Music mavens revisited: Comparing the impact of connectivity and dispositions in the digital age

Abstract

This article aims to contribute to work on music mavenism with the explicit goals to analyze how important connectivity (e.g. someone social network) is in the way music spreads, compared to dispositions (schemes of perception and appreciation). Concretely, I study how individuals find new music and who are involved in spreading information on music. I use data from a representative survey on the Dutch population to examine this. Music discovery can be characterized by three types of practices: individualized networks, traditional expertise, and mass media. The results also show that successful influencers – mavens – differ only slightly from these listeners who are highly involved in exchanging music tips. Dispositions – particularly cultural capital – remain highly influential in explaining not only how people find music but also who recommends music and still seem to be more important than connectivity.

Keywords

Cultural consumption mavens cultural markets connectivity cultural capital word-of-mouth

Introduction

How certain cultural expressions become popular or fashionable, and others not, remains an intriguing question in the study of cultural consumption. Traditionally, cultural mediators have held a pivotal role in informing audiences about the latest supply – often via mass media, such as newspapers, television, radio, and magazines (Kristensen and From, 2015; Janssen and Verboord, 2015) – by making selections and providing evaluations of market supply. Still, a significant proportion of this influence tends to run via individual consumers who spread word of mouth (WOM) among people they know – the most influential of these peers are known as “opinion leaders” or “mavens” (Rogers, 2003). Parallel to this scholarship, the social stratification of taste has been mapped in the past decades building on Bourdieu’s (1984) cultural capital model, which emphasizes that underlying dispositions in terms of cultural taste and knowledge are strongly related to social status of and the media that are used by a particular class. While the work of Bourdieu and followers often revolves around taste reproduction – emphasizing stability rather than change – it also recognizes the importance of individual networks for the (mediated) transmission of taste with the important addition that not all peers have equal status or influence (see McLean, 2017).

In this article, I argue that the rise of Internet and social media in recent years has made it imperative to bring these research strands (WOM and cultural capital) closer together as the structures of cultural markets are in radical transition. The availability of mediatized cultural products has increased tremendously (e.g. Aguiar and Waldfogel, 2016), and the circulation of products has intensified due to new modes of distribution and acquisition such as streaming and downloading (López-Sintas et al., 2014; Ordanini and Nunes, 2016). Among younger generations, traditional media are losing ground to websites, vlogs, social media networks, and other new media, implying that they are less connected to legitimated mediators but more to their peers. At the same time, cultural legitimacy as a structuring device of cultural fields is under pressure due to the gradual decline in highbrow cultural participation and the new conceptions of cultural capital that are simultaneously emerging (e.g. Prieur and Savage, 2013). Cultural taste is often said to be more loosely connected to social position (see also theoretical section) and, as such, more individualized.

This article aims to study how music consumers find information on new music and to what extent they are also involved in spreading information on music. Not only does this contribute to unresolved questions on how ideas and fads diffuse across populations, it also increases our understanding of how emerging technologies are incorporated in the everyday practice of cultural consumption (Leguina et al., 2017; López-Sintas et al., 2014). Affordances of new media have not often been connected to fields of culture. There is a large body of research on opinion leaders, market mavens, and WOM in the field of marketing, but this work is generally focused on identifying relevant people for implementation in persuasive communication (e.g. Walsh and Mitchell, 2010). Cultural consumption studies often analyze what is consumed or liked or how cultural expressions are consumed or appropriated within larger consumption repertoires (e.g. Jarness, 2015; Michael, 2015). How individuals discover products and, through habitual exploration strategies, shape their taste repertoires is less clear. One of the few papers that provide an in-depth account of music discovery is the study by Tepper and Hargittai (2009) that shows for a sample of American students that digital media – by that time – had become an additional – but not dominant – source of information. This article – in some respects – replicates that study, but it also goes beyond it by updating the role of technology, offering a more fine-grained explanatory model of finding and spreading music, and testing explanatory mechanisms on population survey data that are representative for the Netherlands.

Theoretical background

Cultural markets, mediators, and WOM

While there is a large literature on cultural consumption (including music consumption), which studies the frequencies, repertoires, and the meanings of culture and consuming culture (e.g. Van Eijck, 2001; Warde, 2014; Yaish and Katz-Gerro, 2012), less is known about how individuals arrive at their specific choices – in other words, how they discover products. Choice behavior in cultural markets links consumers with the product categories created by gatekeepers and tastemakers (DiMaggio, 1987; Janssen and Verboord, 2015). However, this does not rule out the possibility that some categories contain much larger supply than others; consumers use different, more nuanced genre labels when searching for products (e.g. Goldberg, 2011); and that new classification labels come into existence through new modes of consumption (e.g. Airoldi et al., 2016).

Ultimately, discovery implies finding culture among a set of cultural items that has already been filtered, edited, given meaning, and sometimes marketed, by mediators in the cultural industries such as marketeers, cultural journalists, and critics (Janssen and Verboord, 2015; Smith Maguire and Matthews, 2012). Reconstructing how music spreads thus means that discovery and influence are two sides of the same coin. Individuals searching for new products rely on sources that make supply transparent, whereas mediators who propagate items from this supply depend on audiences that are partial to their efforts. Not only is the actual influence by cultural mediators on concrete purchases debated (e.g. Holbrook and Addis, 2008), mass communication research has also emphasized, from the 1950s onwards, the role of recipients of media messages in their studies on opinion leadership and WOM (Katz and Lazarsfeld, 1955). These theories conceptualized media influence as a two-step process with influential individuals (opinion leaders) gatekeeping information from the mass media to a larger public via interpersonal communication (see Weimann, 2015). Since then, various research traditions have developed their own perspectives on how and why communication in social interactions is used for discovery and can be influential, and yet, few focus on cultural consumption.

