The grand opening? The transformation of the content of culture sections in European newspapers, 1960–2010

1 The Grand Opening? The Transformation of the Content of Culture Sections in European Newspapers, 1960–2010 Semi Purhonena†, Riie Heikkiläa and Irmak Karademir Hazirb a University of Tampere, Finland. b Oxford Brookes University, UK. † Corresponding author. Abstract This study presents an analysis of changes in the coverage of culture in major European newspapers from 1960 to 2010. Employing a content analysis of the newspapers, we examine how culture sections have changed in terms of the cultural areas covered. We ask whether the content of cultural coverage has become more heterogeneous and whether newspapers embedded in different geographical and cultural contexts differ in this respect. To assess heterogeneity, we first focus on the opening of culture through the increase in coverage of emerging art forms at the expense of the coverage of established art forms. Next, we turn our attention to music and measure openness according to the increase of pop-rock content. The results suggest a cultural opening, albeit mostly measured by the second proxy. Established art forms hold their mainstream position, while emerging forms are validated at a slower pace than expected. Regarding music, the transformation of coverage from classical music to pop-rock is very dramatic. The findings challenge expectations about the order in which the newspapers manifest the timing and thoroughness of the opening of culture, highlighting the complexity of the factors shaping newspapers’ cultural coverage. Keywords Cultural legitimacy, classifications, cultural change, newspapers, popular culture, music This is an accepted manuscript. The original article has been published in Poetics 2017, 62, pp.29-42.DOI: https://doi.org/10.1016/j.poetic.2017.02.001. 2 The Grand Opening? The Transformation of the Content of Culture Sections in European Newspapers, 1960–2010 1. Introduction How have the classifications and hierarchies of arts and culture changed over the last decades? More precisely, has the “content” of legitimate culture become more heterogeneous and open; that is, have new forms of (popular) culture been included in the conception of legitimate culture in addition to more established highbrow arts? If so, what are the geographic and cultural limits of this post-1960s cultural transformation towards increased openness and heterogeneity? This paper provides a bird’s-eye view of long-term changes in cultural hierarchies and the content of legitimate culture through the lens of a specific institution in the field of cultural production: culture sections in quality European newspapers. We employ a quantitative content analysis of cultural coverage in six major European newspapers from 1960 to 2010. The analysis of a large sample of newspaper articles on culture allows us to assess the extent to which the heterogeneity of cultural coverage has increased over the last fifty years, if increased heterogeneity signifies an emergence of new cultural areas and art forms that have become recognized with the status of “art” or “culture,” or, rather if heterogeneity has increased through the transformation of content inside the boundaries of established arts, and, finally, whether newspapers published in different parts of Europe are similar regarding these trends. This paper opens with a look at different perspectives through which the heterogenization of the content of legitimate culture can been conceptualized, as well as at previous work analyzing newspapers as a means to understand cultural change. After a section clarifying our hypotheses, the comparative setting of the analysis, and the research design, the results section will firstly examine the possible increase in the coverage of emerging art forms at the expense of the established arts. The second analysis will focus on articles on music and ask to what extent there has been a shift in the balance between popular music and classical music over the fifty-year period. The paper concludes by reflecting on the results and methodological aspects of the study. 1.1 Heterogenization and the opening up of legitimate culture: Theoretical perspectives Cultural hierarchies and classifications vary historically and cross-nationally (DiMaggio 1982, 1987; Levine 1988; Peterson 1997; Daloz 2010; Lamont 2012). Bourdieu famously argues that 3 cultural and social hierarchies are homological (e.g., that a taste for “high” arts is highly class dependent), resulting in a situation in which prescriptions of legitimate culture contribute to the reproduction of inequality (Bourdieu 1984). However, a wide range of studies casts into doubt the role of traditional highbrow culture as a status marker and suggests that legitimate culture itself is becoming more inclusive, combining items from both highbrow and popular milieus (e.g., Peterson 1992; Peterson & Kern 1996; DiMaggio & Mukhtar 2004; Van Eijck & Knulst 2005; Prieur & Savage 2011). Most famously, this trend of “opening up” and increasing the heterogenization of legitimate culture’s content is conceptualized as the rise of the cultural omnivore (Peterson 1992, 2005; Peterson & Kern 1996; for a review, see Karademir Hazir & Warde 2015). The omnivore scholarship shows that high status groups tend to appreciate broad repertoires of taste, whereas low status groups’ tastes are more narrow and univorous. While this notion does not necessary contradict Bourdieu’s model (if omnivorousness is taken as a new form of cultural capital yielding various benefits; Lizardo & Skiles 2012), it nevertheless implies that the concept of legitimate culture has widened and become more heterogeneous. Omnivorousness, however, is only one way to approach the post-1960s cultural transformation. There are plenty of more or less intertwined perspectives and research strands that conceptualize the “grand opening” and