Political ecology and recorded music

Political ecology is multifaceted and difficult to summarise (cf. Hayward Reference Hayward1994; Robbins Reference Robbins2012; Neumann Reference Neumann2014). At a basic level, though, political ecology is defined by critical attention to the principles of action and the forms of social order that link material environments and human cultures. Political ecologists demonstrate a particular interest in the environmental stresses that result from resource extraction and processing, as well as product manufacturing, consumption and disposal. From this perspective, a political ecology of music would study how the stuff of musical culture is made and possessed, dispossessed and unmade.

Although a political–ecological approach could be applied to any aspect of music's material culture – such as sheet music, instrument manufacturing, concert-going – my particular interest is in recordings.Footnote ¹ Just as political ecology is heterogeneous, so are there numerous existing ways of thinking about recordings as items in the world: as commodities, technologies, media, formats, objects, emblems, things, etc. (e.g. Kenney Reference Kenney1999; Kittler Reference Kittler1999; Straw Reference Straw1999–2000; Maisonneuve Reference Maisonneuve2001; Born Reference Born2005; Hennion Reference Hennion2007; Sterne Reference Sterne2012). There are also plenty of ways to think about how such items come to be and move through musical culture: production, consumption, distribution, reception, exchange, attachment, mediation, migration, circulation, and so on (e.g. Jones Reference Jones2002; Gaonkar and Povinelli Reference Gaonkar and Povinelli2003; Ochoa and Botero Reference Ochoa and Botero2009; Straw Reference Straw, Crow, Longford and Sawchuk2010; Toynbee and Dueck Reference Toynbee and Dueck2011; Born Reference Born, Clayton, Herbert and Middleton2012; Novak Reference Novak2013; Santoro Reference Santoro, Shepherd and Devine2015). What is more, there are various frameworks for conceptualising the connections that are enacted between people and materiality in the creation and circulation of cultural artefacts more generally: topology, actor-network-theory, posthumanism, thing theory, a variety of new materialisms, and many other ideas (e.g. Thompson Reference Thompson1979; Straw Reference Straw1999; Brown Reference Brown2001; Pels et al. Reference Pels, Hetherington and Vandenberghe2002; Barad Reference Barad2003; Latour Reference Latour2005; Miller Reference Miller2005; Appadurai Reference Appadurai2006; Morton Reference Morton2007; Suchman Reference Suchman2007; Curry Reference Curry2008; Bennett Reference Bennett2010; Born Reference Born2010; Coole and Frost Reference Coole and Frost2010; Ingold Reference Ingold2012; Lury et al. Reference Lury, Parisi and Terranova2012; Barry Reference Barry2013; Sterne Reference Sterne, Gillespie, Boczkowski and Foot2013). In combining aspects of many of these perspectives, the goal here is to highlight the co-formative associations of biology, geology, capital and culture that animate the social life (Appadurai Reference Appadurai1986) and social death (Gabrys Reference Gabrys2011) of recorded music artefacts. Building on both this diverse range of literature and the above definition of political ecology, it is possible to offer a more precise conception of a political–ecological approach to music.

On the side of making and possession, a political ecology of music has an interest in the movement and processing of materials that are required before a recording can be bought or sold, created or listened to. There is a whole economy of raw materials and supply chains that undergirds what is traditionally called the recording industry. This economy is able to mobilise, synchronise and aggregate massive global contingents of people and materials, intertwining numerous national governments, local economies and environments. While such processes and materials may seem peripheral, they are actually central to what the recording industry is and how it works.

On the side of unmaking and dispossession, political ecology is interested in how even our favourite recordings, like all good things, must come to an end. That is, recordings eventually enter into circuits of dispossession (cf. Lucas Reference Lucas2002). Here, one mode of accumulation (as an expression of fandom) gives way to another (as a measure of disinterest): where we once collected records, our records now collect dust. And when the value of billions of such recordings is truly exhausted (Straw Reference Straw2000), the question becomes: where do they go? Political ecology is here interested in waste, in the sense of both the ravages of production and the remnants of consumption.

Thus, the issue for the political ecologist is more than who is making and listening to music, or buying and selling it – or even what that music sounds like.Footnote ² Political ecology is also about where the raw materials of musical artefacts come from and how they are processed and manufactured, as well as how the artefacts are treated and where they go when we are done with them.

The potential contribution of a political ecology of music seems twofold. Analytically, for music studies, the approach offers an expanded conception of musical culture. It highlights more fully the global reach and power of the music industries. It asks how the aesthetic and industrial circulations of musical culture hinge on the material and economic flows of resource networks. It considers dispossession not merely as an afterthought or an adjunct to ironically linear models of the product lifecycle (cf. Gregson and Crang Reference Gregson and Crang2010; Lepawsky and Mather Reference Lepawsky and Mather2010) but, rather, as part of a topology of musical materiality in which the question of disposal is immanent in desire and manufacturing, in which ‘afterlife … is anticipated before exchange’ (Hawkins Reference Hawkins, Gabrys, Hawkins and Michael2013, p. 51). In other words, if research into consumption and possession has revealed much about music as an agent in social life (e.g. Peterson Reference Peterson1992; DeNora Reference DeNora2000; Bull Reference Bull2007; Savage and Gayo Reference Savage and Gayo2011; cf. Bourdieu Reference Bourdieu1984; Bennett et al. Reference Bennett, Savage, Silva, Warde, Gayo-Cal and Wright2009), then so might rituals of disposal and dispossession reveal various social logics and inform our understanding of musical culture (e.g. Straw Reference Straw1999–2000; cf. Strasser Reference Strasser1999; Lucas Reference Lucas2002; Hetherington Reference Hetherington2004; Hawkins Reference Hawkins2006). Together, these ways of looking at musical culture call for ‘a change in focus, from the “objectness” of things to the material flows and formative processes wherein they come into being’ (Ingold Reference Ingold2012, p. 431) – and wherein they cease to be. A simple way of thinking about a political–ecological approach is as an expanded and radically material ‘art world’ (Becker Reference Becker1982): if coffee porters matter in the world of literature, then so do farmers, chemical engineers and waste workers in the world of music.

