection between the producers and the movie theatre managers (Abisel, 2005;Erkılıç and Ünal, 2018). The whole country was divided into six distribution regions, and the producers' main financial income was the advance payment from the movie theaters in each region. Accordingly, movie theatre managers met the producers every year to commission films to be produced in the upcoming season. Due to their advance payment, managers were entitled to participate in the decision making processes relating to the actors, subject, genre, plot, etc. of any particular film. This system, later named as "manager hegemony" (Abisel 2005:105), constituted the industry's modus operandi. Secondly, a better understanding of Yeşilçam's productivity can be provided by an evaluation of the visual style and sound space of the films. Starting from the mid-1960s, Yeşilçam relied on the strategy of "speeding up production instead of increasing capacity" (Erdoğan 1998b:174). As part of this strategy, dubbing (post-synchronization) became the standard sound process, and filmmakers minimized the number of camera placements, particularly for dialogue scenes (Erdoğan 1998b:174-75). Thirdly, originality appears to be a significant issue in making sense of the industry's production capacity. Only a limited portion of the films were based on original screenplays, which is a consequence of the high-speed production technique. The majority were either remakes and adaptations, or based on plagiarized plot lines. All agents involved, including producers, directors, screenwriters, and theatre managers, favored pre-tested and successful formulas (Erdoğan 2003;Gürata 2006). In addition to these circumstances, one should also note that the industry was working under highly institutionalized and strict censorship mechanisms (Kaya Mutlu 2013;Öztürk and Karadoğan 2020). Multiple stages of censorship approval combined with increasing demand from the 200 audience and movie theatre managers was the driving force behind this fast and practical approach.
Yeşilçam's mode of production was profitable and sustainable, yet it was also vulnerable, shown by the industry's multiple crises throughout its lifetime. One of these crises took place in the second half of the 1970s as a consequence of the contraction of public life, triggered by political unrest in the country, and the increasing popularity of television broadcasts. Yeşilçam turned to making erotic films, targeting male adolescent audience as a strategy to prevent the downsizing of the film market. This strategy resulted in dozens of films, which marked the "sex influx" (TR seks furyası), a period starting in the mid-1970s and ending with the 12 September 1980 coup d'état. 4 It is worth mentioning that not all players of the industry took part in the making of erotics, some preferred to stay out of the trend, becoming less active in their professional careers.
Another crisis occurred by the end of the 1980s. After the short-term impact of the 1980 coup d'état, Yeşilçam revived to a certain extent around the mid-1980s, and maintained its business 202 these films are now available online, and the number is yet to increase. In this study, we greatly benefit from the first of these recent developments, the emergence of open and digital databases specialized in Turkish cinema.
Whenever a new digital database is introduced, initial admiration for the great efforts involved yield to a more routine appreciation, as the users discover the textual resources lying behind that database. In most cases these textual sources are created by researchers or collectors who have devoted their lives to the subject. The same story stands for the digital databases on Turkish cinema, since they rely to a great extent on two fundamental published resources: Agâh Özgüç's If the digital databases on Turkish cinema are created out of previously published materials, then should these online databases be considered simply digital versions of these publications? As pointed out above, Yeşilçam film production was rushed in the heyday of the industry to meet the increasing demand from the audience. Given the context, producers sometimes released two versions of a film with different titles, re-released a previously made film with a new title to resemble a trending film, or made an unplanned film out of unused footage from another work.
Moreover, inconsistencies between the available credits information obtained from different sources are prevalent. 5 In most cases, under these circumstances, it is very difficult to identify whether a film is new and original, or whether the available credits information is accurate. The digital databases developed various verification methods for distinguishing between different resources, for instance, both professional and non-professional users may provide data input to the IMDB, but the new contribution becomes available only after being "reviewed by a member of the data editing team" (IMDb Help 2021).
At this moment, four major databases provide credits information of the films in Turkish cinema.
These databases, along with the number of the films they cover between 1950-1989, are listed below.
• IMDb, 5683 films [Advanced title search, Country: "Turkey", Language: "Turkish", "all titles Each database has different advantages and shortcomings based on their data input and verification methods. For instance, one can observe faulty categorization of entries on Wikipedia, because categorization fully depends on the user providing the input, such as in the case of the film L'immortelle (Robbe-Grillet 1963). This mystery film, shot in Istanbul with a significant orientalist gaze, is mistakenly listed in the "List of Turkish films of the 1960s". 6 Even though set in Istanbul, and featuring Turkish actors in supporting roles, it is highly controversial to define L'immortelle as a Turkish film, since there is no individual or institutional involvement from Turkey in the creative or the financial aspects of the film. Similar faulty data are even more likely to be found on Sinematurk.com, where many foreign films are marked as "domestic". 7 Among these databases, we find the TSA as the most reliable source regarding the accuracy of data covering the Yeşilçam period. The database project of the TSA started to be developed in 2013 and its outcomes were released online in 2014. Data on this database were generated solely by researchers, without user-generated content or crowdsourcing. Their methods utilize higher levels of verification, providing a systematic data creation protocol and cross-checking mechanisms among different sources (Saydam 2021). This not only eliminates faulty data, frequently found in other databases, but also minimizes the number of misleading inputs, as in such cases where two different releases are given for a single film.