Discovery of cultural products

In cultural markets, consumers use WOM to deal with the uncertainty of consuming products whose value can only be established by experiencing (Caves, 2000). However, WOM is also symbolically potent as consumers like to converse about creative goods: cultural products provide topics of conversation both in established relations and in more casual exchanges (DiMaggio, 1987; Lizardo, 2016). The advent of Internet technologies has extended consumers’ options to articulate opinions, find information, and to spread WOM beyond their face-to-face contacts, which has instigated the study of electronic word of mouth or eWOM (Hennig-Thurau et al., 2004). Consumers more actively search for information online that they cannot find in their personal networks, for instance on customer websites, blogs, or, in the case of music, on streaming services (e.g. Kjus, 2016). For certain groups, online tools have become part and parcel of how they connect to culture in everyday life: they watch music clips on YouTube, visit platforms where they engage in discussions, and exchange recommendations through social media (Nowak, 2016).

Still, we do not know a lot about how cultural consumers look for information and make music discoveries in contemporary cultural markets. In their mid-2000s study of American college students, Tepper and Hargittai (2009) found three main ways to find new music: their social networks, traditional media (radio, television, etc.), and technology (file sharing, browsing the Internet, etc.). These pathways of exploration were hardly aligned to specific socioeconomic or demographic profiles and only slightly to behavioral patterns. More recent studies tend to concentrate on streaming services such as Spotify (e.g. McCourt and Zuberi, 2016). However, they generally focus on how users experience and make sense of streaming technologies rather than on how users incorporate them in broader discovery repertoires.

Influencing the purchase of cultural products

At the core of most studies into personal influence lies the proposition that person-to-person communication, or WOM, has certain properties which increase the likelihood of influence as well as the diffusion of influence. First of all, WOM involves communication with other people, which lends more credibility to the source of information than, for instance, paid media. Second, WOM often takes place between individuals who know each other (family, friends, and acquaintances), which further increases the trustworthiness. Third, WOM is embedded within networks implying that information can travel exponentially fast if individuals who receive information pass this on to several others. Due to these properties, WOM is recognized – particularly in the field of marketing – as an important mediating force. Both opinion leadership and opinion seeking are considered integral to the construct of WOM, since it constitutes a basic form of social interaction (Flynn et al., 1996; Sun et al., 2006).

At the same time, some people – not necessarily the traditional opinion leaders – are highly involved in producing and distributing online information (cf. user-generated content or UGC), which has reinvigorated the interest in a specific category of influencers, the market mavens (Feick and Price, 1987). These influencers have information about many kinds of products, shopping channels, and other facets of markets and are seen as more pro-active in disseminating information to others. They also make strong usage of online technologies in order to keep themselves updated with the latest developments (Boster et al., 2011; Goldsmith et al., 2006; Walsh et al., 2004; Walsh and Mitchell, 2010).

One of the recurring challenges in the field of WOM and eWOM, however, remains how to identify and mobilize opinion leaders and mavens, as well as those likely to be influenced by them (opinion seekers). Marketing-based approaches tend to emphasize psychological profiles (Goldsmith and Clark, 2008; Ruvio and Shoham, 2007), whereas sociological studies often take the broader scope of how ideas and products diffuse across society (Rogers, 2003) and therefore see opinion leaders as innovators and influencers in social networks (Burt, 1999; Wejnert, 2002). With regard to culture, in the study by Tepper and Hargittai (2009), mavens were found to be more experimental, more omnivorous, and also receiving more recommendations. Other studies (also using non-representative data) suggested that online music mavens have information on many genres, know their way across many websites, and often are active on such websites (e.g. by initiating discussions; Sun et al., 2006; Walsh and Mitchell, 2010).

Dispositions versus connectivity

Cultural taste is subject to processes of social stratification, as many studies into cultural consumption have shown, including those on music (Bennett et al., 2009; Leguina et al., 2017; Van Eijck, 2001). Individuals belonging to higher-status groups tend to show greater interest in cultural expressions that are considered more legitimate in society, which has been explained by their larger cultural capital (Bennett et al., 2009; Bourdieu, 1984). While cultural capital can refer to a broad range of social assets (competences, skills, credentials, and behaviors), it is often thought to be structured by underlying dispositions that are shaped by both previous experiences (e.g. socialization) and adjustments to new situations (e.g. “watercooler talk” at work). Dispositions are schemes of perception and appreciation. In their most legitimate form – aesthetic dispositions – they are geared toward aesthetic preferences that favor form above content and, as such, entail familiarity with and skills to appreciate more complex forms of culture (Bourdieu, 1984; Daenekindt and Roose, 2014; DiMaggio, 1991). As scholarship on cultural capital and cultural consumption developed, also other perspectives have been put forward, such as the concept of cultural omnivorousness: the tendency to combine culture with different degrees of legitimacy in single repertoires (Jaeger and Katz-Gerro, 2010; Peterson, 2005).

Discovery of culture is generally less associated with social inequalities, since opinion leaders are found in all realms of society (Rogers, 2003). However, as symbolic boundaries between cultural forms are blurring and cultural capital is changing accordingly, being in the know appears to be increasingly a distinctive feature (Prieur and Savage, 2013). Showing awareness of new cultural trends can provide status when consumers are able to display their dispositions as a form of position taking that meets acceptance among peers (Bourdieu, 1984; Prieur and Savage, 2013). While showing knowledge of the latest cultural expressions could allure the admiration of relevant others, often this is limited to one’s social network. Nevertheless, this social network does act as a resource for knowing what is trendy and what is acceptable (McLean, 2017: 34ff). On one hand, social networks reiterate the impact of socioeconomic status (SES) and associated dispositions: individuals are more likely to interact with people of similar sociodemographic background than with dissimilar others (the homophily principle), which affects their cultural consumption practices (Mark, 2003; Rössel and Pape, 2016). Concretely, sharing certain dispositions enables people to bond through discussing preferences and non-preferences (which also implies that the relationship between cultural taste and networks is reciprocal, see Lizardo, 2006). On the other hand, social networks also allow new influences to enter. They shape cultural consumption by way of information channeling (Kane, 2004) – an argument which can be traced back to Granovetter’s (1973) work on strong and weak ties and to previously mentioned studies of opinion leadership (Katz and Lazarsfeld, 1955; Rogers, 2003).

