PDF Publication Title:
Text from PDF Page: 009
G Model POETIC 1239 No. of Pages 13 8 M. Airoldi et al. / Poetics xxx (2015) xxx–xxx the title of each video. Most of the tightly connected groups of videos detected by our network analysis refer to conventional music genres (e.g. ‘Polka’), local music scenes (e.g. ‘Grime/Uk Hip Hop’) and generational music preferences (e.g. ‘Rock 90s’). Yet, some of these clusters also seem to point at different semantic dimensions that depart from conventional definitions of music genre (e.g. ‘Music for Babies’). 4.2. Music clusters as crowd-generated music categories In order to analyse this discrepancy between cluster types in more detail, we compared the musical content in each cluster by coding the ten most frequent relevant words in the title of the videos. This way, we scrutinize the underlying patterns of semantic association between different musical contents that are implicitly emerging from the network structure and stem from users’ aggregated practices. Building on Webb’s idea of milieu (2007), that encapsulates geographical, cultural and social aspects of music scenes, we coded with the tag ‘milieu’ all words explicitly referring to a local context of music production (e.g. ‘South Sudan music’, ‘Irish music’). With the tag ‘reception’ we coded all words that refer to the context of music reception (e.g. ‘sleep’, ‘relax’). The tag ‘genre’ was used for words pointing to conventional music genres, while ‘cross- genre’ classified common words generically referring to the features of a music video (e.g. ‘live’, ‘remix’, ‘instrumental’). Furthermore, the tag ‘venue/radio/label’ indicates words that relate to the world of music making (music venues, radio shows, music labels), and the tag ‘Youtuber/other media’ refers to the so-called ‘YouTube stars’ (cf. musicians and video- makers who are popular because of their videos on YouTube) and media platforms such as iTunes or Spotify. The remaining categories (artist, song/album/tour) are self-evident. There is evidence that clusters differ considerably in relation to the relative frequency of the categories of words described above. We see that several clusters are characterised by an extremely high rate of videos containing artists’ names and conventional genre names, and relatively few common names that are related to the musical content, while many other clusters display an opposite trend. For instance, in ‘Ambient/Chillout’, ‘Jazz/Classical’, ‘House/Lounge’ and ‘Jazz/Bebop’ more than 90% of the video titles share the same references to genres and artists, whereas less than 50% include generic words referring to the content of the videos. At the same time, more than 90% of the videos belonging to clusters such as ‘Sounds of Nature’, ‘Hair Dryer Sound’, ‘Glee Music’ and ‘Best Songs/Top Hits’ are described by ‘cross-genre’ words in the titles, and fewer than 25% by shared references to specific artists and music genres. We therefore observe an analytical polarity in the axis between those clusters mainly connoted for ‘genre’ and those mainly connoted by ‘cross-genre’. Another similar polarity emerges if we pay attention to how more than 60% of the videos included in ‘Ugandan Music’, ‘Pop Stars Interviews’, ‘Grime/ UK Hip Hop’ and ‘Irish Music’ are discursively framed via a reference to a common cultural milieu (Webb, 2007) and/or to specific music venues, radios or music labels (which are key elements in any field of music production, see Bourdieu, 1993). These very same clusters stand out by the absence of references to the contexts of music reception. On the contrary, for clusters in which the category ‘reception’ is significantly important – such as in ‘Guitar Tutorial/For Musicians’ (21.8%), ‘Relaxing Background Music’ (60.6%), ‘Hair Dryer Sound’ (100%), ‘Sounds of Nature’ (100%), ‘Music for Babies’ (100%) – the categories ‘milieu’ and ‘venue/radio/label’ never appear. A second semantic axis is thus established based on the prominence of production or reception-related coding. See Table 1 for examples of video titles per cluster. In order to better interpret and compare the positioning of each cluster according to these two dimensions – and following Middleton’s suggestion to “locate musical categories topographically” (1990:7) – we constructed a Cartesian graph and assigned x and y coordinates to each cluster. The x-axis corresponds to the ‘genre’/‘cross-genre’ continuum. For each cluster we computed the offset between the relative frequency of cross-genre labels on the one hand, and artist and genre words on the other. We aggregated ‘artist’ and ‘genre’ labels because both semantically refer to the actual musical content of the video, implicitly or explicitly qualified in terms of genre. The y-axis analogously refers to the offset between the percentage of words in the title related to the context of reception and those related to the context of production. Here, we aggregated the categories ‘milieu’ and ‘venue/radio/label’ for purposes similar to those above (Fig. 2).9 On the left side of the map, among those clusters sharing genre-identity of sorts (negative x-values), one can see the opposition between the negative y-values of clusters close to classical definitions of music genre (e.g. ‘Ugandan Music’, ‘Grime/UK Hip Hop’, ‘Irish Music’, ‘Ethiopia/South Sudan Music’, ‘K-Pop’, ‘Celtic Music’, ‘Jazz/Bebop’) and the positive y- values of clusters closer to newer genres (such as ‘House/Lounge’, ‘Ambient/Chillout’, ‘Pop Hits’, ‘Rock 90s’, ‘80s Pop/Rock’). Whereas the former can be described as a “constellation of styles connected by a sense of tradition”, the second show resemblances to “marketing categories” instead (see Holt, 2007: 18). On the right side, (x > 0; ‘cross-genre’ polarity), conventional genres are replaced by less common music categories, such as ‘Sound of Nature’, ‘Hair Dryer Sound’, ‘Best Songs/Top Hits’, ‘Glee Music’, ‘YouTube Stars’ and ‘Flute/Piano Cover/Tutorial’. All these clusters refer to videos explicitly sharing cross-genre commonalities, such as the ‘relaxing’ tones of a thunderstorm, the sound of a fan or a hairdryer, merely being a part of a ‘top 10 songs’ ranking or of the soundtrack of a television series such as ‘Glee’, the strong presence of online user-generated contents (van Dijck, 2009) and/or of specific musical instruments. The highest y-values which denote the discursive prevalence of the contexts of music reception over those of music production are associated with ‘Music for Babies’ (y = 3), ‘Sounds of Nature’ (y = 2.7), ‘Hair Dryer Sound’ (y = 2.7), ‘Relaxing 9 These are the adopted formulas: x = f cross-genre (f artist + genre); y = f reception (f milieu + f venue/radio/label). Values on the map are normalised according to axes’ means and standard deviations. Please cite this article in press as: M. Airoldi, et al., Follow the algorithm: An exploratory investigation of music on YouTube, Poetics (2016), http://dx.doi.org/10.1016/j.poetic.2016.05.001PDF Image | Follow the algorithm: An exploratory investigation of music on YouTube
PDF Search Title:
Follow the algorithm: An exploratory investigation of music on YouTubeOriginal File Name Searched:
AiroldiBeraldoGandini2016Preprint.pdfDIY PDF Search: Google It | Yahoo | Bing
Cruise Ship Reviews | Luxury Resort | Jet | Yacht | and Travel Tech More Info
Cruising Review Topics and Articles More Info
Software based on Filemaker for the travel industry More Info
The Burgenstock Resort: Reviews on CruisingReview website... More Info
Resort Reviews: World Class resorts... More Info
The Riffelalp Resort: Reviews on CruisingReview website... More Info
CONTACT TEL: 608-238-6001 Email: greg@cruisingreview.com (Standard Web Page)