Sunday, September 29, 2019

Music's Carbon Footprint (2)

        I've been thinking more about music's carbon footprint, and in doing so this month I prepared a short essay for possible publication on the subject. Corrections and suggestions are invited:

Music’s Contribution to Global Warming

            How much does music contribute to global heating? If ethnomusicologists are to contribute our understandings of people making music to discussions surrounding carbon emissions, greenhouse gases, and the climate emergency, it would be useful to know at the outset how much all the activities surrounding music production, delivery, and consumption contribute to the overall emission of greenhouse gases (GHGs) that have been heating the Planet.
            In the preindustrial era, of course, musical activities did almost nothing to raise global temperatures. Musical activities made a larger contribution to GHG output and climate change during the industrial, mass consumption era. Musical instrument and sheet music distribution, radio, recordings and television utilized energy resources on a larger scale, as did the shift to electronically amplified instruments. Ironically, in our post-industrial era, the delivery and consumption of music via the internet requires more energy than the manufacture, distribution and consumption of music on vinyl records, cassettes, and CDs did during their years of peak use (Brennan and Devine 2019).
Music GHG statistics are few and far between; much more research needs doing, but on the face of it musical activities contribute a very small proportion of greenhouse gas emissions to the world’s total. We do have reliable GHG estimates for all musical activities in the UK in 2009. We also have them for recorded music in the US from 1977-2016. With US figures adjusted upwards to include live performances, in 2009 the musical activities in the UK and US accounted for about 3 million of the then-total 48 billion tonnes of GHG emissions, or .00625%, a miniscule amount.[1] Extrapolating from the UK and US to the rest of the world, the contribution of the music industry overall in 2009 is unlikely to have exceeded .02% or two-hundredths of one percent; still seemingly insignificant. When only 90 corporations, the vast majority being fossil fuel producers such as Exxon and state entities such as GAZPROM (Russian Federation) and Aramco (Saudi Arabia), contribute more than 70% of GHGs annually (Heede 2014), one wonders whether ethnomusicologists’ time wouldn’t be better spent in convincing our institutions to divest from fossil fuels than in encouraging the music industry to reduce GHGs.  Here’s a more optimistic way to think about it: the US music industry contributes almost as much to the US’ annual gross domestic product (GDP) as the automobile industry.[2] Everyone knows that the auto industry is moving, however slowly, towards more efficient gasoline engines and hybrid and electric vehicles. Few doubt that an all-electric vehicle future would make a significant reduction to GHGs, so long as most of the electricity comes from renewable energy sources. The same could be said if the music industry moved to a carbon-neutral future.
            Let’s look at GHGs more closely (fig. 1).  

Those emitted by human activities consist of carbon dioxide (CO2) from burning fossil fuels (76%); methane from agricultural activities, waste management, energy use and biomass (16%); nitrogen oxides from fertilizer use and other agricultural activities (6%); and fluorinated gases from industrial processes, refrigeration and so on (2%). Considered in terms of energy source (fig. 2), global GHGs come from electricity and heat production (25%), industries that burn fossil fuels on site for energy (21%), agriculture, forestry and other land use (24%), fossil fuels burned for transportation (14%), fuel burned to heat and cool buildings (6%), and other (10%) (United States Environmental Protection Agency). 

