The Society for Cinema and Media Studies held—virtually, for a second year in a row, due to the ongoing pandemic—its 63rd annual conference in the spring of this year. Reprinted here and lightly edited for clarity, is one panel from the conference, “Sounds of Accompaniment: Music, Technology, and Labor amid Capitalist Aesthetics,” a fascinating collection of talks from four thinkers of music, sound, and media discussing historical and recent phenomena that are relevant to our current-day social and technological environments. This panel was chaired by Andy Stuhl of McGill University and sponsored by two SCMS Scholarly Interest Groups (SIGs): the Radio, Audio Media, and Podcasting Studies SIG and the Sound and Music Studies SIG.Today I’m going to talk about two of the main industrial music programs of the 1940s: Muzak’s piped-in service and RCA Victor’s subscription program. And I’m going to try and show how they relied upon a new understanding of music as functional—music could affect workers’ bodies and minds and emotions and their morale towards greater efficiency. So, this was an understanding of music as functional and therefore laboring. Surveys of worker music preferences, especially, both collected data on the effects of music, and they also reinforced this functional understanding of music. And this was both for the workers and for management.Further, industrial music that was based on worker preference surveys gave workers a sense of individuality that boosted their morale. And I think by reading this discourse against the most emphasized (to management) features of the RCA system, ability to contact any worker immediately through their public address system, created an interesting loop. So, according to RCA, the music played over the loudspeakers conveyed to the workers that management was listening to them and that management could then assume that the workers were also attentively listening to them through the public address system. And so, in this way, the sounds of the industrial music system facilitated a factory-wide culture of what I want to call pan-aural listening. I’m floating that concept out there for all of you to pick apart.I also want to try a conceit here: figure 1 is an image—or rather, a series of three images—from the Picture Arrangement Test that had been developed by Silvan Tomkins and Daniel Horn in 1942, further refined in 1952.1 The Picture Arrangement Test, the PAT, was an adaptation of the thematic apperception test, the TAT, which functioned like the Rorschach Test. Essentially, the test subject was asked to order the images here into a “story that made sense.” And then the subject’s answers were analyzed for reasoning abilities to give information about the subject’s intelligence, personality, and possible psychopathologies. This was developed in the 1940s to assist with the selection and the training of personnel for industrial settings. So, a lot of the images are placed in industrial settings. Each plate consisted of three images, arranged in 120-degree rotations from each other in order to randomize the test subject’s engagement with them. In Figure 1, which was Plate 13, the three images are a man (Tomkins’s text terms this person the “hero”) standing at a machine, touching it with his hands, the man turned away from his machine with a thought bubble of a second man talking hovering overhead, and both men facing each other, the second man talking.There were originally 25 of these plates. They were used, again, for training and placement in industrial settings. And later Tomkins actually started incorporating visual representations of his affect theory that he was developing at this time. Affect theory was, briefly, the theory that there are universal physical expressions of the building blocks of emotion, essentially, and that these are universal and standardized and that there are also standard reactions to them. Affect theory took off in a slightly different direction after Tomkins. So, what I want to do is to use plate 13 here to help frame my talk today. So how might we arrange these various images?First, let’s start with the hero working. We know that he’s working because his hands are on some machine here and he’s looking at his hands. So, we’re seeing this as a representation of working. The next scene he is being talked to, maybe yelled at, by his boss. We know this person is his boss because he has a hat and a name tag, and because he is a little bit taller. We might note that everybody’s frowning a little bit. And then the third scene is the hero-worker resentfully remembering his boss either in general or that specific encounter of the previous frame. So, I’m calling this the management-versus-labor interpretation of this. The grumpy laborer version. And this is to underscore that industrial laboring spaces at the beginning of World War II were this sort of top-down structure. There were pressures for efficiency and possible grumbling from organized labor that couldn’t be fully articulated due to the national cause of war. I want to talk today about the use of sound in these laboring spaces during the 1940s to potentially redefine labor and how this inflected management-labor relations, worker morale and established a system for pan-aural listening.I don’t think I actually need to chart the complete arc of music as an accompaniment to labor for this group. But I want to note that by the early 1920s, phonographs had been introduced to factories for use during breaks. Additionally, in this period, there were studies and marketing efforts by the Edison Phonograph Company, as well as others that established that music could consistently affect mood, mind, and even the body.Following a British report on the role of music and increased efficiency among workers in 1937, there were similar efforts to employ music in factories in the United States. There were two main models that came out of this, both dependent on new loudspeaker technology. There