its key elements: the increased heterogenization of legitimate culture and the two sides of that process, the rise of popular culture and the decline of traditional highbrow culture’s status. Studies analyzing cultural participation longitudinally ask whether audiences of highbrow arts are shrinking and aging, thus causing the “meltdown” of highbrow arts’ status as a means for distinction (DiMaggio & Mukhtar 2004). A more recent discussion on the “emerging forms” of cultural capital argues that distinctions and hierarchies are increasingly developing within the realm of popular culture (Prieur & Savage 2011, 2013; Friedman et al. 2015; Savage et al. 2015). As a consequence, clear hierarchies structure the fields of popular culture (Varriale 2016). An intersecting line of research deals with the legitimation of popular culture via intellectualization and aestheticization (Regev 1994; Baumann 2001; Schmutz 2005, 2016), highlighting how new forms of culture can be transformed “from entertainment to art” (Baumann 2007) by extending the high-art criteria of evaluative practices (Van Venroiij & Schmutz 2010) or by other processes of “artification” (Heinich & Shapiro 2012). Then again, studies and debates on the commercialization of culture probe whether changes in the balance of established arts and popular culture result from the competition between commercial legitimation and aesthetic legitimation, in which commercial 4 values might take over artistic values and the autonomous status of the arts (Bourdieu 1993; Hesmondhalgh 2007; Scardaville 2009; Verboord 2011; Schmutz 2016). One related stream of research focuses on the role of cultural globalization and aesthetic cosmopolitanism (Tomlinson 1999; Regev 2013; Varriale 2015, 2016, forthcoming). Classical highbrow arts—the core content of traditional legitimate culture—have represented essentially a European cultural canon, which has been challenged by global flows and influences, especially the emergence of popular world culture. The most notable example of this is the “poprockization” of music throughout the world, namely, the growing influence of American (and British) pop-rock music and its local adaptations, which is conceptualized as “expressive isomorphism” standardizing the national peculiarities in world culture (Regev 2013; Varriale 2016). Finally, technological changes, especially digitalization, may play a role in the increase of cultural heterogeneity, as the proliferation of distributable and portable cultural forms and the ease of consuming them are clearly factors behind the present “cultural abundance” (Wright 2011, 2015). Without doubt, these myriad aspects of the heterogenization of legitimate culture could be considered parts of a more general sociological narrative about changes in the Western world over the latter part of the twentieth century, in which the key has been a shift from traditional and unquestionable authority to more flexible and plural authorities (be they aesthetic authorities or other types of rules of conduct). This shift can be called, for instance, “informalization” (Wouters 2007) or “expressive revolution” (Martin 1981). In any case, all the outlined processes contribute to cultural change which DiMaggio (1987:452) calls a trend towards “more differentiated and less hierarchical” and “weaker and less universal” cultural classifications. 1.2 Newspapers as tastemakers The media, as an institution, plays a key role in how different cultural classifications and hierarchies appear, spread, become legitimized, and are debated (Janssen 1999). For this reason, newspapers, with their predefined culture or arts sections, provide useful data (Janssen et al. 2008, 2011; Jaakkola 2015). Moreover, sociologists often encounter a lack of suitable data sets enabling the tracking of long-term trends over several decades and across a variety of national contexts (Peterson 2005; Reeves 2015), and using newspaper data is a means to solve this problem. Specialized sections on arts and culture were established in many major European newspapers around the middle of the twentieth century (Jaakkola 2015). These form a structural frame that 5 enables and moulds taste dispositions by producing symbolic value and hierarchies (Bourdieu 1993; Janssen 1999; Jaakkola 2015). Scholars consider quality newspapers especially influential in shaping public opinion on the value and legitimacy of art forms and cultural objects (Verboord et al. 2015).1 Cultural journalists and critics are cultural intermediaries—both gatekeepers who select what will be showcased and tastemakers who define what counts as good taste (Bourdieu 1984; Smith, Maguire, & Matthews 2014; Janssen & Verboord 2015). Thus, changes in cultural legitimacy should be evident in their writings, or, more generally, in the cultural classifications and hierarchies represented in the media (Janssen 1999; Jaakkola 2015; Verboord et al. 2015). Therefore, the relative amount of newspaper space dedicated to particular art forms and genres is “indicative of their cultural status at a given point in time” (Janssen 1999:330). Given their potential as a longitudinal and comparative data source for research on the supposed opening of culture during the last decades, newspaper cultural pages remain surprisingly understudied. A major exception, however, is a recent large-scale Dutch research project (e.g. Janssen et al. 2008, 2011; Verboord et al. 2015). This project’s various outputs shed light on how the evaluations of literature (Berkers et al. 2014); film (Kersten & Verboord 2014); music (Schmutz 2009; Schmutz et al. 2010; Van Venrooij & Schmutz 2010); and fashion products (Janssen 2006) have changed in different national contexts. In addition to understanding the dynamics of these distinct fields, the project uses newspaper content to reveal how broader processes of social and cultural change—such as globalization—influence cultural appreciation and