From 1970 to 2011 global CO2 emissions increased by nearly 100%, with 3/4 of the increase coming from industrial processes and fossil fuel burning. In 2012 the total amount of GHG’s emitted as a result of human activities had risen to 53 billion tonnes from 48 billion in 2009 (The World Bank).
Three studies of the UK and US music industries help in understanding music’s contribution to global warming. First, the Environmental Change Institute at Oxford University conducted a report for the environmental organization Julie’s Bicycle on the sources and amount of GHGs generated in 2007 by the UK music market (Bottrill et al. 2007; Bottrill et al. 2010). The report calculated an annual total of 540 million kilograms (kg.) ( = 540,000 tonnes) of GHGs, coming from the following sources: manufacture and distribution of musical instruments, books and sheet music, and recorded music, 138 million kg.; live performances (including audience travel which generated 231,00 million kg.), 402 million kg. This 540,000 tonnes applied only to the UK. Although it included GHGs generated by music’s share of distribution over the internet, in 2007 internet downloads and streaming were fewer than today while correspondingly more music was delivered via CD.
            A second useful study compared the amounts of GHGs required to manufacture and deliver music on plastic (vinyl, cassette, and CD) versus the internet. The study took into consideration that some albums delivered digitally over the internet were subsequently burned to CD-R discs for CD use and kept in jewel cases. The authors concluded that “Despite the increased energy and emissions associated with Internet data flows, purchasing music digitally reduces the energy and CO2 emissions associated with delivering music to customers by between 40 and 80% from the best-case physical CD delivery, depending on whether a customer then burns the files to CD or not” (Weber et al, 2009). Ten years later, consumers are burning many fewer CDs; indeed, most computers today no longer contain optical drives for disc burning.
            This second study predicted that a shift to internet music delivery would significantly reduce GHGs. However, as its authors acknowledge, it was conducted for two IT corporations (Microsoft and Intel) that had an interest in seeing that kind of result. What is the situation today, when most recorded music is delivered to customers via internet streaming on Spotify, Apple Music, YouTube, Pandora, and so on? The results are not quite so sanguine. In third environmental cost study, Matt Brennan and Kyle Devine chose to sample GHG outputs from certain key years from 1907-2016. Devine terms this shift from plastic to internet delivery the “dematerialization” of music. They reported total GHG outputs for the US only, from vinyl, cassettes, and CDs, versus GHG outputs from internet delivery via mp3 and streaming. They reported (fig. 3) on the years 1977, 1988, 2000, and 2013-2016 (Brennan and Devine 2019; Devine 2019). Figure 3 shows that in 1977 US CO2 emissions from the manufacture of and 

sale of music (chiefly on vinyl records) were 140 million kg. The first CDs appeared in 1982. In 1988 when vinyl records made up roughly 19% of sales revenues, CDs 20%. and cassettes and 8-tracks 60%, the total CO2 figure for the US was slightly lower, 136 million kg. In 2000 when CDs were at their peak revenue (and not many vinyl records and cassettes were sold) that number had increased to 157 million kg., almost all from CDs. And yet by 2016, when the revenue from CDs was 1/8 of what it had been in 2000, and when most music recordings were sold via mp3 downloads or by subscription streaming on the internet, the US energy costs of music delivery had increased to as much as 350 million kg. of GHGs, or 350,000 tonnes (Brennan and Devine 2019; RIAA US Sales Database).
            How could this be? It seems counterintuitive that the environmental costs of making plastic CDs and jewel cases could be about half as much as internet music delivery with little or no plastic product (and waste). It seems inconceivable that a shift from a manufacturing economy to a service economy for recorded music could result in double the CO2 emissions. Yet what drove up the cost of internet delivery was “the energy used to power online music listening. Storing and processing music in the cloud depends on vast data centres that use a tremendous amount of resources and energy” (Brennan and Devine 2019, 3).[3]
To be sure, Brennan and Devine did not calculate the environmental cost of distributing LPs, cassettes and CDs to consumers in record stores. Nor did they calculate the environmental costs of making the record, CD and DVD players. Doing so would have increased the GHG figures for 1977, 1988, and 2000. But they also did not calculate the costs of making the smartphones, mp3 players, and computers that consumers rely on today to download and stream their music. Moreover, as the US population grew from 203 million in 1970 to 307 million in 2010, it’s likely that there were proportionately half again as many music consumers in the latter year than the former, consuming 50% more energy, And of course, unlike the Oxford University study of UK music’s carbon footprint, Brennan and Devine did not attempt to calculate the environmental costs of live music which, extrapolating from the UK study (Bottrill et al. 2007), would have been for the US about three times more than for recorded music, or approximately 1.05 million tonnes of GHGs. Add that to the carbon footprint of recorded music and in 2016 the US likely generated around 1.4 million tonnes. If the UK’s footprint was 540,000 million tonnes in 2009, extrapolating to 2016 from that and then adding the US figures would result in a UK plus US annual total somewhere between 2.5 and 3 million tonnes ( = 6.6 billion US pounds) of GHGs for 2016.
            Efforts to reduce the music industry’s contribution to global warming have been underway for more than a decade. The Julie’s Bicycle-sponsored 2007 Oxford Study resulted in a variety of actions to lower music’s carbon footprint, such as the change from plastic CD packaging to cardboard digipacks. In 2013 REVERB was created and dedicated to reducing touring bands’ carbon footprints insofar as possible. On their tours, Dave Matthews Band, Phish, Dead and Company, Drake, Walk the Moon, and others reduce carbon use by employing solar energy, distributing reusable water bottles, providing solar charging stations at concerts, and handing out information about environmental issues, green products and tech, and so on. REVERB sponsors a Farm-to-Stage program that works with local farmers to provide artists and their crews with locally sourced food (REVERB). By 2018, Spotify had closed almost all of its data centers and reduced its carbon footprint by 1,500 tons of CO2 while switching to Google cloud services, which like competitor Apple has “gone green.” By converting to solar power and purchasing renewable energy certificates (that work like carbon offsets) they can claim their data centers are carbon neutral. In July, 2019 a group of music industry professionals formed an organization called Music Declares Emergency, calling for “the music industry to acknowledge how its practices impact the environment and to commit to taking urgent action” and to “work toward making our businesses ecologically sustainable and regenerative.” Some of the suggestions include reducing the energy used on tours, greening merchandise, using sustainable materials, and purchasing carbon offsets. Many musicians have signed on as supporters (MUSIC DECLARES EMERGENCY). And because music has the power to raise environmental consciousness and incite environmental activism its impact goes well beyond the boundaries of the music industry to galvanize the environmental movement more broadly, whether targeting fossil fuel corporations, agrochemical producers, or threats to species extinction.[4]