Another reason to take the role of connectivity in the study of music mavenism more seriously lies in the rapid advent of digital media, and, more specifically, social media platforms (Baym, 2015; Beer and Burrows, 2010; Van Dijck, 2013). Social media have made contact moments with relevant others more frequent and more ubiquitous (Van Dijck, 2013). Importantly, being connected online also expands the network beyond friends and close relatives. Not only are the costs of linking to “weak ties” virtually the same as to “strong ties,” alternative connections enable consumers to contact like-minded strangers (e.g. fan communities and discussion forums) as well as to being exposed to social influencers, such as vloggers, on content sharing platforms (e.g. YouTube; Carter, 2016). Online media have also boosted the importance of “sharing” (information, pictures, clips etc.) as smartphones have enabled consumers to do this in one click (John, 2013; Kang and Namkung, 2016). Producing and finding UGC can thus also be seen as important ingredients of online connectivity. Arguably, these developments impact the way information flows and thus how individuals discover culture.

Method

Data were collected via the Longitudinal Internet Studies for the Social sciences (LISS) panel: an online panel that is based on a true probability sample of households drawn from the population register by Statistics Netherlands. It can be considered representative for the Dutch population also because individuals without computer or Internet are facilitated to participate (https://www.lissdata.nl/about-panel). In total, 848 individuals filled out the survey that was distributed in June 2015. In the analyses, we excluded the 82 respondents who said they never listen to music (these respondents did not get follow-up questions on music). Furthermore, we lost 74 respondents who stated they had no social network and therefore were not asked follow-up questions on their friends and connectivity.

The empirical research focuses on two aspects of mavenism: (a) the extent to which peers and social media are important for general music audiences when looking for new music, and (b) identifying mavens among general music audiences, as well as characterizing them in terms of social position and music involvement. Thus, I study WOM from two perspectives: the individuals that seek recommendations and those that provide them (cf. the mavens).

Measurements

The survey asked which channels individuals use to find new music (artists or songs they did not know before). Eight options were provided (in some cases with examples): friends/acquaintances mention artist or song, friends/acquaintances recommends music via social network site (e.g. Facebook), friend/acquaintance gives CD or vinyl album, listening to radio or watching television, reading about music in a newspaper or music magazine (print), reading about music on a website (e.g. on Pitchfork, Allmusic, Amazon), listening to music on an online source (e.g. iTunes, Spotify, YouTube), searching for music in a record store, and other (all have four response categories from “never” to “often (1× per week or more)”).

To measure mavenism, I followed Tepper and Hargittai’s (2009) operationalization that emphasizes concrete behaviors. Respondents were asked to which degree they considered themselves as people who sometimes recommend music to others, and, if so, how often they do so (no; yes, every now and then; yes, regularly; yes, often). Individuals who answered affirmatively were also asked (a) how often people that were recommended the music also listened to the music (again four response categories: never, every now and then, often, and always) and (b) the number of people they regularly or often recommend music to (no one, 1–2 people, 3–5 people, and more than 5 people).

Dispositions

Dispositions are measured through three indicators that signal how individuals perceive and appreciate cultural expressions: cultural capital, parental pop music socialization, and music omnivorousness. For pragmatic reasons, I operationalize cultural capital as familiarity with legitimate culture in the traditional meaning (Bourdieu, 1984; DiMaggio, 1991). Cultural capital is measured via participation in nine cultural activities: visiting theater performances, visiting ballet/modern dance performances, visiting art museum or gallery, visiting classical music concert or opera, visiting art house movies, watching Dutch quality TV-drama, watching foreign quality TV-drama, reading literary novels, literary essays or poetry (five categories), plus appreciation of two music styles: classical music and opera (5-point Likert-type scales). Cronbach’s alpha = .790.

To gauge specific dispositions in the domain of music, I analyze parental socialization (e.g. Willekens and Lievens, 2014). Parental pop music participation was measured using three questions on pop music socialization. One probes how often the parents (alone or together) visited pop concerts (including jazz and blues) when the respondent was about 12 years old. The other two ask how often the parents listened to pop music at home; this was asked for separately for the father and the mother. The response categories are as follows: never, sometimes, and often. I calculate the mean score.

Music omnivorousness is measured via the appreciation of 12 music styles. Since omnivorousness assumes diversity in taste, I group these styles together – based on the outcomes of a factor analysis – in four subcategories: (a) classical music, opera, jazz; (b) rock, alternative rock, hard rock; (c), top 40 music, hip hop, R&B; and (d) country, Dutch music, cabaret music. For every style, dummy variables are created that distinguish between, on one hand, liking and liking very much and, on the other hand, not liking (very much) or being indifferent. I then count the number of preferences within each category and calculate the distribution of counts over the four categories using an entropy measure (see also Verboord, 2010). The final variable is recoded so that it runs from completely univore (only choices in one category; 0) to completely omnivorous (perfect distribution across all four categories; 1). Note that respondents who do not appreciate any genre are recoded as completely univore (this amounts to 107 individuals).

Connectivity

In the analysis, I distinguish between general connectivity (in the traditional sociological conception) and online social activities (following studies of eWOM). It is increasingly difficult to disentangle offline and online social lives. The first dimension aims to take stock of network properties without making claims of where contacts are made or maintained, but focuses mostly on “strong ties.” The second dimension turns attention to online behavior that is likely to put individuals in contact with others, yet restrains from asking about amounts of actual contacts realized.