hierarchies (Janssen et al. 2008). Newspapers are a fertile medium for exploring such issues, since depictions of cultural products in different national contexts are largely consistent with country-specific structural dynamics. For instance, the relatively less open cultural classification system identified in the analysis of German newspapers seems to be in accordance with the German educational system (conducive to cultural hierarchy), Germany’s journalistic field (state-dominated media system), and Germany’s cultural policy system (decentralized) (Janssen et al. 2011). Moreover, newspapers are a crucial data source because there seems to be a relationship between the form of journalistic attention devoted to particular art forms and genres and the location of those genres within the wider cultural hierarchy. For instance, the increase in popular music reviews—especially rock music—since the 1960s in the US press granted those genres legitimacy earlier than in other national contexts (Schmutz et al. 2010). By broadening geographical coverage, our research contributes to existing knowledge on changes in the valuation and legitimation of culture through the visibility in quality newspapers and thus 6 provides one answer to the call to “uncover further national or longitudinal differences” (Janssen et al. 2011:161). The inclusion of newspapers from both Northern and Southern Europe—including one newspaper from Turkey, located literally on the border of Europe—allows us to question the limits and variations in cultural declassification observed by Janssen et al. (2011) and Schutz et al. (2010) as a general (albeit varying) trend in Central Europe and the United States. Enlarging the sample to include national contexts outside the core of Europe in many regards (i.e. the level of integration with globalization; history of democratization; and degree of interconnectedness in the fields of politics, art, and journalism) accentuates cross-cultural differences and allows us to test the cultural opening thesis more thoroughly. Moreover, our research differs from earlier studies (Schmutz et al. 2010; Janssen et al. 2011) in the sense that it covers only newspapers’ cultural sections rather than their entire content. By concentrating on explicit culture sections, we can let the newspapers decide on our behalf what counts as culture and what is valuable enough to be included under the culture section. 1.3 Aims, hypotheses and the comparative setting We investigate if cultural content in quality European newspapers has become more open and heterogeneous, as implied by the abovementioned various debates in cultural sociology, from 1960 to 2010 in terms of the cultural areas and art forms (topics) discussed in articles found in cultural sections. These newspapers come from six European countries, ranging from two Nordic countries (Helsingin Sanomat, hereafter HS, from Finland, and Dagens Nyheter, DN, from Sweden) to two large Western European countries (Le Monde, LM, from France, and The Guardian, GU, from the UK), and, finally, two Mediterranean countries (ABC/El País, ABC/EP, from Spain, and Milliyet, MIL, from Turkey). Our analysis is guided by the hypothesis (H1) that, between 1960 and 2010, the conception of culture has opened in terms of cultural areas and genres discussed in newspaper culture sections, revealing that the dominance of classical established arts has dwindled. We should be able to track this trend in two ways: (H1a) through the increased heterogeneity of cultural content caused by the appearance of emerging art forms that have gained space at the expense of traditional and established art forms (which, in other words, should be manifested as an upward trend in the proportion of articles discussing emerging cultural areas); and, (H1b) through the shift in the volume of the articles on classical and popular music (seen as an upward trend in the proportion of articles on pop-rock music). Besides focusing on emerging cultural areas, we concentrate on music in particular because it enables us to measure the relative amount of space dedicated to traditional 7 “high” culture and more recent popular cultural forms within perhaps the most established and institutionalized art form in modern Western societies (Frith 1996; Peterson & Kern 1996; Roy & Dowd 2010). Our second hypothesis concerns differences between newspapers embedded in different European contexts characterized by specific cultural, historical, and socio-political features, in addition to differences in the newspapers’ institutional and field-level (organizational) characteristics. This hypothesis (H2) states that there is significant variation in the thoroughness and timing of the opening of cultural content among the newspapers studied (including both cases of H1a and H1b). We do not claim that the newspapers selected are nationally representative cases and thus directly express national differences. Nevertheless, the newspapers are linked to specific European contexts with their own sociohistorical and structural characteristics, which the newspapers’ content reflects. There are also several other influential factors to consider, and we expect the differences between the newspapers to signify at least three society-level dimensions (cf. DiMaggio 1987; Van Venrooij & Schmutz 2010; Janssen et al. 2011). First, we must pay attention to the different locations in the global—and Western-dominated—cultural system, our expectation being that the opening happened earlier and more thoroughly in central countries (Appadurai 1996; Janssen et al. 2008). According to this dimension, the UK and France clearly represent the centre, whereas Spain, Nordic countries (among which Sweden is clearly closer to the global cultural centres than Finland), and especially Turkey are examples of more peripheral locations in the global cultural system. Second, we assume that the different media system models play a