Implicit in the above is a second potential contribution of political ecology: a relatively new avenue of critical social engagement for music scholarship. Policy is an established and, it seems, increasingly important issue in popular music studies (Cloonan Reference Cloonan2007; Stanbridge Reference Stanbridge2007; Frith Reference Frith, Brandl, Prévost-Thomas and Ravet2012; Street Reference Street2013; Behr Reference Behr, Shepherd and Devine2015). Such work rightly focuses on difficult questions of cultural ‘value’ and attendant arts funding issues in the face of the continued hardening of neoliberal economic philosophies. A radically material approach to music may add to the political relevance of music studies by spurring scholars to look at the place where music's economic and cultural significances encounter its environmental consequences, thus opening the door to critical interventions at the intersection of music, media and environmental policy (cf. Bottrill et al. Reference Bottrill, Lye, Boykoff and Liverman2008). The issue, then, is one of making us more aware as scholars, musicians and listeners – and thus more responsible, too.

In these ways, a political ecology of music is somewhat different from other existing ecologies of music. It is not the naturalistic and aesthetic environmentalism that defines the acoustic ecology of R. Murray Schafer (Reference Schafer1994) and Barry Truax (Reference Truax2001). Nor is it the perceptual–ecological framework that defines the social psychologies of Eric Clarke (Reference Clarke2005) and Tia DeNora (Reference DeNora, Clarke and Clarke2011) – although there is much to be learned from such work. The politics of political ecology are most closely aligned with ecomusicology (Rehding Reference Rehding2002; Grimley Reference Grimley2006; Guy Reference Guy2009; Ingram Reference Ingram2010; Allen et al. Reference Allen, Grimley, Von Glahn, Watkins and Rehding2011; Echard Reference Echard2011; Pedelty Reference Pedelty2012; Allen Reference Allen2013). Although my relative inattention to the musical imagination and its semiotics of landscape seems to differentiate my approach from the hermeneutic ecology that is central to much mainstream ecomusicology, the field does demonstrate an emerging concern for resources and sustainability as well as the kind of reconstructed materialism at the conceptual base of this article (e.g. Ingram Reference Ingram2007; Allen et al. Reference Allen, Grimley, Von Glahn, Watkins and Rehding2011, Reference Allen, Titon and Von Glahn2014; Allen Reference Allen, Auricchio, Cook and Pacini2012; Pedelty Reference Pedelty2012, pp. 17–36; Allen and Dawe Reference Allen and Dawe2015). While I thus assume that a political ecology of music is a partner to ecomusicology, I am also attempting to extend some of that field's existing methodological inclinations. In doing so, this project draws on literature that complements a longstanding emphasis on what happens when old media were new (Marvin Reference Marvin1988) by examining what happens when new media grow old (Acland Reference Acland2007). It is also closely connected to an emerging style of media ecology, which is not the study of ‘media environments’ after Marshall McLuhan and Neil Postman (cf. Lum Reference Lum2006) but, rather, the study of how the manufacture and delivery of mediated culture ‘consumes, despoils, and wastes natural resources’ (Maxwell and Miller Reference Maxwell and Miller2012, p. 1; cf. Gabrys Reference Gabrys2011; Acland Reference Acland2014; Parikka Reference Parikka2014; Maxwell et al. Reference Maxwell, Raundalen and Vestberg2014; Smith Reference Smith2015).Footnote ³ I develop such an approach to the material politics of recorded music, which we might also call a critical musicology of infrastructure (cf. Larkin Reference Larkin2013; Parks and Starosielski Reference Parks and Starosielski2015), in the hopes of responding to a challenge that has been well articulated by Jonathan Sterne: ‘A real social critique of music markets’, he says, ‘must extend beyond the usually named music industries, to broadband, consumer electronics, computers and software, and a range of other related fields’ (Sterne Reference Sterne2012, p. 218; emphasis in original).

With varying degrees of explicitness and implicitness, the rest of the article kneads its three main materials into this political–ecological framework, moving through shellac, plastic and data. The effect of organising the article in this way is not necessarily a cumulative argument. Rather, it is to cover some largely uncharted territory, or to re-cover and re-chart a territory that has been mapped in other ways, so that several themes can emerge and recur. First, the article consistently highlights industries such as natural resource extraction, chemical engineering and energy production, none of which is normally seen in music scholarship as integral to the business of making records. Indeed, whereas it has been necessary to distinguish cultural commodities from other kinds of goods and services (Miège Reference Miège1979), and to examine music's specificity within that subcategory (Straw Reference Straw1999–2000; Taylor Reference Taylor2007a; Morris Reference Morris2010), this article suggests that conceptualisations of the recorded music commodity also stand to gain from reconnecting the artefact with its baser but not exactly ‘raw’ materials (cf. Barry Reference Barry2005; Bennett Reference Bennett2010; Ingold Reference Ingold2012). Secondly, practices of production, consumption and accumulation are seen, not quite as versos, but as topological functions of waste, construed broadly to include fallout from production (Murphy Reference Murphy1994), rituals of dispossession (Lucas Reference Lucas2002), renewal through wrecking and recycling (Lepawsky and Mather Reference Lepawsky and Mather2010) as well as plain old garbage (Maxwell and Miller Reference Maxwell and Miller2012). Finally, in moving roughly chronologically from shellac to data, I do not mean to imply a teleology in which materials replace one another in a succession of improvements on the way to ‘perfect fidelity’. This article views the history of sound reproduction less in terms of the continuous progress of technology than the contingent practicalities of political ecology.