Respondents are asked to assess the size of their personal network via the following question: “Not regarding family and partner, with how many friends and acquaintances do you know well enough to discuss personal or important issues with?” This question is adapted from previous measures of network size (e.g. Lizardo, 2006). In addition, the survey contains a network name generator to elicit information on the three closest friends of the respondent. One of the questions probes the similarity in music taste: “To which extent do these friends have the same music taste as you do?” (4 response options: 0 = “friend does not like music,” 1 = “absolutely different taste,” 2 = “partly same taste,” and 3 = “(almost) the same taste”). I first calculate the average score and then create four dummy variables that overlapped with the original categories. Thus, average scores just below and above 1 are coded as “absolutely different taste,” and so on.

Besides the estimated network size, I also rely on a more subjective measure of connectivity. Regardless of the size of a network, some individuals may have more central positions in their networks. The survey asks the extent to which respondents think they are “connectors”: bringing other people into contact with each other. Concretely, the following statements on friendships were presented: “I am often the link between friends in different groups,” “I often introduce people to each other,” “I try to get persons in contact with each other if I think they would like each other,” “Many people that I know, know each other via me” (Likert-type scales from 0 = “very untrue” to 4 = “very true”). Connector status turned out to be a reliable scale (Cronbach’s alpha = .869). I calculate the mean score of the statements.

Three questions are used to assess how connectivity extends online. First, respondents were asked the frequency at which they use Facebook and Twitter (Five categories from 0 “never” to 4 “(Almost) daily”). I calculated the mean score. Two more indicators concern the involvement of individuals with UGC. I measure the degree to which respondents consult UGC for finding information on culture, via the mean score for six forms of consulting UGC: reading reviews about a cultural product by other consumers, reading Internet Movie Database (IMDb), reading websites that collect media reviews, reading webzines or blogs on music, reading webzines or blogs on various cultural forms, and using webstores with reviews. The reliability proved to be good (Cronbach’s alpha = .730). Finally, I examine the production of UGC on cultural matters. This is measured by calculating the mean score for five forms of generating UGC: writing reviews of cultural products; rating a product; writing stories/articles on culture for a blog or website; chatting with other people on a forum; and following an artist, writer, or cultural celebrity on Twitter or Facebook. The reliability is moderately good (Cronbach’s alpha = .651).

The following variables have all been rescaled between 0 and 1 (by dividing by the maximum): music listening frequency, cultural capital, pop music participation parents, music omnivorousness, network size, self-perceived connector, using Facebook and Twitter, producing UGC, and consulting UGC.

Control variables

In the explanatory analyses, I control for the usual demographic characteristics: age, sex, ethnicity, and highest attained educational level (in six categories). For ethnicity, a dummy variable is created that probes whether individuals had a migrant background (either first or second generation of non-Dutch descent, both western and non-western).

Data analysis

The mapping of discovery strategies is done by conducting an exploratory factor analysis (see Field, 2013: 665ff). The explanatory analyses employ two forms of multiple regression analysis: ordinary least squares (OLS) regression analysis for the continuous variables and ordinal regression for the variables that have an ordinal level of measurement (see Field, 2013: 293ff). Assumption checks for the OLS regression did not show problems regarding multivariate normality, multicollinearity, homoscedasticity, and auto-correlation.

Results

Strategies of finding music

To establish which patterns can be found in the ways music listeners find music nowadays, I conducted a factor analysis with the nine variables that represent different search strategies. The results are presented in Table 1. The first dimension consists of three variables: friends mention artists or songs in conversation, friends recommend music via social media, and listening to music via an online source, such as Spotify. This factor is therefore labeled the individualized network strategy (to find music). The second factor signifies search strategies that have a more traditional bearing. Here, the following variables have high loadings: friends give CD or album, reading about music in a magazine or newspaper, searching in a record store, and – with a slightly lower loading – reading about music on a website. These variables share a focus on traditional mediators and sound carriers (record stores, physical products, and music journalism), implying a holding on to institutional sources for finding music (hence the label traditional expertise). Finally, the third factor is composed of only one variable: finding music via radio and television.

Table 1.

Ways of finding new music.

	Individualized network	Traditional expertise	Mass media
Friend mentions artist/song in conversation	.734	.088	.052
Friend recommends music via social network site (e.g. via Facebook)	.771	−.011	−.131
Friend gives me CD or vinyl album	−.027	.752	−.031
Via listening to the radio or watching television	.228	.016	.817
Reading about music in a newspaper or music magazine (print)	.048	.698	.281
Reading about music on a website (e.g. Pitchfork, Allmusic, Amazon)	.374	.414	.010
Listening to music on an online source (e.g. iTunes, Spotify, YouTube)	.802	−.085	.072
Searching in a record store	−.070	.763	−.177
Other	.286	.041	−.568
% of variance	30.007	13.941	12.485
Cronbach’s alpha	.674	.628	NA

Rotation direct oblimin, pattern matrix; factor loadings in bold are used for interpretation factor.

Note that these results imply that there is not only differentiation in how networks operate but also in the role of online media. Listeners who still draw a lot on traditional expertise also use networks, but in a different way (exchanging CD or albums) than the more individualized listeners. And the former are also found online, but for consulting websites – in similar ways as they consult music magazines – rather than downloading or streaming music.

It is also important to signal that the three factors indicate which underlying dimensions are found in the data, not which strategy is employed most often. In fact, inspection of the frequencies of the three new variables shows that two-third of the respondents regularly or often use mass media, while traditional expertise is (on average) hardly used more than once a month. The individual network strategy is used more often, but also here not many people do this on a regular basis.

The second step is an analysis of who uses what strategy of finding music. Here, I test the importance of dispositions and connectivity on the choice of strategy. The three dimensions are converted to variables by calculating the mean score of the variables that have high loadings on the factors (as described before). Since the third dimension only consists of one variable, this remains unchanged. I conduct two OLS regression analyses (individualized network strategy and traditional expertise strategy) and one ordinal regression analysis (mass media strategy), which are reported in Table 2.

Table 2.

Explaining repertoires of finding music (N = 656).