significant role, our expectation being that the opening happened earlier and more thoroughly in “liberal” model countries (represented in our study of the UK) rather than “democratic corporatist” (Finland, Sweden) or “polarized pluralist” (France, Spain, Turkey) models (Hallin & Mancini 2004). Third, we should remember that different cultural policy regimes have a strong influence on the dynamics of cultural fields and hierarchies (Looseley 2011; Katz-Gerro 2015). Our expectation is that the opening of culture happened later in countries where established and traditional art forms were heavily state subsidized (Janssen et al. 2011). This dimension suggests that France and Nordic countries, where state funding for the established arts is high (ACE 1998), might have adopted new cultural forms later than the other countries examined (UK, Spain, and Turkey). Taken together, and to specify our second hypothesis, these three society-level dimensions suggest that the newspapers can be ordered, according to how their national contexts have facilitated the 8 opening of culture, in a way that (H2a) the British GU represents the earliest opening and the Turkish MIL represents the last. Among the other newspapers, we expect (H2b) the Finnish HS to be the second latest. Overall, our study echoes Fishman and Lizardo’s (2013:215) questioning of the assumption that the mechanisms and processes behind cultural change “operate in a largely casefree manner.” Thus, we argue that differences in the geographical and cultural locations provide the background for a fruitful comparative analysis of the transformation of the newspapers’ cultural content and, therefore, of the changes in cultural classifications and hierarchies over the last five decades. 2. Research design 2.1 Data Data from the culture sections of ABC/EP, DN, GU, HS, LM, and MIL from 1960 to 2010 were collected between 2013 and 2014. The sample was collected at ten-year intervals, including volumes from six time points (1960, 1970, 1980, 1990, 2000, and 2010). To eliminate seasonal variation in arts coverage, the method of “constructed weeks” was used (Riffe et al. 1993; Janssen et al. 2008). Each research year was divided into thirds from which one full week was randomly constructed, totalling three weeks per selected year. The sample includes 18–21 editions of each newspaper per year (depending on whether the newspaper was published six or seven days a week), which amounts to 711 editions altogether. The data were coded, following a predefined coding system, into a statistical matrix.2 The unit of our content analysis is an individual article (N=13,161). The data only include articles with journalistic content published in culture sections—both the core cultural pages inside the newspaper as well as thematic cultural supplements. All the newspapers analyzed are nationally leading quality papers with comparable levels of circulation, albeit in different geographical and cultural regions in Europe (cf. Janssen et al. 2008; Jaakkola 2015; Verboord et al. 2015). HS, founded in 1889, has a moderate social-democratic tendency and is Finland’s most widely read newspaper, with almost no national competition (Jaakkola 2015). HS set the model for choosing the other national cases. GU is similar to HS in both its moderate centre-left voice and circulation levels (Taylor 1993). DN is among the biggest newspapers in Sweden and shares HS’s political leanings (Hadenius 2002). LM, the most widely read newspaper of France, shares the previously mentioned newspapers’ moderate leftist political 9 tendencies (Eveno 2004). In Spain, EP, founded just after the end of Franco’s right-wing dictatorship in 1976, is a logical counterpart to the other selected newspapers. However, for the period of 1960 to 1970, with no comparable alternative available, we had to use the monarchist, right-wing ABC, the most influential Spanish newspaper at the time (Olmos 2002). For Turkey, we chose MIL because it is the most widely circulated newspaper, published without interruptions between 1960 and 2010. Moreover, MIL is a politically centred paper that falls in line with the rest of the sample (Topuz 2003). The online supplement (Table A.1) presents more information on the newspapers. Table 1 shows the distribution of the data across years and among newspapers. The way our sample is distributed by year for each newspaper reveals how the culture sections have been structured and changed in terms of volume (as reflected in the number of articles). Overall, the data are quite evenly distributed both in terms of years and newspapers, albeit with some exceptions. The total number of articles generally increased towards the 2000s. In the sense of pure volume, the newspapers fall into two categories: larger (ABC/EP, DN, HS) and smaller (GU, LM, MIL). Moreover, the Nordic papers, HS and DN, differ from both the Central European and Southern European cases in the amount of journalistic attention given to culture in 1960. In the case of MIL, there was quite a dramatic increase in 2010, whilst in the case of EP there was a clear decrease in 2010 compared to 2000 or 1990. Table 1. Breakdown of the data by year and newspaper (percent of total and number of articles). ABC/EP DN GU HS LM MIL Total 1960 1.4 (182) 2.7 (355) 2.0 (258) 3.0 (399) 1.4 (186) 0.7 (87) 11.1 (1,467) 1970 3.6 (475) 3.6 (478) 1.1 (145) 3.0 (392) 3.0 (397) 1.0 (128) 15.3 (2,015) 1980 2.4 (311) 2.3 (309) 1.1 (144) 2.5 (334) 2.3 (306) 1.0 (134) 11.7 (1,538) 1990 4.1 (540) 3.3 (430) 3.1 (403) 3.6 (479) 1.8 (237) 1.0 (130) 16.9 (2,219) 2000 7.3 (955) 3.3 (433) 2.6 (348) 4.3 (572) 3.7 (490) 2.0 (260) 23.2 (3,058) 2010 2.3 (305) 3.9 (517) 3.4 (441) 5.0 (660) 2.2 (294) 4.9 (647) 21.8 (2,864) Total 21.0 (2,768) 19.2 (2,522) 13.2 (1,739) 21.5 (2,836) 14.5 (1,910) 10.5 (1,386) 100.0 (13,161) 16 3.3 Music: The rise of pop-rock To test the hypothesis of openness according to the increase in popular genres, we