Shellac

In the early days of commercial sound reproduction, recordings came in all shapes and sizes, spun at various speeds, and were made from all kinds of stuff – from household tinfoil to high-tech Condensite, beeswax to beef tallow, celluloid to spermaceti.Footnote ⁴ As early as 1895, though, the race to mass duplicate recordings led Emile Berliner to use a composite shellac formula in his discs. The combination of the disc form and the shellac formula was preferred, not only to Thomas Edison's cylindrical phonograph recordings, which were at the time made of vulnerable wax, but also to the hard vulcanite rubber then used in pressing discs, which was prone to warping (Read and Welch Reference Rehding1976, p. 127). With the arrival of electrically powered turntables in the 1920s, the speed of recordings was standardised at 78 revolutions per minute.Footnote ⁵ In these ways, the shellac 78-rpm disc became the key material form of recorded sound for nearly half the history of commercial sound reproduction.

Shellac is a resin made from lac, the sappy secretion of a beetle native to southern and eastern Asia. Twice a year these insects swarm ‘like blood-red dust’ in the forests of India and Thailand, and twice a year the twigs they encrust with lac are harvested, crushed, washed, dried, melted and ‘made into thin sheets which are broken into the familiar flakes of the shellac of commerce’ (Walker and Steele Reference Walker and Steele1922, p. 278). Although the modern shellac trade began in the late 19th century (Anon. 1913) and although shellac was used in everything from varnish and buttons to poker chips and electronics components (Walker and Steele Reference Walker and Steele1922, p. 280), it was the popularity of the gramophone record that fuelled the explosion of the shellac industry after 1900.

The demand for shellac during these years was such that producers struggled to keep up. As a result, prices varied and the quality of the supply was uneven: ‘Often it was loaded with impurities such as grass, weeds, gravel, plain dirt, and it seemed, anything else that would add weight to the shipment’ (Isom Reference Isom1977, p. 719; cf. Parthasarathi Reference Parthasarathi, Bel, Brouwer, Das, Parthasarathi and Poitevin2005, p. 29). What is more, WWI and its aftermath saw the price of shellac jump from a prewar £50 per tonne to nearly £900 per tonne in early 1920, settling below £500 by the end of 1921 (Sturdy Reference Sturdy1920a, Reference Sturdy1920b, Reference Sturdy1921; cf. Martland Reference Martland2013, pp. 201, 271). This was partly because of scarcity due to the use of shellac in anticorrosive paint on naval ships (Walker and Steele Reference Walker and Steele1922, p. 280) and it was partly due to weak crop yields in India (Anon. 1919, 1920). The upshot of this inflation was that the record industry used lower quality shellac, less virgin shellac and less total shellac per disc, which resulted in rougher records (both in terms of acoustic definition and needle wear and tear). Still, the postwar boom in consumer demand was enough to offset the overall rise in price (Martland Reference Martland2013, pp. 210, 271, 276) and shellac maintained its centrality until about 1950. By then, shellac production in India peaked at 40 million kg per year – the majority of which went to record companies.Footnote ⁶

While that amount of resin could go a long way in the recording industry, yielding about 300 million discs, it took millions of beetles about a year to produce a single kilogram of shellac.Footnote ⁷ Given average shellac yields (Singh Reference Singh2007), that level of production would have required around 100,000 ha of land.Footnote ⁸ When we consider that billions of discs were manufactured during the half-century long shellac era, this raises questions about the environmental consequences of such activity.Footnote ⁹ The process of protecting crops, for example, can involve the use of toxic pesticides. Further, while harvesting techniques (which basically involve sawing, striking and scraping tree branches) could be ‘painstaking and selective’, they were often ‘gross and ravenous’ (Isom Reference Isom1977, p. 719). This could lead to the death of host trees. That said, the rapid lifecycle of the lac beetle is such that shellac is a renewable resource.Footnote ¹⁰ In these ways, commercial shellac cultivation and harvesting shares in many of the controversies that mark forestry and agribusiness writ large.

Key as shellac was to the political ecology of early recorded music, the resin was actually a minority ingredient in gramophone discs. The majority of these records – up to 85 per cent – consisted of filler, a mixture of minerals, fibres and lubricants in which shellac acted as a binding agent. The chief fillers were crushed slate and limestone, mixed in equal parts. Limestone was particularly important, not only because it was inexpensive (so was slate) but because it was smooth; it sounded better, flowed well under the needle. This junction between the acoustic signature of limestone and its price point is one among many instances in which the contours of the history of recorded music, as an aesthetic form, have developed in tandem with the material politics of resource networks.

The best and most plentiful limestone came from Indiana. It was partly for this reason, remembered sound engineer Rex Isom (Reference Isom1977, p. 720), that in the late 1930s RCA Victor moved a significant portion of its production facilities to Indianapolis. Columbia followed suit, and to this day places like Terre Haute, IN remain important manufacturing centres in the recording industry. Isom (Reference Isom1977) noted that ‘the advantages of being near the source of the best filler … showed up larger than was expected on the bottom line of the balance sheet’. But the RCA engineer's cost–benefit analysis does not consider the ecology of limestone extraction, which is infamously destructive of landscape, habitat and water supplies. From this perspective, using ‘shellac’ as shorthand for the 78-rpm record is actually something of a misnomer: the critical ecosystems for 78s were as much the quarries of Indiana as they were the forests of India.