	Individualized network			Traditional expertise			Mass media
	Model 1a	Model 1b	Model 1c	Model 2a	Model 2b	Model 2c	Model 3a	Model 3b	Model 3c
Female	−.034	−.015	−.008	.026	.039	.061*	.325*	.405**	.439**
Age	−.017***	−.017***	−.011***	.000	.001	.002*	.003	.005	.007
Educational level	−.024	−.018	−.018	−.022*	−.017	−.014	.070	.091∼	.082
Migrant origin	.140*	.141*	.162*	.067	.063	.066	−.403*	−.312	−.284
Music listening frequency	.476***	.465***	.390***	.316***	.324***	.287***	1.616***	1.655***	1.586***
Dispositions
Cultural capital	.664***	.582***	.364***	.715***	.652***	.468***	−.935**	−1.147**	−1.400***
Pop music part parents	.348**	.310**	.176∼	.100	.089	.010	.151	.099	.036
Omnivorous music taste	−.075	−.068	−.054	−.050	−.041	−.044	.459*	.670**	.682**
General connectivity
Network size		.048	.019		.082	.093∼		.436	.451∼
Friends do not like music		−.195	−.170		.285*	.215∼		−.584	−.523
Friends have different taste		−.163**	−.146*		−.081∼	−.075∼		−.332	−.318
Friends have partly same taste		Ref.	Ref.		Ref.	Ref.		Ref.	Ref.
Friends have same taste		−.032	−.036		−.028	−.014		−.240	−.215
Self-ascribed connector		.442***	.360***		.207**	.166*		.354	.391
Online social activities
Use Facebook and Twitter			.389***			−.137*			.038
Producing UGC			.276**			.390***			−.448
Consulting UGC			.738***			.417***			1.148*
Constant	1.230***	1.035***	.640***	−.006	−.155	−.243**
Threshold cat.1							.656	1.186*	1.268*
Threshold cat.2							2.178***	2.748***	2.872***
Adj. R²	.337	.365	.434	.174	.196	.255
Nagelkerke R²							.083	.105	.112

Unstandardized coefficients; individual and traditional: OLS regression; mass media: ordinal regression (variable 3 categories); test of parallel lines, Model 1: Chi² = 9.357, p = .313; Model 2: chi² = 10.718, p = .634; Model 3: Chi² = 13.180, p = .660. Significance levels: ∼p<0.10 *p<.05 **p<.01 ***p<.000.

I discuss the analyses by comparing the results for the same models (a, b, and c) for the three dimensions. When comparing the first models – that contain the demographic variables and disposition indicators – some clear differences emerge. Stronger adherence to the individual network strategy is – as expected – more common among younger listeners, but there is not a lot of evidence that the traditional expertise strategy is the exclusive domain of the older generation. The influence of dispositions differs across search strategies. Possession of cultural capital positively affects both the individualized network strategy and the traditional expertise strategy, but is negatively associated with using mass media. This seems in line with previous studies of cultural capital in the sense that the first two – more active – strategies represent the more “engaged” relation with music (see Bennett et al., 2009: 48ff). At the same time, for the mass media strategy, the negative effect of cultural capital goes together with a strong impact of music listening frequency, suggesting that one’s general set of cultural competences and dispositions is decisive for choosing more pro-active or more lean-back ways of finding music. Pop music socialization by the parents is positively related to the more individualized network strategy, although this effect mostly disappears when producing and using UGC are taken into account (Model 1c). Showing a more omnivorous music taste only contributes to using mass media.

Next, the connectivity variables are added to the models (1b, 2b, and 3b). Network size does not seem to have an impact, although additional (unreported) analyses show that for the traditional expertise and mass media strategy, its positive effect is dampened by the connector variable. The more people state that they are strong connectors, the more often they use the first two strategies. Thus, being a connector is most of all related to engaging in more active search strategies, not just to the individualized network strategy. I find only limited influences of the taste structures in networks. Compared to having more or less the same taste as one’s friends, having a different taste than one’s friends decreases the use of the individualized network strategy as well as the use of the traditional expertise strategy. Finally, I add the online social activities to the model. Clearly, people who are active in producing UGC and who regularly consult UGC are more likely to engage in both the individualized network strategy and the traditional expertise strategy, but not the mass media strategy. Social media usage is only positively related to the individualized network strategy.

Mapping and explaining music mavenism

Which listeners are involved in recommending music and who is more prone to be recommended music? Tables 3 and 4 provide descriptive results. As can be seen in Table 3, a little more than a third of the respondents claim never to recommend music to others. A slightly larger group (43%) sometimes recommends music to others, while regularly or often making recommendations is rarer: only 10% of the respondents can be categorized as active recommenders. The same pattern was found for the other aspects of music recommendation. The majority of people recommending music stated their advice was “sometimes” followed. Most recommenders only addressed one or two people, just like the majority of all listeners who received recommendations from one or two people.

Table 3.

Giving music recommendations.

	Ever recommend music to others?		If you do, do they follow your advice?	To how many people do you regularly recommend music?
No, never	307 (35.8%)
Yes, sometimes	370 (43.1%)	No, never	30 (3.5%)	None	23 (2.7%)
Yes, regularly	60 (7.0%)	Yes, sometimes	350 (40.8%)	1 or 2	322 (37.5%)
Yes, often	29 (3.4)	Yes, often	64 (7.5%)	3–5	93 (10.8%)
		Yes, always	15 (1.7)	More than 5	21 (2.4%)
N	766		459		459

Table 4.

Receiving music recommendations.