Methodology
In this exploratory study, we present an unconventional methodological approach to the broader area of film studies, while contributing particularly to the Yeşilçam literature with our findings from macro-level relational analysis. Here, we applied network analysis in order to examine the interconnections of various players in the Yeşilçam film industry, and to reveal the patterns of their relations. We aimed to identify how the whole film industry presented itself, using a network map based on data from the Center for Turkish Cinema Studies (TSA) database, to which we gained access through a bilateral agreement.
We began with scraping the film credits information from the TSA servers, which came in JSON format, and then converted these data to CSV format and merged them into a spreadsheet. The data covered extensive information on the films, including their metadata, along with additional elements, such as directors, screenwriters, producers, actors, production companies, and advertising agencies. For this research, our scope was limited to identifying the relations between the directors, screenwriters and producers. We ran a pilot study to test the outcome from a variety of maps covering four different sections in the Yeşilçam history. After sorting the data for standardization, we decided to filter the data for the specific period between 1960 and 1979, considered as the heydays of the Yeşilçam industry and defined as the High Yeşilçam Period (Arslan 2011).

206
We employed Gephi for the visualization of the networks. Using the Force Atlas 2 layout algorithm, we generated two undirected bipartite maps, which demonstrate films and players as separate groups of nodes. In the first map, based on modularity, we mainly focused on the relations between the players, and considered films as intermediary entities around which these players revolve. Here, we wanted to illustrate the key players, their positioning, and clusters in the network representing the whole Yeşilçam industry. The second map, which focuses on the professions, shows each one of the professional roles as a separate node; each profession is color-coded to reveal their distribution across the map. In the cases where a single player took on more than one profession, we disjoined them into two or three separate nodes in order to highlight each profession of the same player independently. For example, a player who both directed and produced either the same or different films between 1960 and 1979 acquired two separate nodes on the map. For this reason, we did not include name tags on this map, rather, concentrated specifically on the distribution of professions.
Moreover, we created stacked area and radar charts derived from the statistical data that were the basis for the maps, in order to better interpret the networks. We listed the most prominent players and their professions based on weighted degree, and examined these roles with regards to the 207 industry's dynamics, showing the apportionment of players' roles in directing, producing, and screenwriting.

A bifold industry
The network maps generated using the abovementioned methodology and the distribution of roles in these networks reveal two main findings. The first derives from the positioning of players in the network. As clearly seen in Figure 1, based on the modularity of players, the industry consists of two distinct subnetworks: A type and B type. The map reveals, regardless of profession, all the players of the Yeşilçam industry who undertook at least one directing, producing or screenwriting role in the making of 3834 films between 1960 and 1979. Thus, the map shows the whole scheme of the industry's relations with a particular focus on the clustering of players. 4361 nodes were identified in this network, including the films, connected by 8546 edges. The graph density is low, as expected, indicating a rather loose-knit network; however, the average path length is 5.144, which makes short distance connections possible.  (1960)(1961)(1962)(1963)(1964)(1965)(1966)(1967)(1968)(1969)(1970)(1971)(1972)(1973)(1974)(1975)(1976)(1977)(1978)(1979). Colored nodes represent the players, while grays are the films, and edges indicate relations between players and films. Node size is proportional to weighted degree. Modularity with resolution: 3.086.
Several differences are evident between these two subnetworks. Firstly, A type players constitute a larger portion on the map in terms of both the weight of the nodes and the number of edges between these nodes. The number of films, represented as the nodes in grey, are also significantly higher in this subnetwork. In terms of their weighted degrees, four names are the most prominent in the A type subnetwork: Osman Fahir Seden (306), Safa Önal (283), Bülent Oran (252), and Ülkü Erakalın (223). In the B type subnetwork, however, the intensity of relationships is much lower, and there are no players with a central role similar to these four names. Thus, this subnetwork occupies a more limited and non-dominant area in the overall map.