The various constituent materials of 78s converged, mixed and took shape in pressing plants. With their smoke stacks, assembly lines and heavy duty machinery, these plants were obvious sites of standardisation, mass production and industrialisation. Although recording was not the first instance of mass musical production (precedents can be found in the sheet music and instrument trades), there is a sense in which music had not been seen as an ‘industry’ before sound recording (Eisenberg Reference Eisenberg2005, p. 18). Some commentators were obviously impressed by record factories, praising the ‘forest of belting’ and ‘wonderful human-like machines’ that were ‘busily stamping, punching or pressing’ records and playback equipment ‘in musically pleasant contrast to the war-time production of shell cases, fuses, primers, ammunition cases, aeroplane parts, etc.’ (Sturdy Reference Sturdy1919, pp. 167–8). In other contexts, though, the coupling of music and industrialisation was a target of cultural critique. As Richard Osborne (Reference Osborne2012, pp. 70–1) notes, René Claire offered a cinematic comparison between the ‘prison regime and the production line in a gramophone factory’, while Theodor Adorno derided the resemblance of gramophone production to the progressive assembly of the Ford Model T. George Orwell, for his part, noticed the record industry's potential for environmental damage and complained gracefully about a ‘Ruined Farm near the His Master's Voice Gramophone Factory’. Still, what neither the proponents nor the critics of musical industrialisation consider are the less obvious ways that the recorded music industry also began as part of various natural resource industries.

The political–ecological history of the 78-rpm disc thus highlights seemingly unlikely connections between the industries of music, forestry and mining. Such relationships were not incidental; they were integral to the everyday operations of making and selling recorded music. To understand how the price and availability of raw materials influenced business decisions is to appreciate more fully how the record industry functions as an industry. What's more, to observe the import–export patterns of feedstocks such as shellac and limestone is to extend back the timeline of musical globalisation, past the debates about ‘world music’ that emerged in the 1980s. Similar points have been made by others, albeit in relation to hybrid aesthetic and identity formations (Kassabian Reference Kassabian2004; Taylor Reference Taylor2007b) and, implicitly, transnational movements of musical commodities as finished products (Gronow Reference Gronow1983). Together, these emphases have defined the study of musical globalisation as the study of what Keir Keightley (Reference Keightley, Toynbee and Dueck2011) perceptively calls ‘song networks’. But there are other factors at play. To paraphrase Arjun Appadurai (Reference Appadurai1986, p. 13), although the biographical aspect of some things (such as 78s) may be more noticeable (because they garner more symbolic investment) than that of some others (such as shellac and limestone), these other materials are no less important. Indeed, the industrial–aesthetic textures of song networks take shape in relation to the industrial–material flows of resource networks. The political–ecological viewpoint thus shifts not only the timeline of musical globalisation but also the character of the questions that can be asked about that phenomenon. In this way, the 78 era is part of a longer history of musical globalisation – a history which is not only about the movement of recordings around the Earth but also the movement of earth to make recordings.

Plastic

The dominance of the 78 format was doubly challenged in the late 1940s, with the introduction of Columbia's 33⅓ rpm ‘long play’ (LP) album and RCA-Victor's 45 rpm ‘single’. Following a drawn out ‘war of the speeds’, in which LPs and 45s jockeyed for position both between themselves and with 78s, the LP was crowned the ‘core commodity’ of the Western recorded music industry, occupying an 80 per cent market share by 1960 (Keightley Reference Keightley2004, p. 378). The core materiality of that core commodity – and of recorded music commodities for the rest of the century – was plastic.

In accounting for the shift from shellac to plastic, in particular the hazardous polyvinyl chloride (PVC) from which LPs are made, most writers point to technological advances (e.g. Kolodin Reference Kolodin1957). That is indeed one kind of explanation, as LPs were said to have ‘very quickly assumed a position of general superiority as to quieter playing surface, frequency range, and clarity’ (Read and Welch Reference Read and Welch1976, p. 324) while their ‘hair-width’ microgrooves were celebrated for their increased playback time (Gelatt Reference Gelatt1977, pp. 290–1). However, it is possible to suggest, with Osborne (Reference Osborne2012, p. 67), that the ‘downfall’ of shellac was not strictly a matter of sound quality: ‘despite the “frying-bacon” sizzle of its surface noise’, he says, ‘shellac was a format capable of withstanding continued audio improvement’ (cf. Myers Reference Myers1946). Accounting for the downfall of shellac therefore requires a more complicated explanation than straightforward technological progress. Politics and ecology are important parts of that story.

World War Two looms large here, as the Japanese occupation of key shellac-producing regions reduced supply in the West. While the Canadian and British governments instituted salvage campaigns that recycled millions of 78s (Anon. 1942a; Osborne Reference Osborne2012, p. 67), the US War Production Board instituted rationing policies that shrank non-military uses of shellac by 70 per cent (Anon. 1942b, 1942c, 1942d; cf. Chasins Reference Chasins1943). The scarcity of shellac caused prices to continue rising in the immediate postwar period, from a ‘normal’ US$14 per tonne to approximately US$45 per tonne (Myers Reference Myers1946, p. 413). Unlike the shellac situation around WWI, though, where demand offset inflation, the embargoes of WWII created a conjuncture in which the recording industry was impelled to consider other materials (Winner Reference Winner1944).

