Introduction

Noise has been a subject of deep fascination to those working in the interdisciplinary field of sound studies (for a helpful overview see Novak 2015). It is a key tool in our lexicon of ways of describing and defining categories of sound, and although we may encounter these categories as apolitical and objective, they are “socially, culturally and morally informed and motivated,” and because of their subjectivity, can be considered “essentially political in nature” (Chandola 2012: 391). It is this subjective, “political” attribute of noise that is of interest to sound scholars, who hope that in studying noise they can read “the codes of life, the relations among men” (Attali 1985: 6).

Due to its subjective and political nature, formal attempts to define noise have had limited success. Novak (2015) helpfully identifies three broad “discursive contexts” within which noise takes on different definitions: the aesthetic, the technical, and the social. In aesthetic contexts, noise is a category antithetical to “music.” Here the two form a classic Hegelian dialectic (Elbow 1993), with their combination yielding the third category of “all audible sound.” This definition of noise is most commonly experienced in the context of musical performance: for example, if one were to watch an orchestra, the notes played would constitute music, while the accidental clattering of a violin-bow against a music stand or the creak of the pianist’s stool would constitute noise. Alternatively, noise may be defined in aesthetic contexts as roughly synonymous with “bad,” often used as a categorical critique of emerging styles of music (Novak 2015). In the early twentieth century, for example, jazz was considered barbarous, uncivilized noise (Moore 2007), and in more recent times hip-hop and punk have been similarly described.

In technical contexts, noise is contrasted with “signal,” and refers to any kind of interference with a message being transmitted. The “static” produced by a long-range walkie-talkie, for example, is referred to as “white noise,” while the black and white “fuzz” of a badly tuned cathode ray television is “visual noise.” More than solely technical, this definition of noise is fairly common in everyday language. Complaining “I can’t hear you – it’s too noisy!” in a busy café implies the noise/signal distinction, where the desired signal (speech) is drowned out and obscured by all other sound (noise).

In social contexts, noise is used as a descriptor for deviant sonic behavior, highlighting societal norms and delineating what is acceptable and what is not. A neighbor's dog barking late at night is “noise,” despite the fact that it is not opposed to “music” or “signal,” because it occurs outside the accepted sonic order. Its designation as “noise” implies that the night is a quiet time and denotes an expectation that dogs be kept under control.

Novak’s “discursive contexts” are in practice less different than they might at first seem. Uniting all three is the fact that noise is a relative category that can “only take on meaning by signifying something else” (Novak 2015: 2). Whether defined as not-music, not-signal or not-appropriate, noise as a descriptive term or category of sound only makes sense in relation to other terms and categories. Equally, the discursive contexts all specify that noise is unwanted, begging the question “unwanted by whom?” and reminding us that noise is predominantly a subjective category. To call something noise is to make a value judgement, and we can easily call to mind situations that use alternative classifications of sound. For example, the knocking of the violinist’s bow and creaking piano seat are the music in John Cage’s influential “silent” composition 4’33, while the “noise” of a busy café could be an asset when having a private conversation; correspondingly, the barking of a dog could be construed as “signal” warning of a potential intruder.

R. Murray Schafer takes a similar approach to Novak in his influential book The Tuning of the World (1977: 273), pointing out that while the word “noise” has “a variety of meanings and shadings of meaning,” there are four particularly important definitions. Three of these map broadly onto Novak’s discursive contexts: “unwanted sound” (social), “unmusical sound” (aesthetic), and “disturbance[s] in any signaling system” (technical). Schafer’s fourth definition, “any loud sound,” acknowledges that “noise” is also often used in common language to refer to sounds that are “particularly loud” (Schafer 1977: 273). However, while volume does indeed often play a part in the experience and designation of sounds as noise, loudness alone does not account fully for why some sounds become “noise,” not least because of the variability with which volume translates into experience. Two sounds of similar loudness may each be experienced as noise or not, depending entirely on their context (consider the loud “noise” of a passing police siren compared to the similarly loud “music” of a concert). Equally, sounds of quite different volume may be experienced in counterintuitive ways. A person snoring “quietly” in the bed next to you may seem very noisy, for instance, while the much louder sound of a commuter train might not be experienced as noisy and may even be deemed relaxing or sleep inducing. In situations such as these, there may be analytical benefit in making a separation between the terms “loud” and “noisy.” We can take “loudness” to refer to the volume of a sound and “noisiness” to refer to a sound’s symbolic capacity to be perceived as unpleasant. Therefore, a train may be “loud” but not necessarily “noisy,” while a person’s snoring may be “noisy” though not necessarily loud.

One suggestion might be that there is a point at which a sound becomes objectively “noisy” when it reaches a certain volume, perhaps one at which physical damage begins to be done to the ear. But, again, this approach misses the richness and plurality of noise as an experiential and classificatory phenomenon. As Paul Hegarty puts it “[although noise] can be loud, it is much more about what is deemed to disturb, and loudness is only part of that overall sense of noise” (Hegarty 2007: 4, original emphasis).