focus on the music field and investigate how music coverage is divided according to different musical genres and how it has changed over time (Table 4). Two genres, classical music and pop-rock, differ from the rest of the sample regarding their size. Almost 44% of all articles on music cover classical music, whereas one-third focuses on pop-rock. Whilst their proportions are not entirely different from each other, they have opposite trajectories: pop-rock coverage has increased dramatically, whereas classical music coverage has sharply decreased. Jazz, with its 9% coverage, and other smaller genres—such as domestic hit music, traditional folk music, and world music—have remained relatively stable over time, apart from the fact that world music was fairly unrepresented until 1990 and that jazz saw a clear rise from 1960 to its height in 1980–1990. Altogether, pop-rock accounted for one-third of all music articles, and there is a huge and steady increase in its proportion, as shown in Table 4. But how has this proportion evolved between 1960 and 2010 in each of the six newspapers? In the case of the proportion of articles on emerging cultural areas, are there—again—significant differences in the victorious trajectory of pop-rock? Table 5 shows that not only decades but also newspapers have significant and powerful associations with the prevalence of articles on pop-rock. Table 4 evidences the unadjusted association of time points, but there are also significant differences between the newspapers, with GU, HS, and LM, in particular, having published fewer articles on pop-rock than ABC/EP (or DN) and MIL clearly having published more. Moreover, there is a significant interaction between decade and individual newspapers. In contrast to the powerful interaction when predicting articles on emerging cultural areas in Table 3, the strength of the interaction here is much more modest. As shown by model 2, Table 4. Articles on different musical genres (of all articles on music) by the year (percentages). 1960 1970 1980 1990 2000 2010 Total Pop-rock 1.6 14.9 26.6 26.7 41.7 55.5 33.4 Classical music 83.9 61.5 45.8 50.9 33.4 22.8 43.6 Jazz 2.7 7.5 12.2 11.7 9.1 7.1 8.7 Domestic hit m 4.7 4.2 4.3 2.0 2.8 4.3 3.6 Traditional folk 2.4 4.7 3.5 2.7 2.6 1.9 2.9 World music 0.4 0.5 1.0 2.9 7.1 4.4 3.4 Other 4.3 6.8 6.6 2.9 3.4 4.0 4.4 Total (N) 100.0 (255) 100.0 (429) 100.0 (395) 100.0 (546) 100.0 (650) 100.0 (773) 100.0 (3,048) 17 only MIL shows a statistically significant deviance from ABC/EP, this time suggesting that the increase in the proportion of pop-rock articles over each ten-year interval is less pronounced in MIL than in ABC/EP (the reference category). In fact, if the analysis is re-run excluding MIL from the data, the interaction between decade and newspaper is statistically insignificant. Table 5. Articles on pop-rock music (of all articles on music) by year, newspaper and other independent variables (odds ratios and Wald statistics from logistic regression analysis). Unadjusted Model 1 Model 2 Model 3 Decade (Wald) 334.72*** 309.55*** 48.26*** 21.98*** 1960=0 1.71*** 1.72*** 1.77*** 1.51*** Newspaper (Wald) 209.44*** 182.30*** 70.04*** 64.29*** ABC/EP (ref.) 1.00 1.00 1.00 1.00 DN 1.06 1.39* 1.83 1.52 GU 0.74* 0.66** 0.57 0.65 HS 0.43*** 0.42*** 0.19*** 0.21** LM 0.29*** 0.32*** 0.51 0.33* MIL 2.63*** 2.08*** 5.49*** 5.64** Decade x Newspaper (Wald) 20.27** 22.65*** Decade x ABC/EP (ref.) 1.00 1.00 Decade x DN 0.91 1.10 Decade x GU 1.04 1.01 Decade x HS 1.22 1.26* Decade x LM 0.87 1.09 Decade x MIL 0.77* 0.74* Number of cultural pages (Wald) 178.06*** 0.94 1–2 (ref.) 1.00 1.00 3–5 1.78*** 0.90 6–10 3.29*** 0.85 11 or more 4.34*** 0.82 Supplement (Wald) 276.41*** 48.29*** Not supplement (ref.) 1.00 1.00 Supplement 4.33*** 2.87*** Size of the article (Wald) 21.16*** 1.06 Smaller than a quarter page (ref.) 1.00 1.00 At least a quarter but less than full page 1.11 1.00 Full page or more 2.19*** 1.22 Type of the article (Wald) 0.00 0.70 Other (ref.) 1.00 1.00 Review 1.00 0.92 Constant 0.11*** 0.10*** 0.13*** χ² change of the model *** ** *** Nagelkerke pseudo R² change 0.25 0.01 0.03 N=3,044. * p < 0.05; ** p < 0.01; *** p < 0.001. 18 When predicting the prevalence of articles on pop-rock music, the control variables measuring different elements of cultural content play a minor role. The only significant factor when controlling for other associations is whether the article appears in a supplement or in the core culture section; articles on pop-rock appear much more frequently in supplements than other types of music. Unadjusted associations also show that articles on pop-rock appear more frequently in the longest culture sections, as well as in the form of the longest articles. All these features are typical to the 2000s when there were longer culture sections, longer articles and more thematic cultural supplements—thus, with the exception of supplement, these associations disappear when controlling for the decade (model 3). Figure 1. The proportion of articles on pop-rock (full line) versus the proportion of articles on classical music (dashed line) of all articles on music by year, separately on each six newspapers (percentages). 