	From how many people do you regularly receive music recommendations?
None	271 (31.6%)
1 or 2	396 (46.2%)
3–5	80 (9.3%)
More than 5	19 (2.2%)
N	766

Can we explain who receives or spreads information? Table 5 presents the results of three regression analyses that examine which music consumers are more likely to receive information, give information, and be more maven-like. First, I find a strong positive influence of cultural capital for all indicators. Having a lot of cultural capital at one’s disposal increases the likelihood that one is recipient and also supplier of music recommendations. Being socialized into popular music also has a positive impact on all indicators, though most of that disappears when the connectivity and market involvement variables are included in the Models 4b and 4c. More omnivorous music consumers are less likely to give recommendations. Similar to the results for finding new music, I find that the influence of connectivity mainly runs via the self-assessed connector status. Individuals who think they initiate more connections receive and provide more recommendations and score higher on the mavenism scale. There is a consistent negative influence of taste difference with friends. Finally, there is a consistently strong impact of consulting UGC on all three indicators.

Table 5.

Regression of giving and receiving information (N = 658).

	Receive recommendations			Give recommendations			Mavenism
	Model 1a	Model 1b	Model 1c	Model 2a	Model 2b	Model 2c	Model 3a	Model 3b	Model 3c
Female	−.152	−.241	−.144	−.103	−.254	−.167	−.174∼	−.138	−.084
Age	−.024***	−.022**	−.016*	−.021**	−.018**	−.010	−.016***	−.016***	−.010**
Educational level	−.033	−.046	−.054	−.119*	−.107∼	−.110∼	−.058	−.046	−.048
Migrant origin	.226	.309	.359	.323	.415∼	.469*	.243∼	.258∼	.282*
Music listening frequency	1.219***	1.408***	1.281***	2.007***	2.088***	1.986***	1.162***	1.159***	1.053***
Dispositions
Cultural capital	2.857***	2.299***	1.668***	2.500***	1.962***	1.290**	1.618***	1.419***	.969***
Pop music part parents	1.490***	1.183**	.929*	1.423***	1.104**	.806*	.732**	.669**	.482*
Omnivorous music taste	−.251	−.014	.010	−.789**	−.686**	−.694**	−.255∼	−.236	−.232
General connectivity
Network size		.872**	.930**		.478∼	.501∼		.221	.238
Friends do not like music		.556	.556		−.302	−.439		.026	−.040
Friends have different taste		−.529*	−.509*		−.645**	−.619*		−.441**	−.416**
Friends have partly same taste		Ref.	Ref.		Ref.	Ref.		Ref.	Ref.
Friends have same taste		−.321	−.265		−.185	−.138		−.132	−.106
Self-ascribed connector		1.077**	1.095**		.973**	.906*		.795***	.728**
Online social activities
Use Facebook and Twitter			−.170			.088			−.085
Producing UGC			−.033			.701			.412
Consulting UGC			2.386***			2.218**			1.390***
Constant							1.307***	.884*	.578∼
Threshold cat.1	−.291	.386	.777	.079	.313	.858
Threshold cat.2	2.717***	3.560***	4.039***	2.904***	3.355***	4.004***
Adj. R²							.205	.233	.260
Nagelkerke R²	.234	.255	.281	.246	.254	.286

Receive (0, 1, 2)/give (0, 1, 2): ordinal regression; maven (0–6): OLS regression; test of parallel lines, Model 1a: Chi² = 27.197, p = .001; Model 1b: chi² = 25.870, p = .018; Model 1c: Chi² = 33.631, p = .006. Model 2a: Chi² = 34.068, p = .000; Model 2b: chi² = 39.680, p = .000; Model 2c: Chi² = 45.510, p = .000. Significance levels: ∼p<0.10 *p<.05 **p<.01 ***p<.000.

Conclusion and discussion

The study of how products spread through discovery and influence has not received much attention in the study of cultural consumption. However, the rise of digital technologies and the declining legitimacy of traditional mediators make this a relevant question for the study of cultural markets. This article examines how individuals orient themselves on one particular cultural market (i.e. finding new music) and whether indeed connectivity is becoming more important at the expense of traditional markers of distinction (dispositions). The empirical research partly replicates, but also extends, the mid-2000s studies by Tepper and Hargittai (2009) and Sun et al. (2006) on how students find music in the digital age. Not only do I offer more fine-grained measures of dispositions and connectivity, the data are also representative for the Dutch population. The main findings can be summarized as follows.

First, music discovery in the mid-2010s can be characterized by three types of practices. One group of consumers combines individual contacts (online and offline) with listening via online sources such as Spotify and YouTube. Particularly, younger listeners adopt this strategy. Another group sticks to traditional expertise, such as record stores, print media, exchange of physical products, and extended with information on websites. This group is slightly older and less often uses social media. Despite these differences, both ways of finding music are otherwise influenced by very similar characteristics: large cultural capital, being a connector, and consuming as well as producing UGC. In contrast, the last practice of finding music – which involves using mass media – is employed mostly by individuals with less cultural capital but a more omnivorous music taste.

Second, the results show that about two-third of the music listeners is involved in music recommendation practices; most consumers both offer and receive advice, and the extent to which they do so is reciprocal: the more people recommend, the more they receive recommendations from others. Dispositions – via cultural capital and, to a lesser extent, parental music socialization – remain highly influential in explaining not only who recommends music but also who receives recommendations. Being connected is also relevant – not so much in size, but mainly in two other ways: (a) consumers who consider themselves to be a connector are more active and (b) consumers who report having friends with different taste do less. Remarkably, recommending and being recommended is strongly associated with consulting UGC, but not with producing UGC or using Facebook and Twitter. It seems that music consumers who are involved in the circulation of info use UGC mainly to inform themselves about new trends. Also for mavens – who are generally conceived to be eager to communicate information – I did not find an association with generating and spreading content online.