Seeking a concise definition, many sound scholars have described noise as “sound out of place,” a reformulation of Mary Douglas’ classic definition of dirt as “matter out of place.” This definition is especially useful, as it emphasizes the contextual nature of “noise” and, crucially, recognizes that it is a category into which sounds are placed rather than an essential quality of sounds themselves. Indeed, when reading the literature, it is striking just how often this appropriation of Douglas’ famous line takes place. Peter Bailey is perhaps the first to make explicit the connection, writing that to define noise we should “echo” Douglas and refer to it as “sound out of place” (Bailey 2004: 23). Bailey’s suggestion then reappears in the same words in Emily Cockayne’s history of dirt and noise in seventeenth Century England (Cockayne 2007: 113) as well as Making Noise, Making News (Chapman 2014: 30), Mark Chapman’s history of suffragette print culture. It also appears in Religion Out Loud (Weiner 2014: 7-8), Isaac Weiner’s exploration of religious sound and public space. Karin Bijsterveld, too, makes the connection, writing in her history of twentieth century noise that Douglas’ work on dirt and the idea of “sound out of place” “have much in common” (Bijsterveld 2008: 37). The idea is raised again in the same terms in Kim-Cohen’s In the Blink of an Ear (Kim-Cohen 2009: 111) and Garret Keizer’s The Unwanted Sound of Everything We Want (Keizer 2010: 27). Finally, Pauline Destrée makes the Douglas link in an article on the politics of noise in “estate” communities (Destrée 2013: 16), and Jonathan Willis connects the two in his book on church music and social discord (Willis 2010: 225).

In addition to the above, there are also several implicit references to Douglas in the literature, moments where noise is defined or described as “sound out of place” without specific mention of Douglas or “matter out of place.” Foremost among these is British physicist G.W.C. Kaye, who in 1931 wrote that noise should be considered “sound out of place” (Kaye 1931: 443). So similar is Kaye’s description of noise to Douglas’ description of dirt that some have subsequently described his work as a reinterpretation of Douglas (e.g. Bijsterveld 2008: 240; Keizer 2012: 27), despite the fact that its publication predates Douglas’ “matter out of place” by thirty-five years. Mike Goldsmith also describes noise as “sound out of place” in the introduction to his book Discord (Goldsmith 2012: 2), as does Roland Atkinson in an article on the ecology of sound (Atkinson 2007: 1905). Similarly, Siegmund Levarie describes noise as “sound that is disorderly” (Levarie 1977: 21), and, finally, while he does not refer explicitly to sound being “out of place,” Jacques Attali evokes Douglas when he describes noise as “destruction, disorder, dirt, [and] pollution” in his seminal work Noise (Attali 1985: 27).

Despite the seeming consensus on a definition of noise as “sound out of place” suggested by the above, there has been, to date, no detailed or systematic analysis of the link between noise and dirt, either within the field of sound studies or without. This is a problem for several reasons. Firstly, those quoting Douglas often use the “sound out of place” definition as a convenient, tidy way to conclude difficult discussions on the nature of noise, without further investigation or a dissection of the appropriateness of the comparison. This raises the potential of academics offering an inadequate definition if noise and dirt are not analogous. Additionally, by not fully investigating the links between Douglas’ work on dirt and noise, academics are missing out on the potential of understanding noise more deeply. Although Douglas wrote extensively, and literal discussions of dirt in her writing total only a handful of pages, these pages say far more about dirt than simply describing it as “matter out of place.” If the noise-dirt link is “correct,” so to speak, then Douglas’ other writings on dirt will be useful in understanding noise. Finally, exploring Douglas’ writing and opening links between sound studies and a wider body of anthropological ideas helps to preserve and reinvigorate the strong interdisciplinarity of sound studies.

In this article we argue that the link between dirt and noise is indeed productive in terms of understanding noise. By linking Douglas’ key points on dirt to selected work on noise from sound studies literature, we aim to show how each can inform the other. We begin with a brief history of the idea of “matter out of place” before giving an overview of Douglas’ work on dirt from Purity and Danger and the essay “Pollution.” This overview is not a comprehensive summary of all that is contained in each, but rather a deliberate selection of the most important sections for gaining an understanding of Douglas’ ideas on dirt. We then begin applying these ideas, taking each of Douglas’ main points in turn and illustrating how they relate to specific pieces of research on noise from within sound studies. Finally, in view of Douglas’ ideas, we argue that noise should not be thought of as “like” dirt; it is both more accurate and more productive to say that noise, in fact, is dirt: an aural type of pollution. We then point to some of the benefits of thinking of noise in this way.

• The Theory of Dirt