Bottrill, Catherine, Geoff Lye, Max Boykoff and Diana Liverman. 2007. First Step: UK Music Industry Greenhouse Gas Emissions for 2007. Executive summary. London: Julie’s Bicycle.
Bottrill, Catherine, Geoff Lye, and Max Boykoff. 2010. “Carbon soundings: greenhouse gas emissions of the UK music industry.” Environmental Research Letters 5, doi 10:1088/1748-9326/5/1/014019.
Brennan, Matt, and Kyle Devine. 2019. “Music streaming has a far worse carbon footprint than the heyday of records and CDs—new findings.” The Conversation, April 7. Accessed 30 August 2019.
Devine, Kyle. 2019. Decomposed: The Political Ecology of Music. Cambridge: MIT Press.
ECOSONG. N.D. Ecosong.Net. Accessed 30 August, 2019.
Heede, Richard. 2014. “Tracing anthropogenic carbon dioxide and methane emissions to fossil fuel and cement producers, 1854-2010.” Climate Change 122 (1-2): 229-241.
MUSIC DECLARES EMERGENCY. N.D. Accessed 30 August, 2019.
REVERB. N.D. Reverb: About Us. Accessed 30 August, 2019.
RIAA U.S. Sales Database. N.D. U.S. Recorded Music Revenues by Format and U.S. Recorded Music Sales Volume by Format, 1977-2018. Accessed 25 September 2019.
Siwek, Stephen, and Joshua Friedlander. 2018. The US Music Industries: Jobs and Benefits. Prepared for the RIAA (Recording Industry Association of America). Accessed 19 September 2019.
The World Bank. N.D. “Total greenhouse gas emissions.” Accessed 18 September 2019.
United States Environmental Protection Agency. N.D. “Global Greenhouse Gas Emissions Data.” Accessed 18 September 2019.
Weber, Christopher L., Jonathan O. Koomey, and H. Scott Matthews. 2009. “The Energy and Climate Change Impacts of Different Music Delivery Methods.” Final report to Microsoft Corporation and Intel Corporation. Accessed 19 September 2019.

[1] Calculated by combining and extrapolating from Bottrill et al. 2010 and Brennan and Devine 2019. A (metric) tonne is 1000 kilograms, or 2200 pounds, which is 1.1 times the US ton of 2000 pounds.
[2] Siwek and Friedlander 2018 calculated that in 2015 the music industry contributed $143 billion to the US economy, 4/5 of the auto industry’s contribution.
[3] Their study, “The Cost of Music,” was widely reported in the press in April, 2019. Devine’s book on the subject, Decomposed (2019) offers more detail. Some of these data centers now are powered partly if not fully by solar energy, however.
[4] Ecosong, a collaboration among musicians, media makers, scientists and community organizations, is one of many such efforts (ECOSONG).