was a Muzak model of piped-in music programs via electric wires from a central source, and they would go out to the individual factories’ public address systems. And there was the RCA Plant Broadcasting System that instead required a local phonograph to be connected to the public address system. And then also, because of that, required an onsite music director that was responsible for switching out the recordings as well as a subscription to the RCA record library.Both of the industrial music models depended on employee surveys to measure efficacy as well as to refine the playlists, play time, and to develop these new morale-building strategies. These systems again were connected to the factory amplification systems to play music for workers as they worked. So, this was no longer music during breaks, this was music during work. Music itself was being put to work.So again, Muzak was founded in the early 1930s; it recorded its own music and then piped it in to these curated music programs via electrical wires to subscribers. It described itself not as entertainment. It was instead, “music specially designed to help workers without distracting them from their tasks.”2 Again, because it was a subscription service, you didn’t need to have your own equipment. They would come and get everything set up for you. Muzak claimed to have specially -designed channels for different types of labor: there’s industrial work versus white-collar office work versus public lobby spaces like hotels and restaurants. They claimed to have specially engineered sound quality that could “slice” through the sounds of the machinery on the factory floor. They described themselves as management-, labor science- and government-approved.The science part—which, I’m a historian of science, so this is the part I get extra excited about—is, they had a number of psychologists working on measuring worker efficiency. Muzak measured efficiency by counting things. They counted objects; they counted units of things produced by the workers. They also counted absenteeism and they counted injury and would try and chart these incidents against the intervention of music (see figure 2). Also, though it’s not totally obvious in figure 2, but they often just played music for like 30 minutes at a time as an injection into the system. They recognized that playing music continuously could lead to boredom and reduced effects.Muzak also refined their programming through worker surveys. So, this collected data towards that. It also, I think in the case of Muzak, much more so than RCA, worked as a branding exercise. Workers were asked for demographic information and to indicate whether they liked or disliked various musical genres. Then they were directed to agree or disagree with a series of sentences that I find delightful; “Muzak gives me a lift,” “Muzak helps relieve my fatigue,” “I wish the music would be stopped immediately.”3 And so, in all of this it allows for them again to refine their programming. It also offers the appearance of caring about worker opinions and in internal memos and discussions, they talk about asking for worker feedback in order to keep the unions happy and make them think that we care about what they want. It also standardized the ears of the workers and underscored the functional applications of music. It created affect.In brochures aimed at management, Muzak described their product as a lubricant that eased tensions with workers.4 It allowed management to look like it cared about the humanity of its employees; music was a gesture of courtesy. And then they go on to say that it’s a practical gesture because “it’s real.” Although again, this feels a bit like this is about keeping up appearances of caring about workers.Now we’re back at plate 13 of the Picture Arrangement test and I’ve rearranged the scenes. So, here we have what I am calling the Affect-Realized Interpretation. First is the worker-hero thinking about his boss talking/yelling. But maybe the worker is not remembering the boss yell. Instead the worker is anticipating his interaction with his boss. Next is the image of him working. Then third is the imagined interaction coming to pass, the boss yells at our worker-hero. And so we can think about the realization of this affect and that in Muzak’s case they are actively training workers to realize, maybe not conflict as in this picture, but a harmonious relationship via the piped-in Muzak.RCA Victor’s Plant Broadcasting System, in contrast to Muzak, offered a subscription service. The plants would regularly receive new records as well as guidance for the plant’s music director. And part of this guidance was the monthly Industrial Music Newsletter, a wonderful archival resource for thinking through the listening experience of workers.Music was absolutely a part of this listening experience, but there were also a number of other worker morale-oriented programs that RCA Victor, via the Industrial Music Newsletter, recommended. RCA endeavored to collapse morale and production together. Because morale was presumed to have direct consequences for production, it could be optimized to increase production. Many of the underlying features of the RCA industrial music system—the practice of highlighting the accomplishments or shortcomings of individual workers in plant-wide broadcasts, quiz competitions and awards, and surveying worker opinions about music—were promoted by RCA as a way to foster individuality or personhood, and they use this language very clearly, in turn, to increase morale. The system could also, through this very cultivation of individuality, control the factories’ workers. And so I would argue that this counterintuitive system of control was only possible through sound and a very specific understanding of the technology of industrial sound systems at this time. RCA promoted an idealized plant of attentively listening workers that in turn, because of management’s use of the industrial sound