They found what they were looking for in PVC, a synthetic polymer derived from a petrochemical called ethylene, which Richard Maxwell and Toby Miller (Reference Maxwell and Miller2012, p. 59) have called ‘the most … pernicious chemical in use’ due to the health and environmental concerns associated with its manufacture and transport. Although PVC had been available since the 1930s, it was prohibitively expensive. Only in the midst and aftermath of WWII did PVC become a commercially viable material in the record industry. When the LP had firmly taken hold of the market in 1960, British record companies annually consumed 5 million kg of PVC (Scaping Reference Scaping1979, p. 136). With the LP's peak in Britain during the late 1970s, at nearly 130 million units, the amount of vinyl jumped to 22 million kg (5 per cent of the country's total yearly PVC output). Extending this to the height of worldwide sales – 942 million units worldwide in 1978 – the LP weighed in at 160 million toxic kilograms.

In sourcing all that PVC, and in refining its suitability for sound reproduction, the record industry forged new partnerships with plastics and petrochemical industries. The largest single PVC producer in the UK at that time was British Petroleum (Scaping Reference Scaping1979, p. 136); the major US suppliers were the DuPont, Union Carbide and Dow Chemical corporations, as well as diversified automotive interests such as Goodrich, Goodyear and Tenneco (Anon. 1949; Martin Reference Martin1951, p. 78; Duston Reference Duston1974, p. 3). Indeed, the LP was seen as ‘a monument to the cooperation of the record industry with the plastic industry’ (Isom Reference Isom1977, p. 723; cf. Khanna Reference Khanna1977) and developments in one field were followed closely in the other. Music industry trade papers such as Billboard covered the invention of new plastic resins, while chemical engineering trade papers such as Plastics kept a watchful eye on the record industry's need for new materials.

Both industries relied on petrochemicals and so variations in the oil market could directly impact their profit margins and production capacities.Footnote ¹¹ Such variations were tracked with interest in music journals such as Billboard (Kirsch Reference Kirsch1973; Anon. 1974; Kozak Reference Kozak1976; Traiman Reference Traiman1979). As part of a more general ‘triumph’ of plastic in the 20th century (Bensaude-Vincent and Stengers Reference Bensaude-Vincent and Stengers1996, p. 201; cf. Meikle Reference Meikle1995), this era of the recorded music commodity brought about a new system of relationships between the recording industry, chemical engineering and energy corporations – and it raises a new series of questions about the bulk and toxicity of recorded music. It is also the moment when the recording industry's ecological centre of gravity shifted from the forests of the Southeast to the oilfields of the Middle East.Footnote ¹²

Nowhere is the finality of the shift from shellac to plastic better illustrated than in a story told by the comedian Barry Humphries (Reference Humphries1992, pp. 129–33), also known as Dame Edna Everage. In the mid-1950s, Humphries took a job at EMI in Melbourne. After apprenticeships in ‘push[ing] boxes of records around the warehouse’ and ‘wandering amongst the aisles of steel shelving’, Humphries got his first big break: he was instructed to go down into the basement of EMI, to grab a hammer and destroy the company's stock of discontinued 78s – to shatter it, disc by disc, making way for microgrooves. What a waste! Surely these recordings should have been sold off to collectors, donated to museums and archives or, as Humphries himself protested, given ‘to hospitals, to old people's homes’. But they weren't. While part of this has to do with copyright legislation (deletions cannot legally be sold or given away) it also says something about the temporal logic of the recorded music commodity.

Unpacking this temporal logic actually takes us back to the beginning of sound reproduction, where the new technology offered manifestly peculiar temporal possibilities. Indeed, says Sterne (Reference Sterne2003, p. 287), ‘If there was a defining figure in early accounts of sound recording, it was the possibility of preserving the voice beyond the death of the speaker’. But the possibility of preservation actually gave rise to a tension between permanence and ephemerality – a ‘preservation paradox’ – which has ‘been a fundamental condition of recording throughout its history’ (Sterne Reference Sterne, Bijsterveld and van Dijck2008, p. 59). In the words of media historian Daniel LeMahieu (Reference LeMahieu1988, p. 88), writing about the rise of the gramophone in Britain, ‘Recorded sound transcended time, and yet in another sense, this modern, technological form of permanence contributed to the transience of music’. He explains:

These temporal dialectics gave rise to a ‘collecting impulse’ (Shuker Reference Shuker2010), which became increasingly pronounced with the rise of electrical recording and the standardisation of the 78. It was during the 1920s, for example, that American libraries started seriously collecting sound recordings (Almquist Reference Almquist1987, pp. 13–16). For individual listeners, this transition took on a particular meaning in relation to the recording's temporal logic: ‘The eclipse of acoustical recording’, says LeMahieu (Reference LeMahieu1982, p. 379), ‘accelerated the trend whereby old records were traded in, destroyed, or forgotten in favor of more current performances. By the late 1920s … the records of many opera singers from the earliest days of the gramophone became distressingly rare’.

One response to that distress came in the form of a new Gramophone column: ‘Collector's Corner’, which ‘during the 1930s defined the art of collecting historic recordings’ (LeMahieu Reference LeMahieu1982). What's more, this Gramophone column emerged alongside related commentary in venues such as Melody Maker and Down Beat, as well as a new literature in collector guides and discographies (Shuker Reference Shuker2010, pp. 25–7). Such commentary existed in relation to both classical and popular repertoires, especially jazz (Millard Reference Millard2005, p. 252), and the practice of collecting was guided not only by the logic of cultural preservation but also completism, discrimination and distinction (Shuker Reference Shuker2010, pp. 17–21).

