This is an important and brilliant book. Kyle Devine makes the case for the ‘political ecology of music’. This represents an expansion of both musicology and music business studies. Rather than confining ourselves to ‘the conditions of making and hearing music as music’, we must address ‘the conditions surrounding all the other efforts and substances that go into making, moving, and unmaking the objects that constitute music's material culture’ (p. 13). This is a transformative undertaking: ‘A critique of political ecology may add to the public relevance of music research by spurring scholars to look at those places where music's economic and cultural significances encounter its environmental and human consequences, thus opening the door to interventions at the intersections of music, media, and environmental policies’ (p. 26).

As his part in this ecology, Devine addresses recording formats. He looks at the natural and human resources expended in their creation, as well as the consequences of their eventual disregard. The results are chastening. Decomposed is divided into three eras. First there is the period of shellac (1900–1950), with its exploitation of ‘resources and workers’ (p. 76) and its industrialisation of music. Next comes the plastic-based production of vinyl, cassettes and CDs (1950–2000), which implicates the recording industry with the petrochemical industry, producing millions of kilograms per annum of a non-biodegradable material. Finally, we have the period of data (2000 to now). Devine disabuses us of any notion that consumption has been dematerialised. To repeat this mantra is to be ‘be fooled by the commercial rhetoric of streaming and cloud computing, to underestimate the material infrastructure and energy usage of data processing, and to misunderstand their consequences’ (pp. 131–2).

Deservedly this book is gaining attention beyond the confines of academia. It has been previewed in The Conversation (Brennan and Devine Reference Brennan and Devine2019a), the Independent (Brennan and Devine Reference Brennan and Devine2019b) and Rolling Stone (Blistein Reference Blistein2019). Moreover, while it is the relevance of its thesis that has prompted this interest, Devine warrants commendation for the manner in which his project is realised. The research work is excellent, uncovering much new material from a wide range of sources. The level of theoretical erudition is also to be praised. As well as having a grasp of diverse academic concepts, Devine has the ability to make them palatable for the reader. He is an elegant writer too. There are numerous neat turns of phrase and a generally cogent and concise marshalling of material. The use of images should also be noted: the book is peppered with apposite photographs, illustrations and graphs.

There are some caveats. Decomposed has a large number of footnotes. Some of these would have been welcome in the main text, in particular the notes that explain the statistical methodology and the ones which admit that some music-related companies are attempting to stem their environmental damage. It is also arguable that Devine spends too long pre-empting criticisms, most notably in relation to the pedagogic ire he feels his ‘musicology without music’ may solicit (pp. 168–84). The material in this book is also distorted towards the vinyl revival. While Devine is correct to draw our attention to the material consequences of this phenomenon, opening and closing the book with this subject feels like a remnant of an earlier orientation of his research.

And then there is the aspect of this book that is soliciting media attention. Devine reaches the startling conclusion that our data-centric period has raised the environmental costs of music ‘higher than ever before’ (p. 160). He goes through some hoops to reach this point. Firstly, a lot of store is set by the Jevons effect; the idea that ‘greater efficiency in resource usage is actually followed by greater consumption of that resource’ (p. 33). While there is acuity here, the validity of the Jevons effect is much debated (see for example Gillingham et al. Reference Gillingham, Kotchen, Rapson and Wagner2013) and it does not supply a dependable formula for determining consumption. Secondly, in outlining the material impact of data, Devine brings together hardware and software, pointing out how the recording industries’ rapid turnover of music is now matched by the consumer electronics industries’ turnover of devices to play these tracks. This is another crucial point, but Devine does not outline the environmental impact of the machines that played shellac, vinyl, cassettes and CDs; therefore his comparison between eras is uneven here. Thirdly, Devine introduces his most resonant information, detailing the greenhouse gasses that plastic and data emit. In this instance, his different formats are more carefully drawn together: a methodology is outlined and an impressive array of data is employed (pp. 236–7). Nevertheless, to reach the eye-catching conclusion that ‘the amount of carbon dioxide equivalents could be more than double in the age of streaming when compared to even the highest output during the plastic era’ there is some sleight of hand (p. 159). Devine does not concentrate on streaming alone, but instead combines the emissions of streams and downloads with the continuing environmental impact of ‘residual [plastic] formats’ (p. 158). In addition, he is still not comparing like with like. His figures for plastic are based on the material production of the recordings; the figures for data are derived from the environmental costs of storage and transmission. There is also the admission that his numbers are based on ‘assumptions and averages’ (p. 236).

As is common in this work, Devine pre-empts criticism. He states that ‘Finding out that some of my quantities are off by a thousand or even a million would not change the book's overall message. The amounts are huge and the issues ought to be reckoned with’ (p. xi). He is correct in this. While this book begs to be questioned – both in relation to its pedagogic provocations and its statistical methodology – it is also one whose importance cannot be gainsaid. This is a subject with which we all must engage.