19 Figure 1 specifies our interpretation of the interaction between time points and newspapers by contrasting the rise of pop-rock with the decline of the proportion classical music articles. This allows us to inspect the moments when pop-rock articles outnumbered classical music articles in each newspaper. The increase in the prevalence of pop-rock articles has occurred in different ways across the newspapers. This is expected, since previous research also shows that the “size and timing of the shifts towards greater attention to popular music” (Schmutz 2009:305) has varied among countries, even if the main trend is clear. However, the differences are not entirely similar to our expectations. MIL seems to show the earliest opening, yet, we should note that Western classical music has never had a similar status in Turkey as in other European countries. Thus, we cannot expect classical music to have attracted more journalistic attention in earlier decades. The high and volatile proportion of pop-rock articles is due to the fact that the main category against which pop-rock is analyzed has been so marginal in Turkey. Only 16% of all articles on music in MIL are on classical music, whilst the proportion in all the data is 44%. As for other newspapers, DN and ABC/EP interestingly turned their attention to pop-rock music earlier than GU, HS, and LM. In the latter group of newspapers, pop-rock does not begin to outnumber classical music until the 2000s. 4. Discussion In this paper, we engaged with literature that suggests a change in the ways culture and arts have been classified and valued in recent decades (DiMaggio 1987; Peterson & Kern 1996; Baumann 2007; Janssen et al. 2011). To empirically explore the heterogenization and opening of the content of legitimate culture, and to track trends in a longitudinal manner, we utilized newspaper data. We consider newspaper culture sections as a valuable medium for achieving this goal, especially when set against our focus on quality newspapers, as much of the literature is particularly concerned with high status groups’ tastes and practices (cf. Janssen et al. 2011:159). Moreover, for decades print media has been a crucial institution of cultural mediation, capable of attributing symbolic value by framing certain cultural forms as “culture” and “arts” when including them systematically in their culture sections and supplements. Regarding our first hypothesis (H1a), we find the increase in the coverage of emergent cultural areas to be very moderate. Articles on established cultural forms (literature, music, etc.) are still the backbone of the arts and culture content of newspapers, despite the proliferation of cultural forms in 20 recent years. However, film and television ranked among the major cultural areas, so they seem to have managed to cross over validation boundaries (cf. Baumann 2007; Lavie & Dhoest 2015). In contrast, the latter part of our first hypothesis (H1b) is clearly supported, since attention to “poprock” music (Regev 2013) has almost exponentially increased, whereas the coverage of classical music is in sharp decline. This finding echoes Schmutz et al.’s (2010) analysis of the United States, France, Germany and the Netherlands (see also Schmutz 2009). When evaluated with this study, it is easier to suggest that the opening of culture via the diversification of legitimate musical genres resonates on a wide geographical level across Europe. To recap, the most strongly established art forms—like music—are far from being in decline in cultural sections, yet there are major shifts in the balance between traditional highbrow culture and more popular forms within the wellestablished arts. When interpreting these findings, we must remember the nature and context of our research object: European quality newspapers and their culture sections. These are a kind of “home court” of classical highbrow arts through which many people follow the arts even when they lack interest in them otherwise (cf. Jaakkola 2015; Janssen et al. 2008, 2011). As such, it is perhaps not so surprising that established art forms still dominate, yet music coverage’s transformation can be considered very dramatic. Many of the various theoretical perspectives on cultural heterogenization discussed in the introduction offer potential interpretations and conceptual tools for further investigations of this situation. The key question, in line with debates on emerging forms of cultural capital and the legitimation processes of popular culture, would ask, under the circumstances of increased heterogeneity and, in music, even the dominance of pop-rock, what the major hierarchies are and how they are articulated. Is it a matter of extending the high-art criteria of evaluation to the sphere of popular culture, or are there other bases for making distinctions (cf. Van Venroiij & Schmutz 2010; Prieur & Savage 2013)? Furthermore, from the point of view of aesthetic cosmopolitanism, what are the countries of origin of the pop-rock music that has appeared on cultural pages of newspapers? How are different countries of origin related to legitimacy and hierarchies? And is the expansion of pop-rock really a matter of uniform “global culture”, or has it been domesticated differently in different parts of the world (cf. Regev 2013; Varriale 2016)? Questions like these have already generated research, but more is still needed, and this paper’s bird’s-eye view of trends—as well as the data set—might be useful for further analyses. It is also clear that our quantitative content analysis is far from exhaustive; rather, it maps the overall trends that merit closer attention and analysis. Thus, future research also needs to employ qualitative approaches to clarify such questions. 21 While our findings support our second major hypothesis—our selected European newspapers show variance in the timing and thoroughness of the opening of legitimate culture—the manner in which the newspapers differed was not totally in accordance with our expectations (H2a–b). When the evolution of the proportions of emergent versus established art forms over time is used as a proxy for cultural opening, the newspapers in our sample can be placed in the following order of approximate openness: (1) ABC/EP, DN, (MIL); (2) LM; (3) GU; and (4) HS. However, when the between-genre evolutions within an established art form—in this case, classical versus pop-rock music—is taken as a proxy for cultural opening, the order is slightly different: (1) DN, (MIL); (2) ABC/EP; (3) GU; and (4) LM, HS. In both cases, the position of GU as a relative latecomer was unexpected, since the UK is more centrally positioned in the global system