So what can be said about mavens and cultural influence in the digital age? Successful influencers – mavens – differ only slightly from these listeners who are highly involved in exchanging music tips (receive and give). At the same time, they are not more omnivorous than other consumers (as would be expected from individuals with a large knowledge of the market) and are not involved in producing UGC (which is an important way for mavens to spread info to others, according to the literature, see Walsh and Mitchell, 2010). This puts doubt on the myth of the music maven: I do not find much evidence that these consumers are a distinct type of music consumer. Rather, there seems to be a circuit of music aficionados in which there is a gradual level of engagement – very much in line with cultural participation studies (e.g. Bennett et al., 2009; Roose et al., 2012). These are the ones involved in WOM. While individuals influencing others have similar taste as their friends, this does not rule out the possibility that they learn about other types of music via “weak ties.” However, they consider themselves a strong connector to the same degree as those receiving many recommendations, and there is not much evidence that they have more online “weak ties” (more online social activities could have been an indication for that, but no differences were found).

What are the implications for cultural markets? There appears to be a partitioning between the engaged and the disengaged in how culture is discovered and spread: mass media remain important for those without much cultural capital and little online presence, but the highly involved music consumers have a broader array of sources they consult and consider themselves connectors. It is very likely that online discovery will become more important with the increasing popularity of streaming services (Kjus, 2016), but it is not certain that the declining reliance on traditional expertise also means that aesthetic dispositions, and someone’s social background, lose importance. Cultural capital and, to a certain degree, parental socialization influence engagement more than omnivorousness. Note that this study did not address what people recommend. The negative effect of cultural omnivorousness on giving recommendations suggests that consumers mostly recommend one type of music. While this needs more investigation, this result seems to correspond with recent studies that increasingly question the usefulness of the cultural omnivore concept for explaining cultural consumption (Prieur and Savage, 2013: 263).

Obviously, this study has some limitations. An important point of discussion concerns the measurement of mavenism. The survey contained relatively simple ordinal measures to probe receiving and giving information. Contrary to psychological studies of mavens, I asked for concrete behavior, which forces respondents to move beyond self-profiling. Future studies could opt for more detailed measures of mavenism, such as the scales used by Goldsmith and collaborators (Goldsmith and Clark, 2008; Goldsmith et al., 2006), although these seem more fit for special interest populations. I recall that this survey targeted a general population – including people with a moderate interest in music. Because of this, also no specific questions on online recommendation were asked – only more general activities as indicators of eWOM. Still, the results of our questions on finding music confirm that online and offline discovery and recommendation are increasingly intertwined (Nowak, 2016). Thus, follow-up research could employ more nuanced measurements of various forms of media usage. For example, YouTube is a channel where one can explore new music videos, but is also a platform – just like Facebook, Twitter, and Instagram – where many social media influencers are active (Carter, 2016). How this phenomenon plays out in cultural markets needs more investigation.

To summarize, this study shows that there is a group of music consumers who actively search for and recommend music. These individuals have high cultural capital and yet are also well connected with friends who have similar music tastes. Thus, there is little evidence that dispositions have lost influence to connectivity. Importantly, online and offline ways of music involvement are increasingly mixed, but how these pathways are used differs between generations. Whereas older people still rely a lot on traditional expertise, younger people use more interpersonal contacts. In that sense, it is expected that connectivity will become more important in the future.

Footnotes

Declaration of Conflicting Interests

The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author received no financial support for the research, authorship, and/or publication of this article.

Marc Verboord is associate professor in the department of Media and Communication at Erasmus University Rotterdam and member of ERMeCC (Erasmus Research Centre in Media, Communication and Culture). His research focuses on cultural consumption patterns, cultural globalization, classification of cultural products, and the impact of new media on the social valuation of cultural products.

References

Aguiar

Waldfogel

(2016) Even the losers get lucky sometimes: New products and the evolution of music quality since Napster. Information Economics and Policy 34: 1–15.

Airoldi

Beraldo

Gandini

(2016) Follow the algorithm: An exploratory investigation of music on YouTube. Poetics 57: 1–13.

Baym

(2015) Personal Connection in the Digital Age, Cambridge: Polity Press.

Beer

Burrows

(2010) Consumption, presumption and participatory web cultures. Journal of Consumer Culture 10(1): 3–12.

Bennett

Savage

Silva

et al.(2009) Culture, Class, Distinction, London: Routledge.

Boster

Kotowski

Andrews

et al.(2011) Identifying influence: Development and validation of the connectivity, persuasiveness, and maven scales. Journal of Communication 61(1): 178–196.

Bourdieu

(1984) Distinction: A Social Critique of the Judgment of Taste, London: Routledge and Kegan Paul.

Burt

(1999) The social capital of opinion leaders. Annals of the American Academy of Political and Social Sciences 566(1): 37–54.

Carter

(2016) Hustle and brand: The sociotechnical shaping of influence. Social Media + Society 2(3): 1–12.

10.

Caves

(2000) Creative Industries: Contracts between Art and Commerce, Cambridge, MA: Harvard University Press.

11.

Daenekindt

Roose

(2014) Ways of preferring: Distinction to the “what” and the “how” of cultural consumption. Journal of Consumer Culture 17(1): 25–45.

12.

DiMaggio

(1987) Classification in art. American Sociological Review 52(4): 440–455.

13.

Erickson

(1996) Culture, class, and connections. American Journal of Sociology 102(1): 217–251.

14.

Feick

Price

(1987) The market maven: A diffuser of marketplace information. Journal of Marketing 51(1): 83–97.

15.

Field

(2013) Discovering Statistics Using IBM SPSS Statistics, Los Angeles, CA: Sage.

16.

Flynn

Goldsmith

Eastman

(1996) Opinion leaders and opinion seekers: Two new measurement scales. Journal of the Academy of Marketing Science 24(2): 137–147.

17.

Goldberg

(2011) Mapping shared understandings using relational class analysis. American Journal of Sociology 116(5): 1397–1436.

18.

Goldsmith

Clark

(2008) An analysis of factors affecting fashion opinion leadership and fashion opinion seeking. Journal of Fashion Marketing and Management 12(3): 308–322.

19.

Goldsmith

Clark

Goldsmith

(2006) Extending the psychological profile of market mavenism. Journal of Consumer Behaviour 5(5): 411–419.