Shellac records were thus collectible from early on; people felt attached to them, in terms of both the anxieties of heritage (cf. Maisonneuve Reference Maisonneuve2001) and the pleasures of fandom (cf. Hennion Reference Hennion2007). Yet most of the artefacts of the 78 era have disappeared. Part of this has to do with the infamous fragility of shellac discs. But delicacy alone cannot explain why, apart from a minority population of antiquarians, 78s were so readily trashed, so easily lost and so quickly forgotten.Footnote ¹³

An important aspect of that explanation stems from shifts in home economics and care for the self, which had also been emerging since the 1920s. This was a conjuncture in which new forms of convenience and new ideologies of cleanliness shaped a new ‘ethos of disposability: chewing gum, cigarette butts, razor blades, and paper products’ (Strasser Reference Strasser1999, p. 173). As cultural commodities, recordings are of course subject to different temporal logics than throwaway packaging and disposable razors (even if certain types of popular music have been scoffed at for their resemblance to bubblegum). Still, the apparent appetite for the supersession of shellac by plastic must be seen, not only as part of the preservation paradox that marks all recording, but as part of the particular culture of disposability and dispossession that was developing in the early 20th century (cf. Lucas Reference Lucas2002).

Vinyl is also part of this moment, though the two formats have broadly different afterlives. They do not meet the same ends, physically or symbolically. Whereas the cultural and economic value of obsolete shellac bottomed out and met with destruction, Will Straw (Reference Straw1999–2000, p. 162) shows that undesirable vinyl records tend to circulate almost endlessly in secondary economies where, even though their value is exhausted, they accumulate in ‘museums of failure’ in which ‘their bulk nevertheless functions almost monumentally’. In other words, shellac 78s seem predisposed to be disposed of, while vinyl LPs seem prone to pile up.

One reason for this difference has to do with the way LPs were articulated to the seriousness of adult culture and the contradictory anti-commercialism of rock. As Keightley (Reference Keightley2004, p. 386) argues, ‘long play’ refers not just to ‘the extended duration of musical playback’ but also ‘the album's ongoing cultural and economic presence’. Indeed, in contrast to the ephemerality that characterised both its predecessor (the 78) and its contemporary (the 45), the LP possessed ‘heightened symbolic capital’ and was ‘more and more perceived to occupy a cultural space similar to that of books’ (Keightley Reference Keightley2004, p. 383). This is what Keightley refers to as ‘the slower temporal logic of the LP’. While shellac and vinyl were subject to broadly similar forms of attachment, the slower temporal logic and particular accumulative tendency of the LP as a thing stem not only from its physical durability but from the meanings that accrued to it as a format. This is a curious irony, given the wider cultural history of plastic, throughout which the material has been much derided as a symbol of a cheap, artificial, throwaway society (Meikle Reference Meikle1995). Nevertheless, the symbolic and temporal logics of the plastic LP shed light on its recent and widely reported revival (cf. Bartmanski and Woodward Reference Bartmanski and Woodward2014).

Of course, some LPs do not survive. And despite the fact that the PVC in LPs may technically be recycled (Ríos Reference Ríos2011), it is generally not economically worthwhile to do so. This means that, in addition to the health and environmental risks associated with the production of PVC, when truly exhausted LPs finally die they will end up in incinerators and landfills – scenarios in which various toxins, including carcinogenic dioxins, will be released into the atmosphere and groundwater (Wittchen Reference Wittchen2012). However, given their temporal and accumulative logics, for the moment it seems that the ‘social death’ (Gabrys Reference Gabrys2011) of the LP will weigh more heavily on our shelves than the environment.Footnote ¹⁴

Data

Anxieties about the material status of music have ‘wound their way through the long history of philosophical aesthetics’ (Straw Reference Straw, Clayton, Herbert and Middleton2012, p. 229; cf. Bowman Reference Bowman1998). But such anxieties take on a distinctive shape in relation to sound reproduction, and they possess a distinctive urgency in relation to the popularisation of the MP3. Take, for example, the remarks of BBC presenter Michael Smith (Reference Smith2012), for whom the release of Beck's conspicuously old-fashioned Song Reader in Reference Smith2012, a book of sheet music, occasioned a reflection on the fate of music in the digital age:

Smith argues lovingly that vinyl LPs ‘condensed and stored [music's] spirit’, implying that digital music is stripped of its aura and somehow, therefore, colder and deader than previous forms of sound reproduction. This is a familiar argument, and a problematic one. It is problematic because Smith misses the fact that similar critiques go back to the dawn of mechanical sound reproduction (historically speaking, the LP is less ‘precious’ than he realises; cf. Sousa Reference Sousa1906). He also misses the subtlety and ambiguity with which such questions have been treated elsewhere (conceptually speaking, the loss of aura is more ambiguous than he allows; cf. Benjamin Reference Benjamin, Benjamin, Arendt and Zohn1968). Indeed, the rest of this article could be devoted to unpicking what Sterne (Reference Sterne2006, p. 338) would call the ‘dubious metaphysics’ of Smith's grievance. But such critiques are established. Instead, I want to focus on Smith's suggestion that music digitalised is music dematerialised.

Empirically, there is some truth to the equation of digitalisation and dematerialisation. While the MP3 itself is not immaterial (the scale of its materiality is invisible), MP3s do have less ‘physical presence’ than CDs (McCourt Reference McCourt2005, p. 249).Footnote ¹⁵ As Jeremy Morris (Reference Morris2013, p. 2) puts it, ‘Album art, jewel cases, and other packaging remnants have morphed into metadata, tags, software interfaces, and other less tactile forms’. For this reason, some environmental reports claim that digitalisation can reduce the ‘material intensity’ of music by as much as 80 per cent, compared to CD-oriented physical retail and e-commerce scenarios (Türk et al. Reference Türk, Alakeson, Kuhndt and Ritthoff2003, p. 34; Weber et al. Reference Weber, Koomey and Matthews2010, p. 763). But Smith's rhetoric ignores – while the scientific research brackets – two central material components of listening to music as data: delivery infrastructure and accessory hardware.