than other countries in the sample, its media system is arguably more liberal, and it is not included among the countries where state funding for the established arts is high (H2a). In other words, rather than indicating society-level dimensions, the result relates to the characteristics and position of The Guardian itself. Moreover, MIL can be ranked in the first category in both measures, but its position really merits the brackets; the journalistic attention that MIL shows to emergent cultural forms and pop-rock today is not part of a gradual opening process, since Western highbrow culture was never canonized in Turkey (Belge 1996). Rather, Turkey occupies a peripheral position in the globalization scale, and the state’s attempt to Westernize its cultural sphere following the foundation of the secular republic was never effective enough to heighten established art forms.6 In that sense, while MIL appears to be among the most open case in our sample, it demonstrates the geographical and cultural limits of the Western cultural canon rather than suggesting a trend towards cultural declassification. Thus, the main lesson taken from MIL in this study is that, while MIL does not show the same European trends evident in other newspapers, it can nevertheless be interpreted as indirectly validating the results on the opening of legitimate culture in other European countries. Taken together, our results are quite divergent when it comes to the three society-level dimensions discussed: the newspaper’s position in the global cultural system (Appadurai 1996), media system models (Hallin & Mancini 2004), and cultural policy regimes (Looseley 2011). For instance, European (and even global) centres of popular culture, such as the UK and France, have, in the cases of GU and LM, a rather traditional outlook on culture. In addition, Finland and Sweden, similar countries both politically and policy-wise, show clear differences through DN and HS, irrespective of the fact that both newspapers are embedded in a similar cultural policy regime and democratic-corporatist media system. One may ask whether the comparative media system model 22 approach is ultimately a suitable framework in the case of these kinds of “close” comparisons of culture and arts coverage in quality newspapers, as the model heavily focuses on media systems’ institutional, legal, and political dimensions (Hallin & Mancini 2004). The data nevertheless reflect some structural similarities. For instance, the peculiar political history of Francoist Spain could be traced through the cultural coverage of ABC/EP, turning rapidly from a conservative perspective towards popular culture. Also, the difference between DN and HS—HS (together with LM) representing the most conservative newspapers in this sense and DN being characterized by its early breakthrough of pop-rock coverage—is again illustrative. Though HS and DN share similar media system models and cultural policy regimes, the difference may be due to Sweden’s more central position in the global (and European) cultural system, as expected (H2b). Moreover, rather than being reduced to any of the three society-level dimensions considered, the difference could be better accounted for with field-specific and institutional factors. As for music, Sweden has deeper roots in the popular music industry (Seabrook 2015), whereas, in Finland’s field of music and general culture, the position of classical music has been very strong throughout the twentieth century (Korhonen 2002). Furthermore, the difference might be due to generational changes inside cultural departments (Jaakkola 2015). Thus, it is difficult to demarcate the roles of field-specific characteristics and society-level characteristics in the shaping of newspapers’ cultural coverage. It may be tempting to overestimate the role of the latter, whereas the organizational or field-level factors may sound relevant only when the results are not in line with society-level expectations. Perhaps not surprisingly, previous studies by Janssen et al. (2011) and Schmutz et al. (2010) identify the same issue. Janssen et al. show that shifts from highbrow to popular cultural coverage in European newspapers were not entirely consistent with society-level factors. They conclude that the “‘popularization’ of arts journalism may be primarily associated with field-level rather than society-level factors” (2011:160). We can only wholeheartedly concur, while also agreeing with Schmutz et al. that “there is considerable variation in the cultural hierarchies of the [four] countries, even though the overarching trend has been towards more ‘general validation’ of popular music as indicated by its inclusion in the pages of elite newspapers” (2010:507). Overall, we see that newspapers are a good source for studying shifts in cultural legitimacy and hierarchies, while different extra-cultural contexts (whether socio-cultural, historical, policy related, organizational, or simply the results of editorial decisions) also play a role in what newspapers 23 legitimize and present as art and culture. The paradox of such data is perhaps this: on the one hand, newspaper culture sections are sensitive to different types of trends and changes in how culture is valued and evaluated (which is the reason to study culture sections in the first place); yet, this sensitivity can simultaneously lead to muddled results if there are large-scale and rapid transformations occurring in the newspapers’ socio-political environments. On the other hand, the entire premise of studying cultural change through newspaper culture sections rests on the idea that the status and the role of newspapers as institutions are relatively stable. Otherwise, it would be impossible to use newspapers as a fixed point of departure enabling the investigation of large, longterm processes that are of interest to cultural sociologists, such as the de-hierarchization, popularization, and globalization of culture (cf. Baumann 2007; Janssen et al. 2008; Fishman & Lizardo 2013). The case of Turkish MIL exemplifies the social sensitivity of the newspaper data in the first sense, whereas the other newspapers represent the relative stability of the institution, allowing us to show how the opening and expansion of the sphere of legitimate culture has taken place in different parts of Europe since the 1960s. Endnotes 1 In the last few years, the Internet and other digital devices have become significant competitors of traditional print newspapers (McChesney 2011). It is yet unknown what this means for newspapers’ and traditional critics’ authority as cultural mediators (Verboord 2010). This new situation obviously affects our data, albeit only regarding the last year (2010) in our sample. 