20.

Granovetter

(1973) The strength of weak ties. American Journal of Sociology 78: 1360–1380.

21.

Hennig-Thurau

Gwinner

Walsh

et al.(2004) Electronic word-of-mouth via consumer-opinion platforms. Journal of Interactive Marketing 18(1): 38–52.

22.

Holbrook

Addis

(2008) Art versus commerce in the movie industry: A two-path model of motion-picture success. Journal of Cultural Economics 32: 87–107.

23.

Jaeger

Katz-Gerro

(2010) The rise of the eclectic cultural consumer in Denmark, 1964-2004. The Sociological Quarterly 51: 460–483.

24.

Janssen S and Verboord M (2015) Cultural Mediators and Gatekeepers. In James D. Wright (eds) International Encyclopedia of the Social & Behavioral Sciences, Oxford: Elsevier, pp. 440–446.

25.

Jarness

(2015) Modes of consumption: From “what” to “how” in cultural stratification research. Poetics 53: 65–79.

26.

John

(2013) The social logics of sharing. The Communication Review 16: 113–131.

27.

Kane

(2004) A network approach to the puzzle of women’s cultural participation. Poetics 32(2): 105–127.

28.

Kang

J-W

Namkung

(2016) Restaurant information sharing on social network sites: Do network externalities matter? Journal of Hospitality & Tourism Research 40(6): 739–763.

29.

Katz

Lazarsfeld

(1955) Personal Influence, New York: Free Press.

30.

Kjus

(2016) Musical exploration via streaming services: The Norwegian experience. Popular Communication 14(3): 127–136.

31.

Kristensen

From

(2015) From ivory tower to cross-media personas. The heterogeneous cultural critic in the media. Journalism Practice 9(6): 853–871.

32.

Leguina

Arancibia-Carvajal

Widdop

et al.(2017) Musical preferences and technologies: Contemporary material and symbolic distinctions criticized. Journal of Consumer Culture 17(2): 242–264.

33.

Lizardo

(2006) How cultural tastes shape personal networks. American Sociological Review 71(5): 778–807.

34.

Lizardo

(2016) Why “cultural matters” matter: Culture talk as the mobilization of cultural capital in interaction. Poetics 58: 1–17.

35.

López-Sintas

Cebollada Filimon

ÀN

Gharhaman

(2014) Music access patterns. Poetics 46: 56–74.

36.

McCourt

Zuberi

(2016) Music and discovery (Special Issue). Popular Communication 14: 123–126.

37.

McLean

(2017) Culture in Networks, Cambridge: Polity Press.

38.

Mark

(2003) Culture and competition: Homophily and distancing explanations for cultural niches. American Sociological Review 68(3): 319–345.

39.

Michael

(2015) It’s not really hip to be a hipster: Negotiating trends and authenticity in the cultural field. Journal of Consumer Culture 15(2): 163–182.

40.

Nowak

(2016) Consuming Music in the Digital Age: Technologies, Roles and Everyday Life, London: Palgrave Macmillan.

41.

Ordanini

Nunes

(2016) From fewer blockbusters by more superstars to more blockbusters by fewer superstars. International Journal of Research in Marketing 33(2): 297–313.

42.

Peterson

(2005) Problems in comparative research: The example of omnivorousness. Poetics 33(5–6): 257–282.

43.

Prieur

Savage

(2013) Emerging forms of cultural capital. European Societies 15(2): 246–267.

44.

Rogers

(2003) Diffusion of Innovations, 5th edn, New York: Free Press.

45.

Roose

Van Eijck

Lievens

(2012) Culture of distinction or culture of openness? Poetics 40(6): 491–513.

46.

Rössel

Pape

(2016) Who has a wine-identity? Consumption practices between distinction and democratization. Journal of Consumer Culture 16(2): 614–632.

47.

Ruvio

Shoham

(2007) Innovativeness, exploratory behavior, market mavenship, and opinion leadership. Psychology & Marketing 24(8): 703–722.

48.

Smith Maguire

Matthews

(2012) Are we all cultural intermediaries now? European Journal of Cultural Studies 15(5): 551–562.

49.

Sun

Young

et al.(2006) Online word-of-mouth (or mouse). Journal of Computer-Mediated Communication 11(4): 1104–1127.

50.

Tepper

Hargittai

(2009) Pathways to music exploration in a digital age. Poetics 37(3): 227–249.

51.

Van Dijck

(2013) The Culture of Connectivity, New York: Oxford University Press.

52.

Van Eijck

(2001) Social differentiation in musical taste patterns. Social Forces 79(3): 1163–1185.

53.

Verboord M (2010) The legitimacy of book critics in the age of the Internet and omnivorousness: Expert critics, Internet critics and peer critics in Flanders and the Netherlands. European Sociological Review 26(6): 623–637.

54.

Walsh

Mitchell

V-W

(2010) Identifying, segmenting and profiling online communicators in an internet music context. International Journal of Internet Marketing and Advertising 6(1): 41–64.

55.

Walsh

Gwinner

Swanson

(2004) What makes mavens tick? Exploring the motives of market mavens’ initiation of information diffusion. Journal of Consumer Marketing 21(2–3): 109–122.

56.

Warde

(2014) After taste: Culture consumption and theories of practice. Journal of Consumer Culture 14(3): 279–303.

57.

Weimann

(2015) Twostep flow of communication. In: Wright

(ed) International Encyclopedia of Social Sciences, 2nd edn (nr. 4). London: Elsevier, pp. 2316–2319.

58.

Wejnert

(2002) Integrating models of diffusion of innovations. Annual Review of Sociology 28: 297–326.

59.

Willekens

Lievens

(2014) Family (and) culture: The effect of cultural capital within the family on the cultural participation of adolescents. Poetics 42: 98–113.

60.

Yaish

Katz-Gerro

(2012) Disentangling “cultural capital.”. European Sociological Review 28(2): 169–185.