Music, possession and the immaterial mediamorphosis – Jan Hemming (Music Institute, University of Kassel, Germany)
Many things need to be mentioned when talking about the music industry: collecting societies, copyright, digital platforms, ownership, but this last concept needs to be challenged. For this, it is necessary to talk about the concept of mediamorphosis. Mediamorphosis affects conception of copyright. There are six phases of it. Graphic mediamorphosis through writing (first) and printing (second). Chemical-mechanical, with the invention of photography (third). Electronic, with the invention of radio (fourth). Digital, with the introduction of CD (fifth). Here is importan to note that the industry starts investing big energy towards the enforcement of copyright.
The sixth and present phase of mediamorphosis is the immaterial one. Its biggest manifestation in the music industry is the access to it, any time, for free. Thus, we don’t need to bother about possessing music anymore. This implies massive challenges to the understanding of the value of music itself that were up to some point forecasted by Firth back in the 80s, but need further research and awareness.
What if Big Data are Wrong?
– Franco Fabbri (University of Turin, Italy)
Code words, like genre, style, mode and equivalent, have been used for centuries to categorise music. One of the purposes of such taxonomies has been to devise norms of connecting the ways on how music is made and its meaning and social connections. These taxonomies have great impact in how people consume music. In the digital world, these taxonomies are replaced by tags and they are widely used for exactly the same physical purposes. Big enterprises like Amazon, Apple, Spotify, etc., use them to suggest new music and increase sales/streamings. Social sciences use these trends to analyse deeper variables between consumers and products.
Recently, Google published an article pretending to explain the history of popularity of popular music genres based on their consumption. This article signals many errors. The nature of digital tags makes them old, unefficient and static. Therefore, more inter-disciplinary research (Computer science, semiotics, linguistics, sociology and anthropology) needs to be done to fix this, because as it is, the big data, is wrong.
The Austrian sociologist Kurt Blaukopf (1989) first introduced the notion of mediamorphosis to metaphorically describe the influence of technology on social and economic developments. Various authors have subsequently shown the usefulness of this concept to historically trace the development of the music industry. For example, Smudits (2002) distinguishes five stages of mediamorphosis (1) graphical mediamorphosis: The invention of writing; (2) repro-graphical mediamorphosis: The invention of printing / offset-printing which allowed for mass dissemination; (3) chemical- mechanical mediamorphosis: The invention of photography and sound recording; (4) electronic mediamorphosis: Radio, amplification and studio recording; (5) digital mediamorphosis: CD, DAT, lossless copying, illegal and commercial music distribution over the Internet. All these stages rely on material artefacts which can be purchased, owned, stored and collected and thus foster music’s status as commodity. However, most recent changes in music distribution such as streaming finally bring the status of commodity to a close. It remains disputed if streaming means that a music file (or a part of it) is at least temporarily present on a computer. With regard to these developments, I wish to add a sixth, immaterial mediamorphosis and argue for the overall usefulness of this concept for popular music studies.
Franco Fabbri: Millions of ‘songs’ (including classical music pieces) have been tagged in the past years, providing information for web surfers and consumers, as well as for researchers. The very amount of data is promising: a number of studies were made, and stunning results were disseminated (like: ‘jazz is the most popular music from the 1950s’!). But what if data are wrong? Or more or less approximate? Among many sources of errors, the most evident are:
1) The incoherence of tagging systems;
2) Insufficient room for significant descriptions of musical content;
3) Limited and
4) Biased (especially ethnocentric) categories.
Most tags are user-generated, and if a user describes a piece of Greek rembetiko as ‘blues’ there is nothing other users (or researchers) can do. For popular music scholars, the existence of such an amount of unreliable data is tantalizing; as it is obviously impossible to restart tagging from scratch, methods have to be provided in order to correct existing tags, via automatic error-detecting algorithms, or by means of Wiki-like editing. The paper (based on an application for an ERC grant) will present a description of the state of the art on the subject, and possible solutions