2 The data were collected as part of a larger research project aiming to analyze the changing relationship between cultural and social stratification in post-1960s Europe (Purhonen, 2012; Purhonen et al., 2015). 3 Following a careful planning of the coding system, the data were coded (by eleven coders altogether) between 2014 and 2015. Before the final coding procedure, a preliminary, more restricted coding system was established and tested on GU and HS using only three sample years (Purhonen et al., 2015). 4 Because an article may include a discussion on multiple cultural areas, the coding system defined the “primary cultural area” of the article as the cultural area mentioned first, preferably deducible from the headline (or similar) of the article itself. 24 5 If time points had been used as a categorical variable, it would have resulted in too many interaction terms to be reasonably reported. 6 The finding regarding the rise of emergent cultural forms can also relate to the peculiar position of MIL in the Turkish media field (Adakli 2009). MIL was known to lean into the social democratic tradition until it was bought by a major industrial conglomerate in Turkey in the 1980s, which led to a change in the paper’s publishing style. Moreover, the business group that bought MIL established a TV channel in 1993, which has become one of the most popular private channels in Turkey. It is likely that an important portion of the paper’s dramatic increase in TV articles stems from the economic field’s monopoly over media and cultural journalism. Acknowledgements This is a revised version of a paper presented at the annual meeting of the American Sociological Association (regular session of the Culture Section), Seattle, USA, 20–23 August 2016, and at some other seminars and meetings. We are grateful to all those who commented on the earlier versions of the manuscript, especially Clayton Childress, Philippe Coulangeon, Jukka Gronow, Susanne Janssen, Tina Lauronen, Carlos J. Fernández Rodríguez, Marc Verboord, Alan Warde and David Wright. The research carried out in this study has been made possible by the financial support from the Academy of Finland (grant number: 291619), Kone Foundation and the University of Helsinki research grants. References ACE (1998). International Data on Public Spending on the Arts in Eleven Countries. London: Arts Council of England. Adakli, G. (2009). The process of neo-liberalization and the transformation of the Turkish media sector in the context of the new media architecture. In J. Harrison & B. Wessels (Eds.), Mediating Europe: New Media, Mass Communications and the European Public Sphere. New York: Berghahn Books. Appadurai, A. (1996). Modernity at Large. Minneapolis: University of Minnesota Press. Baumann, S. (2001). Intellectualization and art world development: Film in the United States. American Sociological Review, 66, 404–426. 25 Baumann, S. (2007). Hollywood Highbrow: From Entertainment to Art. Princeton: Princeton University Press. Belge, M. (1996). Kültür: Cumhuriyet Dönemi Türkiye Ansiklopedisi, İletişim Yayınları, 5, 1287– 1304. Berkers, P., Janssen, S., & Verboord, M. (2014). Assimilation into the literary mainstream? The classification of ethnic minority authors in newspaper reviews in the United States, the Netherlands and Germany. Cultural Sociology, 8, 25–44. Bourdieu, P. (1984). Distinction: A Social Critique of the Judgement of Taste. Cambridge: Harvard University Press. Bourdieu, P. (1993). The Field of Cultural Production. Cambridge: Polity Press. Çaplı, B., & Dündar, C. (1995). 80’den 2000’lere Televizyon. Cumhuriyet Dönemi Türkiye Ansiklopedisi, 15, 1376–1386. Daloz, J-P (2010). The Sociology of Elite Distinction. Basingstoke: Palgrave MacMillan. DiMaggio, P. (1982). Cultural entrepreneurship in nineteenth-century Boston: The creation of an organizational base for high culture in America. Media, Culture & Society, 4, 33–50. DiMaggio, P. (1987). Classification in art. American Sociological Review, 52, 440–455. DiMaggio, P., & Mukhtar, T. (2004). Arts participation as cultural capital in the United States, 1982–2002: Signs of decline? Poetics, 32, 169–194. Eveno, P. (2004). Histoire du Journal Le Monde, 1994–2004. Paris: Albin Michel. Fishman, R.M., & Lizardo, O. (2013). How macro-historical change shapes cultural taste: Legacies of democratization in Spain and Portugal. American Sociological Review, 78, 213–239. Friedman, S., Savage, M., Hanquinet, L., & Miles, A. (2015). Cultural sociology and new forms of distinction (Introduction to the special issue). Poetics, 53, 1–8. Frith, S. (1996). Performing Rites: On the Value of Popular Music. Cambridge: Harvard University Press. Hadenius, S. (2002). Dagens Nyheters historia: Tidningen och makten, 1864–2000. Stockholm: Bokförlaget DN. Hallin, D.C., & Mancini, P. (2004). Comparing Media Systems. Cambridge: Cambridge University Press. Heinich, N. & Shapiro, R. (Eds.) (2012) De l’artification: Enquêtes sur le passage à l’art. Paris: Editions de l'Ecole des Hautes Etudes en Sciences Sociales. Hesmondhalgh, D. (2007). The Cultural Industries (2nd Edition). London: Sage.