In terms of delivery infrastructure, downloading may indeed use fewer resources (no plastics) and generate less pollution (no physical shipping) than other formats – but only in specific circumstances. Most studies advancing arguments about the reduction of material intensity in online scenarios assume that MP3 listening is a matter of downloading an album once (from iTunes, say) and accessing it forevermore on a local hard drive. However, this form of listening exists alongside (and potentially in the shadow of) streaming and subscription services such as Rhapsody, Spotify and, now, iTunes Radio – not to mention the extremely high traffic in music streaming on sites such as YouTube (IFPI 2014). Such media channels mean that a large proportion of digital music's resource and energy quotients come from ‘indirect sources’ (Bottrill et al. Reference Bottrill, Lye, Boykoff and Liverman2008, p. 7). Although the material intensity of this delivery infrastructure is distant, it is nevertheless substantial (Morris Reference Morris2008). Indeed, emerging scholarship calls attention to the ‘aggregate material effects of discrete acts’ – such as music downloading and streaming – ‘that seem, to the online user, utterly virtual’ (Carruth Reference Carruth2014, p. 358; cf. Gabrys Reference Gabrys, Maxwell, Raundalen and Vestberg2014). In looking at those aggregate material effects, a different picture emerges. A single large server farm, for example, can consume thousands of megawatts of electricity (enough to power millions of homes).Footnote ¹⁶ What is more, the number of server farms and data centres is increasing, with their collective carbon footprint already similar in size to that of the airline industry and set to increase at least fivefold by 2020, raising what Sean Cubitt and his colleagues (Reference Cubitt, Hassan and Volkmer2011, p. 156) call an ‘emerging energy crisis of information’. Additionally, the underground and undersea cable networks that constitute the backbone of this infrastructure, and which transmit the vast majority of its data in the Western world, can raise their own environmental concerns. This is the case both in terms of possible habitat disruption that occurs when the cables are buried underground and laid across the ocean floor, as well as the materials and practices required to make such cables – the demand for which is growing (Maxwell and Miller Reference Maxwell and Miller2012, pp. 57–8; cf. Starosielksi Reference Starosielski2012).Footnote ¹⁷ Music is complicit in all these developments. It is for these reasons necessary that music scholars counter the ‘big white fluffy’ connotations of cloud computing by stressing ‘the cold hard physicality of warehouses, servers, generators and climate control devices’ (Morris Reference Morris2011, p. 3). Such an understanding not only provides grounds for a critique of the commodity status of digital music; it also strikingly asserts the material intensity of digital music.

The question of how to access this network, this infrastructure, opens up another corrective to notions of digital dematerialisation. Hard drives, routers, laptops, data sticks, memory cards, MP3 players, smartphones, headphones – these devices are resolutely material and, in various configurations, absolutely essential to digital music listening. What's more, the amount of such accessory technologies is massive, and it is growing (Maguadda Reference Maguadda2011, p. 19). Other researchers therefore suggest that any dematerialisation wrought by digital music delivery systems is offset by a larger overall throughput of digital devices (Hogg and Jackson Reference Hogg and Jackson2009, p. 338; cf. Berkhout and Hertin Reference Berkhout and Hertin2004). Yet it is more common, as Smith's discussion of Beck indicates, to tune out hardware and speculate on the ostensibly supernatural qualities of the data.

The disembodied character of the MP3 gives rise to some particular ways of thinking about the relationship between permanence and ephemerality. Here is André Millard (Reference Millard2005, p. 405), for example:

Journalist Llewellyn Hinkes (Reference Hinkes2009) provides a snappier summary of this discourse in his article on ‘The transient, digital fetish’: ‘Old formats ooze historical significance’, he says; ‘new ones are deleted with a tap’.

With downloading and data streaming, there is indeed a sense in which the MP3 severs the connection between economic and physical decay that defines earlier formats. For, regardless of the value of a song or album, its conditions of existence pull in two directions at once: availability seems assured in the cloud network, on the one hand, and deletion is always threatened by the fussiness and impermanence of digital storage, on the other. In Elodie Roy's (Reference Roy2013) words, digitalisation is at once ‘a total, continuous archive’ and ‘a space of loss, degradation and ultimate erasure’. Sterne (Reference Sterne, Bijsterveld and van Dijck2008, p. 64) puts this into historical perspective: ‘If early recordings were destined to become lost recordings’, he says, ‘digital recordings move in the same direction, but they do so more quickly and more fitfully.’ Ultimately, then, the future of digital music will be governed by a logic of the trace; it will be ‘a future where most digital recordings will be lost, damaged, unplayable, or separated from their metadata, hopelessly swimming in a potentially infinite universe of meaning’ (Sterne Reference Sterne, Bijsterveld and van Dijck2008, p. 65; cf. Roy Reference Roy2014).