system, believed that management was also listening to them.The Industrial Music Newsletter offered support and guidance for managers, usually opening with a suggestion for a playlist. It would be holiday-themed when relevant, there was also always a report from Victor Vim (more on him shortly), new releases, relevant scientific studies and testimonials about improved morale that were due to the music in the workplace. It also showcased other factories that had used their RCA industrial music system well. The Hollingshead factory, for example, did a weekly trivia quiz over their public address system. Again, it wasn’t just about music.RCA also polled worker opinions, and the Industrial Music Newsletter included several images of workers filling out blank request cards and dropping them in request boxes. They also asked workers to fill out surveys though these forms, compared to Muzak’s, were much more open-ended inquiries about employees’ music attitudes. There was an assumption that the workers were listening to the RCA Victor Plant Broadcast System.A bit more about Victor Vim. A one-eyed, walking phonograph record, pep-squad captain, he was introduced to Industrial Music News in June of 1944. He was actually a crossover from a 1943 promotional pamphlet, “Manpower, Music and Morale.”5 And this jaunty fellow brought music and morale together. He explained that this was his job: to help music aid the workers. The World War II context cannot be ignored. It added production pressure and provided language of morale and patriotism. Here, the newsletter offered suggestions for what type of music directors should consider for when the D-Day invasion news arrived at their plant and this underscores the RCA version of morale as military-adjacent.One last bit on the way in which the RCA Plant Broadcasting System presented itself: apparently there was perennial problem of losing individuals in the factory. This was solved by the use of the paging feature of the RCA industrial music system. Paging required, of course, that everyone was listening for their name to be called and would then respond. So, it presumed that the listening was part of this loop. A close reading of the admittedly brief publication, the Industrial Music Newsletter, reveals a constant refrain: that showcasing individual workers, highlighting their individuality and personhood via the industrial sound system was the best way to demonstrate the humanity of the executive. So, not just the workers, but the manager. He, the manager (and it was always a he), showed that he cared via sound. Sound was a tool of demonstrating his humanity. So again, the central feature of the RCA Plant Broadcasting System helped management to improve morale by embracing workers as individuals through sound. This was only possible via sound.This is my last attempt to rearrange plate 13 of the Picture Arrangement Test. Let’s begin with the image of the boss talking at the worker. Then the worker, remembering or anticipating. Here I want to suggest the possibility of auditory memory and also auditory anticipation at work. And finally, the worker efficiently working. I give you the Efficiency of Pan-aurality Interpretation.To sum up, hopefully I’ve conveyed that the historical record suggests several scenarios playing out parallel or, likely intertwined. The images of plate 13 can be arranged many ways. Management practices were informed by the pressures of organized labor, the war, and profit margins. They were also facilitated by psychological theory, applied psychoacoustics and sound-making technologies. And in these industrial setting sound also labored. What were the sounds of labor? Music, machinery, the voice of the manager, the aggregated voice of the worker (mostly via survey), the manifestations of psycho-acousticians and sound engineers, and the affective auditory memory and anticipation. I’ll stop there. Thank you.Thanks so much, Alix. My talk is called Musical Programming: Automation and Aural Anxiety from 1950s Radio to Spotify. And in this talk, I’m going to oscillate between two vantage points, one in the present, and one toward the end of the 1950s. The reason I will do this oscillating is because I want to bring a media historical approach to bear on a recent tendency among researchers and critics to use the figure of Muzak to critique present day streaming services. These writers give different explanations for why Muzakification, as I’ll call it, is the right way to understand streaming’s effect on music—and for why this is something people should worry about.But even in the critiques that take an active interest in the historical relationship between Muzak and streaming, there seems to always be the same gap in the narrative. Namely, how did this musical system that managers imposed on factory workers and retail customers in the 1940s make its way into twenty-first century entertainment media that people choose to and even pay to hear in their leisure time? In other words, how did this industrial feature of factories that Alix showed us cross over into entertainment media? I want to propose that an important, while certainly not the only, part of that crossover occurred through a venture called Programatic [sic] Broadcasting Service at the end of the ‘50s and into the ‘60s. Programatic was a subsidiary of the Muzak corporation that for a few years became a major player in the then-emerging field of radio automation. It established a business model and a musical aesthetics, which it touted as “adult music,” that would both become enduring structures in American commercial radio.By exploring how Programatic sold its services to station owners, my aim is to untangle the industrial side of a link between automation and Muzakification in popular sound media. Today these trends are the targets of interlocking anxieties that constrain people’s music listening as part of a larger complex that I will call musical programming. But before I get too