Thursday, August 1, 2019

Animal sound communication, human language, and perspectivism

          I return to the question of animal sound communication, which I wrote about in January of this year. Animals—humans included—communicate with one another in song, speech, and various in-between modes such as chant. Animals also communicate by manipulating objects to make sounds, whether beavers slapping the water with their tails or humans playing musical instruments. Because animal communication is necessary for survival, it is necessary for sustainability of life as we know it on planet Earth. Sound, of course, is one of the principal ways in which animals communicate with one another.
         Among the questions that interest scientists about nonhuman animal sound communication are, first of all, whether species’ communication in sound is purely functional, or whether it has an aesthetic component. Do animals always make sounds for a practical purpose, in other words, or might they do it for pleasure, as humans do? Might some of their practical sounds, such as birdsong, offer pleasure as well as having functions such as locating oneself to one’s mate? A second question is whether animal sounds constitute language. That is, do they exhibit the structural and functional characteristics of what we know as language, even if in some simple stage? Or are their sounds merely signals? Songbirds, humpback whales, chimpanzees, elephants and dolphins seem to be good test cases for language. All of them communicate with a variety of complex sounds; in terms of their structure, could they constitute a language? Scientists do not agree on the answer. Generally speaking, the evolutionary biologists who study animal behavior and sound communication favor the idea that nonhuman animal sounds, no matter how complex they are in structure, constitute signals, stimuli that result in automatic responses in the listening animal. At the same time, many scientists are conducting experiments to see if that is really true, or if the complex sounds of some species could be said to constitute anything resembling a language, and whether there is any evidence suggesting that it also has an aesthetic component.
         Recently I came upon an educational website that pursued differences between human and animal communication. The linked to a lecture by a professor at the Leiden University Center for Linguistics, and then examined the supposed differences between human and nonhuman animal communication. As I read through these differences, I realized that some of the evidence from birdsong and humpback whale song casts doubt on those claims for difference, or at least for a sharp dividing line between human and nonhuman.
Black-capped chickadee. Photo by Ryan Hodnett
         For example, humans are said to be able to string sounds together in an infinite number of ways to create meaning via words and sentences, and animals are said not to communicate by arranging sounds in this way. And yet there is scientific evidence that chickadees string together phonemes in a simple manner to indicate, for example, not only that there is a danger from a nearby predator, but also how great is that danger. Humpback whales, also, have been observed to string together sounds in different ways to indicate different meanings. Part of the difficulty here, it would seem, is that humans do not understand animal language. For another example, humans are said to be able to invent new words easily, whereas animals must evolve in order for their signals to change. But the invention of new words and their addition to human language proceeds slowly, whereas animals change their signs before they evolve, and then evolution selects those whose signs are most suitable for survival. A few songbirds, for example, who are permanent residents near highways change the pitch range of their songs to avoid traffic noise, and only then does natural selection favor those birds that have changed their songs, resulting in a higher percentage of birds that have adapted and that can, presumably, survive better. Experiments have shown that this evolution is rapid.
         A couple of other differences turn out to have exceptions, also. It is said that humans acquire language culturally, by learning, whereas animal communication is inborn and cannot be learned. But that isn’t so. Songbirds raised by foster parents of another species learn to sing their foster parents’ songs. Another difference discussed on the website is arbitrariness: human language is symbolic and filled with ideas, whereas animal communication is said not to be symbolic. The jury is still out on that one, but one wonders how we would be able to recognize whether animal communication was symbolic when we don’t understand their sounds in the same way they do. In other words, what we conclude by observation to be no more than stimulus and response may in fact be arbitrary to some extent on account of its complexity, or on account of factors that we ordinarily wouldn’t take into consideration, such as the time of day or night of the utterance, its tempo, how frequently it is repeated, and so on.
         For centuries scientists have claimed that only human beings have language, but it no longer certain that this is true. Instead, it’s possible that language is something best understood to exist on a continuum of development, with humans having a more fully developed language while some other animals exhibit it at an earlier stage of development. If animals exist on an evolutionary continuum, it makes some sense to think of language in this way. Obviously, humans have had the capacity to develop language into a very sophisticated communications tool, and beyond that into an artistic expression with aesthetic qualities. At the same time, we should not be too hasty in assuming that animals communicate only by means of signals that act as stimuli to effect responses. After all, scientists do not know what it is like to communicate as a whale or a dolphin or an elephant.
         On the other hand, many Amerindian indigenous peoples’ ecological knowledges take a very different approach to animals and animal sound communication: perspectivism and multinaturalism. I’ve written about this to some extent in a completed chapter for an as-yet-unpublished book edited by Harris Berger, one that has been in the making for a decade or so: the "Oxford Handbook of Phenomenological Ethnomusicology." In it, I discuss evidence for expressive culture among animals. From a perspectivist viewpoint, animals have intentionality, consciousness, reflective thought, experience, aesthetics, language, and expressive culture because what it is like to communicate as a whale or dolphin or elephant from the animal’s point of view is exactly the same as what it is like to communicate as a human because animals think of themselves as humans.