Apart from any potentially unique aspects of the dynamic of disposability and durability that animates the social life of the MP3 itself, and aside from the work of remembering and forgetting that defines digital music as a politics of the archive, the question of the recorded music commodity's temporal logic is also in a sense transposed onto accessory hardware.Footnote ¹⁸ Of course, the recording industry has always been aligned with makers of accessory hardware, which is to say the consumer electronic goods industry (Frith Reference Frith, Curran, Smith and Wingate1987, Reference Frith and Frith1988, Reference Frith, Frith, Straw and Street2001). Indeed, a more comprehensive political ecology of recorded music would have to account for a longer history of playback technologies such as gramophones, radios, home and personal stereos, as well as all their woods and wires, papers and plastics, tubes and transistors. Of most interest here, though, is the sense in which digitalisation intensifies the relationship between the recording and electrical goods industries.Footnote ¹⁹ This intensification goes beyond standalone MP3 players, given the extent to which the global proliferation of gadgets such as smartphones and laptops has been bolstered by the ways in which such devices function as partly (but significantly) musical devices. Importantly, such listening devices instance a higher turnover rate than earlier ones. The accelerated temporal logic of contemporary electronics devices is rooted in an industrial–cultural conjuncture that demands constant software updates and which insists on newness. While these commodities are not quite ‘made to be wasted’, like plastic water bottles (Hawkins Reference Hawkins, Gabrys, Hawkins and Michael2013), in terms of both technology and fashion they are built to obsolesce – born to die. Digital music thus contributes to another problem: electronic waste.

The Environmental Protection Agency (EPA 2011, p. 1) estimates that 438 million new electronics products were sold in the USA during 2009, with nearly 7 tonnes either in storage or ‘ready for end-of-life management’. About 75 per cent of those devices were disposed of, rather than recycled in some way. Quantitatively, music is only a small part of this much bigger problem.Footnote ²⁰ But the very fact of its contribution to the e-waste stream emphasises that digitalisation does not dematerialise the production or consumption of music. Rather, it changes the scale and character of music's materiality. Analogously, digitalisation neither extinguishes the emissions profile nor sublimates the scrap signature of the recording industry; it changes them in complex ways. Like earlier predictions about the paperless office, claims about the possibility of a weightless musical culture assume an untenable lightness of being.

Digital music's material intensity is additionally consequential in the postcolonial world. This is partly because the so-called Global South is seen AS “a reservoir of First World hand-me-downs and sleepy-eyed memories of its earlier consumer items’ (Taussig Reference Taussig1993, p. 232), thus serving as a dumping ground for 70–80 per cent of the Global North's e-waste (Gabrys Reference Gabrys2011, p. 129). But it is also because the centrality of mobile listening technologies is equally and perhaps more pronounced in the South than in the North. This is due to the fact that, in infrastructure-challenged parts of the world, internet penetration via underground and submarine networking cables lagged behind the availability of personal, mobile digital devices. In the North, by contrast, hardwired and PC-based internet access was more quickly established as a norm. Regions in the Global South have thus constructed their own diverse uses and practices surrounding mobile communications technologies (e.g. Ling and Horst Reference Ling and Horst2011), along with their own listening practices, their own temporal logics of musical possession and dispossession – and, consequently, their own corresponding debris fields.

Take, for example, the ubiquity and significance of data sticks, memory cards, music download vendors and mobile phone listening in India (Deo and Duggl Reference Deo and Duggl2013; Rai Reference Rai, Quiñones, Kassabian and Boschi2013; Manuel Reference Manuel2014). Such developments, which represent a marked change from the earlier ‘cassette culture’ described by Peter Manuel (Reference Manuel1993), are attributed to the ‘increasing affordability of multimedia-enabled phones and voice/data plans and wider penetration of mobile coverage’, and mean that ‘the mobile phone has quickly become the most prevalent digital music device’ (Kumar and Parikh Reference Kumar and Parikh2013, p. 2863). The rise of the mobile phone and digital music, which may be considered part of India's ‘pirate modernity’ (Sundaram Reference Sundaram2010), is linked to an alarming rise in local e-waste generation (e.g. Borthakur and Sinha Reference Borthakur and Sinha2013). Nevertheless, the influx of e-waste from other countries may still represent the most significant challenge. The memory of one northeastern Indian recycler highlights the role of music in a longer history of e-waste: ‘At first, we dealt with record players, radios, VCRs and black-and-white TVs. Later on, CD and DVD players followed. Finally, computers arrived, and we started business with e-waste’ (Mohammed Moinuddin, quoted in Gallagher Reference Gallagher2014). In these ways, the political–ecological conditions of music in India have transitioned, over the course of a century, from tree branches to data sticks. Shifting musical formats and listening practices are thus tied to shifting practices of dispossession and disposal. This is the often hidden topological connection between making and unmaking, at simultaneously global and local levels, which a political–ecological approach to recorded music seeks to expose.

Similar issues arise in other places. In Kenya, for example, where the local record industry ‘has been largely informal and undercapitalized since the multinational record companies pulled out of the country in the 1980s’, the recent explosion of a music-based subsector of the telecommunications industry could mark a new beginning (Eisenberg Reference Eisenberg2012). Kenya's digital music listening practices and economies – as with those of India and increasingly numerous other African nations – take place primarily through mobile and handheld devices (IFPI 2014; Matinde Reference Matinde2014). In addition to streaming and downloading, there are significant markets in ringtones and ringback tones, a service by which subscribers can customise the sound heard by callers as they wait for someone to answer. Regardless of the questionable long-term financial viability of such developments, due to licensing and piracy among other issues (Eisenberg Reference Eisenberg2012; cf. Gopinath Reference Gopinath2013), the increasing centrality of personal electronics in the contemporary global listening formation makes it necessary to face the ways in which m-commerce means e-waste. In other words, shifting conditions of music making and music listening are tied to shifting conditions of politics and ecology – often in unexpected or hidden ways. To uncover those ties and those conditions, to account for their mutual influences and consequences, even when they are obscured by the rhetoric of digital dematerialisation, is the challenge that defines the political ecology of music at the outset of the 21st century.