far into that, I first want to return to the present (or recent past) and take a closer look at this type of writing where we see the Muzak metaphor deployed.When I first planned this project, I imagined I’d alternate between material on Programatic and a lit review of writing that uses this metaphor, but luckily for me, David Hesmondhalgh already did the latter part last year. He was interested in drawing out specific anxieties from literature that addresses “streaming’s effects on music culture”—that is, the effects that streaming seems to exert on “musical texts” and “musical experience” separately from its economic effects on musicians themselves.6 And I should note that Hesmondhalgh explored the latter part in an earlier article, in case it seems like I’m suggesting he dismissed it. Hesmondhalgh’s taxonomy of these critiques includes five themes in total, three of which are near and dear to Muzakification. One is an increase in functional music; two, a trend toward blandness in music; and three, a receding of music into background status. All these categories make frequent use of the Muzak metaphor, as in the 2017 essay by Liz Pelly on Spotify and mood playlists “The Problem with Muzak: Spotify’s Bid to Remodel an Industry,”7 which Hesmondhalgh cites prominently.Hesmondhalgh casts some, I think, needed skepticism on these expressions, pointing out that “the recent concerns about the use of music to accompany other activities can seem rather odd when seen in a larger historical context.”8 I would further his caution by making sure we don’t take the claims of music distributors at face value in our eagerness to critique them. Pelly’s essay, for instance, does seem to take for granted that Spotify’s algorithm is able to manipulate something called mood or affect, much like Muzak and RCA claimed to be able to do. Paul Anderson provided a more extreme formulation in his 2015 article “Neo-Muzak and the Business of Mood” by equating mood playlists and antidepressants under an aggressively cynical frame concerning commercial mood management.9 For the purposes of this paper, I’m less interested in sussing out exactly what various companies and critics have meant by mood and more interested in the fact that both parties believe that in general, these platforms are able to use sound to manipulate their listeners. Whatever mood is, it is also the most commonly attributed vector for this manipulation, the occurrence of which is the basis for the larger anxiety I want to examine here.And an important part of that anxiety is one that Hesmondhalgh doesn’t cover: the automation of music. Automation actually surfaces in Pelly’s essay when she describes how musicians lack an easy way to exempt their work from these automated advertiser-branded playlists. But a larger thrust of this claim sees an imminent or perhaps already happening automation of music itself. The automation claim saw a flurry of expression in connection to Spotify, also in 2017, when a “fake artist” scandal came to light. In this case, listeners noticed that recordings of mysterious origin had earned top placement in high-traffic mood playlists on the platform. Multiple responses to this information, including one by Andrew Flanagan for NPR, raised the issue of AI-generated songs—even while acknowledging that the evidence pointed towards Spotify having commissioned these tracks from very human producers.10 These insinuations tied into an ongoing genre of coverage from both techno-optimist and techno-pessimist angles that seizes on AI music press releases to predict that automated music will soon be a major source of competition for human musicians. It’s important to note that the real fact of musicians competing within Spotify against other musicians who had accepted a different labor arrangement is at least a part of what seemed to have invoked artificial intelligence for Flanagan. And we’ll return to that association in a bit.But I first just want to point out that it seems to me at least, there was no necessary correspondence to begin with between automation and the other qualities, like blandness, that tend to go along with functional or background music. The connection is maybe best conjured by the ambiguity in the term “programmed music” that usually seems closely synonymous to “functional music.” This is how Ronald Radano explained what “Muzak” means today, writing in 1989 that its genericized meaning referred to “all forms of programmed musics.”11 The notion of programmed-ness heightens an ominous specter of social control that recovers Muzak’s industrial purpose. This is the double entendre I like in the phrase “musical programming.” Is the music being programmed or is it helping someone program us? Another recent critical study of mood music, by Nedim Karakayali and Baris Alpertan, draws on parts of the Muzak corporation’s history and on a Foucauldian biopower frame in comparing Muzak to the Panopticon as “two ‘classical’ instruments of social control.”12 These authors make a convincing case that we should be very concerned about the success with which these techniques seem to have migrated into new and newly widespread digital media. But in claiming that “Muzak is the closest ancestor of online functional music,”13 Karakayali and Alpertan throw a particularly sharp spotlight on that gap that I mentioned earlier: how exactly did industrial Muzak gain a presence in entertainment media?Here is where I think Programatic Broadcasting Service can shed light on this present-day puzzle. With Programatic, the Muzak corporation not only entered an entertainment media market, but also pinned that entry on a specific form of sonic automation. Broadcasting explained that the service, which was called Muzak Radiomation Programming System prior to its formal launch, promised to provide “daily tape-recorded music programming plus basic equipment for unattended, completely controlled broadcast.”14 (The magazine was quoting from a Muzak