To better interpret this distinction, we need to examine the prominent players in detail. In the A type, attention is immediately drawn to players taking part in the production of the popular mainstream films. The common point of the players here is that they made relatively more popular films with higher budgets and star actors, and which resulted in commercial success. So much so that these movies are still frequently broadcasted on mainstream television channels, as well as on on-demand platforms. However, we should also point out the great diversity of this cluster, which must not be taken as a homogeneous set.
In the type A subnetwork, the producers of the most popular films are players, such as Türker İnanoğlu, Berker İnanoğlu, Hürrem Erman, and Hulki Saner. Yeşilçam's top screenwriters are Bülent Oran, Safa Önal, Erdoğan Tünaş, and Sadık Şendil, who between them wrote a total of 863 screenplays between 1950-1989. Osman Fahir Seden made dozens of films, as a screenwriter, producer and director. Lütfi Akad and Metin Erksan, on the other hand, made relatively fewer films, but attracted great attention with their distinctive styles in Turkish cinema history (Abisel et al. 2005;Kanbur 2005;Kayalı 2004;Masdar Kara 2017). Halit Refiğ, who has a similar importance, is the founder of the national cinema movement (Refiğ 2009). Holding similar status, Duygu Sağıroğlu is one of the foremost names of the social realist movement. Atıf

210
Yılmaz is an iconic director of both popular, commercially successful dramas, melodramas and comedies, as well as more critical, political films associated with arthouse cinema (Arslan 2007;Özyazıcı 2006). Orhan Aksoy, Orhan Elmas, and Nejat Saydam are the directors identified with the melodrama genre, whereas Natuk Baytan focused more on directing historical adventures. 8 The work of those in the B type subnetwork generally involved relatively lower budgets compared to the A type, and with fewer star players. As a result, their box office income was limited. Their films are characterised by the continual, rapid, and practical line of production identified with the Yeşilçam. Remakes, remixes, and exploitations were common in this set of films (Erdoğan 2003;Gürata 2006). These films mainly targeted a comparatively smaller and more peripheral audience, identified with the broader B movie notion in the cinema literature (Saydam 2019). The dominant tendencies in this subnetwork are seen in the work of Yılmaz Atadeniz, Çetin İnanç, Semih Evin, and Nuri Akıncı. For most of their careers, these filmmakers' productions fitted the B movie concept, among which are fantastic movies, superhero stories, comic book adaptations, adventure and action films with plenty of fighting,, and films about subnetworks, singers. All popular tendencies were adapted rather quickly in this cluster, such as Işıl Toraman and Erdoğan Tilav's work, particularly their adventure and action films. Although it is not easy to make a genre-oriented distinction between the two subnetworks, science fiction, fantasy, western and gangster movies are clearly the more dominant in the type B subnetwork.
Speaking of genre, we should also reflect on the erotic films. As we mentioned above, Yeşilçam turned to making erotic films in the second half of the 1970s as a response to the downsizing of the market, a trend referred to as the sex influx. During the sex influx, some players were more actively engaged with the trend than others in their professional careers. For instance, directors, such as Ülkü Erakalın, Aram Gülyüz, Yücel Uçanoğlu, Naki Yurter, Yavuz Figenli, and Oksal Pekmezoğlu were particularly devoted to these films. B type players more readily adapted to the sex influx than the A type players, so that their overall weight in the industry increased during this period. However, a closer look at Figure 1 reveals that erotic films do not correspond to an absolute distinction between A and B types.
Some of the leading players of the sex influx names, such as Yücel Uçanoğlu, Yavuz Figenli, Çetin İnanç, and Erdoğan Tilav appear in the B type, and others, such as Aram Gülyüz and Ülkü Erakalın, in the A type. Yet, some players in the B type, such as Semih Evin, took no part in the sex influx, quitting his career at the second half of the 1970s. That is to say, the sex influx appears to overlap at the intersection of the A and B types only because its repercussion was limited to the second half of the 1970s.
Our second finding relates to the distribution of professions. As clearly seen in Figure 2, the three major professions, namely directing, producing, and screenwriting, have a certain prevalence in the network. However, a clustering similar to that in Figure 1 is also present, yet again A type and B type films are positioned around the same sections in both maps. The nodes representing screenwriters are predominant, mainly on the left section of the map, where the A type films reside. Two in particular, with the highest degree centrality in the entire map, form the center of the A type subnetwork. These two nodes, Safa Önal and Bülent Oran, are also evident in Figure 1, however, this time we notice the line of work in their contribution to the Yeşilçam industry. Again, in the same subnetwork, the total weight of producers is noticeable. Most of these producers are also directors, examples of the case in which a single player has more than one role. Overall, these two professions are collocated, i.e., side-by-side when it comes to the realization of filmmaking. In the B type subnetwork, there are no such central or large nodes, and directing appears more likely to be the leading role in this subnetwork rather than screenwriting or producing.  (1960)(1961)(1962)(1963)(1964)(1965)(1966)(1967)(1968)(1969)(1970)(1971)(1972)(1973)(1974)(1975)(1976)(1977)(1978)(1979). Colored nodes represent the professions, while grays are the films, and edges indicate relations between professions and films. Node size is proportional to weighted degree.