press release15 that I was actually able to see thanks to Alix’s archival work.) “A station of average size that goes on should be able to cut overhead two to four employees through the use of the service,”16 the company predicted.Radio automation was a few years old at this point. Its basic technical conceit was that special tape players could recognize sub-audible cue tones, usually 25 hertz, which would trigger them to stop and start other tape machines. This allowed station managers to arrange a long set of elements ahead of time, with station announcements and commercials cutting into the prerecorded music sequence at the appropriate times. But though the technology wasn’t quite new, Programatic was the first automation venture to my knowledge that marketed the equipment and the prerecorded, cue-tone-embedded music tape reels as one package. This business model would become the industry standard as radio automation flourished in the ‘60s and ‘70s. And the musical aesthetics that Muzak selected for this service would ride along with that model.The same 1958 Broadcasting article described this aesthetics through a mostly negative definition. It would be “melodic, ‘entertainment music,’ almost wholly instrumental—‘the complete antithesis of the average disc jockey program,’ according to spokesmen. It also would contrast with the background music that Muzak supplies to stores, offices and restaurants. Rock and roll and jazz will be omitted.” I think this is where we can see the most explicit point of departure away from the industrial background music context, in which we saw (in Alix’s presentation) Muzak insisting that it was not entertainment, and into an entertainment context.But Muzak’s marketers were just as emphatic in assuring station managers that Programatic would offer a sharp contrast from disc jockey shows—and from the rock and jazz music that DJs were perhaps significantly more interested in playing than were their station owners or the advertisers who funded them. In their vendor listing for the National Association of Broadcasters conference the following year, Programatic used the term for this music style that would appear in most of their ads going forward: “adult music.” Also typically, the term was coupled to the equipment, emphasizing the package of hardware and programming that the stations could lease or buy as a single unit.“Adult music,” as best I can understand it, is also functioning negatively here. This was, above all else, not youth music. To understand why this would be Muzak’s angle, a quick sketch of the 1950s commercial radio context in America is useful. TV had been rapidly drawing big-name show hosts and audiences, especially youth audiences, away from radio. FM transmitting had meanwhile become an enticing new resource for broadcasters since the ’40s, but the FCC was gradually shutting down hopes that established stations could continue either simulcasting the same program on AM and FM or multiplexing their FM signals to provide “narrowcasting” services, including background music for stores. (Muzak itself had in fact been involved in some of those earlier efforts.)This left a crowded field of established radio stations looking for new ways to find a market niche, and competing for a dwindling youth audience with live rock and jazz programming didn’t seem to be a viable path; or, perhaps, conservative station owners simply hoped that it wasn’t. When industry press announced the arrival of automation in the middle of the decade, they often pointed out that previously unprofitable overnight hours, when not enough ads could be sold to justify hiring a DJ and keeping the transmitter on, could become profitable with an automated and unattended setup. The gentle, nostalgic sounds of “adult music” were in some cases described as a good fit for the late night market that Programatic invited struggling stations to pursue.But there’s another piece of this context where Muzak was able to leverage an existing advantage, and that’s labor dynamics. Not only did automation reduce employment and increase station owners’ powers over DJs and engineers; it also appears to have given Muzak a chance to dodge the usual need for union involvement as it crossed into broadcast radio. In a correspondence from the American Federation of TV and Radio Artists archives at NYU, an inquiry from AFTRA’s executive secretary Kenneth Groot ends in a hostile response from Muzak VP John Andrus, who insists that Programatic’s performers don’t fall under AFTRA representation.17Whether Groot pursued this issue further and whether he was aware of Programatic’s automation aspect are unclear. But the automation-syndication model may have given Muzak and Programatic a useful buffer from scrutiny as labor leaders and regulators caught up to this development. In any case, Muzak was able to use its existing resources to generate consistently styled recordings and to syndicate shows using non-union labor. This presumably held major appeal for anxious station owners, for whom the American Federation of Musicians strikes of the 1940s were still a recent memory.And it’s here that I want to return to that theme that we saw in the Spotify fake artist coverage, where the practice of undercutting artists’ labor power raises the specter of automation all on its own, even when the sound production itself isn’t being automated. We’ve now seen a moment where Muzakification and automation took a firm hold in popular sound media by way of Muzak automating broadcast radio. But how does this connect to our present moment and to the cultural anxieties that attend these projects? I’ve mentioned that Programatic’s automation-syndication business model became the standard for other automation companies that cropped up later in the 1960s. One of them was IGM, which stood for International Good Music. Drake-Chenault was another big name in the field going into the ’70s. A colorful full-page magazine ad from IGM in 1968