In the A type, four players are solely involved in screenwriting, with the exception of Safa Önal, who has also directed 19 films. Five actors with multiple roles in the A type still have a primary profession, three directing, and two producing. Among the players with multiple roles, the distribution of work is distinct, except for Osman Fahir Seden, who evenly contributed to the three professions. Almost all B type players, however, have multiple professions and are mainly directors. Of the ten, nine are directors and one is a producer, namely Işık Toraman. Another contrast between the two types is observed in the distribution of secondary roles. Out of the six players with multiple roles in the A type, three are directors, two are screenwriters, and one is involved in production as the second role. In the B type, for nine players, the secondary role is screenwriting, and for one only, producing.
An overall view of professions in the Yeşilçam industry also supports our second finding ( Figure   5). The quantitatively larger and more productive impression of the A type, which manifests itself in Figure 1 and Figure 2, reflects as well on the distribution of roles in both subnetworks. A total of 2182 professional activities were undertaken in the A type compared to 1309 in the B type. Yet, there is a remarkable difference between the two. Among all professions performed in the A type, screenwriting has the largest share with 54%, followed by directing, with 29%. In the B type, the order is reversed : 55% directing and 29% screenwriting. The overall distribution of professions very much coincides with the notion of high-speed production, highlighted by the higher shares of directing and screenwriting compared to producing.

Conclusion
The tendency to reduce the history of cinema to the history of films and major players (particularly directors and stars) has been abandoned in favour of a focus on the cultural and social dimensions of cinema in the last two decades, in line with the new cinema histories approach. It has become critical to analyze the dynamics of the film industries in relation to narration, audience, and viewing experiences. In this paper, we revisit the practice of film historiography using a novel approach, which considers the key players, not as decision makers and great creators who determine the fate of cinema, but as subjects enabling us to understand the internal dynamics of the industry through the linkages they created. Mapping these linkages among the cinema professionals contributes to our understanding of how Yeşilçam was able to survive for 40 years, despite inefficiencies and crises.
It is a fact that, as well as mainstream films, Yeşilçam produced B type films, for different audiences. However, our findings suggest a clear distinction between the players involved in the two types of films. This distinction, as clearly manifested in our study, reveals the differences between their modes of production, financial conditions, and narrative characteristics. Also evident in our findings is an unexpected outcome from these two clusters, in terms of the distribution and undertaking of professional roles. In the A type, screenwriters are more central, and professional careers are more identified with particular roles, while in the B type, directing is the most common career. In contrast, in the B type, multiple roles among the players were more common. This suggests that filmmakers in the B type had higher levels of adaptability in terms of finance, narration and professional competencies. Interestingly, these filmmakers were excluded from the historiography, and their films were considered immature and low quality. It seems however, that the professional flexibility of the B type filmmakers was a critical factor in the survival of Yeşilçam. Moreover, it is not surprising to see that multiple roles were common in the Yeşilçam industry, considering similar engagements in other film industries, including Hollywood. The multiple roles in Yeşilçam, indicate that the industry's professional and business structures resemble those of prevailing film industries.
From a methodological perspective, network analysis was applied to a major film industry with the aim to reveal the connections among players, and understand how these connections reflected on the dynamics of the industry. Employing network analysis is not very common in film studies, particularly at an industry scale. Yet, it can be useful in drawing the boundaries of an industry, defining its players and their roles, showing its evolution, and identifying anomalies, which overall, allow a further assessment of industry's internal dynamics. The limitations of this paper include the restricted period, confined to the most productive years of Yeşilçam. This was because the inclusion of the pre-and post-periods would risk the overgeneralization of our findings. Still, we believe this study will act as a guide for further research, not only in the Turkish cinema industry, but in other cinema industries as well. Future studies in this line may adopt a longitudinal approach to explore the evolution of an industry and evaluate its disposition in a sociopolitical setting, focus on a certain player or a particular cluster for an analysis of micro structural factors, include the actors and actresses to reveal their long-term commitment to the industry, or perhaps, extend the research to include production infrastructures, with a focus on the capacity of the sector and the role of production companies. In point of fact, applying network analysis to multiple film industries would facilitate comparative approaches, and potentially reveal the common dynamics behind the emergence of similar patterns among different industries, while providing a better understanding of the historical connections between them.