Tanya Clement’s recent book Dissonant Records offers an important and welcome intervention in technologically conscious research that aims to engage with literature beyond the printed or manuscript page. Clement’s contribution to the burgeoning field of literary sound studies is unique for its function as “a form of advocacy for the continued, responsible use of historical audio artifacts in literary study” formulated through the theorization and explication of “close listening as a method.” The main concern of Dissonant Records is “to make possible new studies of different stories in literary history by listening.” This work demands reflection on the status and affordances of audio materials held in archives for research, the formulation of concepts that help frame the work of listening, and the cultivation of what Jonathan Sterne has termed “audile techniques”, or approaches to listening, that enable the effective and ethical use of such archival records.
The following interview with Tanya Clement was conducted by Jason Camlot in his office at Concordia University in Montreal on Friday, December 14th, 2024. Clement’s visit to Concordia contributed to a larger research project called Literary Listening that Camlot had begun that year, pursued with the support of a Social Sciences and Humanities Research Council of Canada Insight Grant. This research project is designed to demonstrate the importance of sound studies for literary study and explore the implications of theoretical concepts and practical approaches to listening from other disciplines for the discussion and study of literary works and ideas of “the literary.” The theory informing the research of Literary Listening suggests that by unpacking the history of both mature and nascent examples of audile techniques (approaches to listening) in the history of literary studies we can better understand and reflect on the cultural formations that develop around something we may not even think about as part of the discipline, that is, listening, and better explain how we listen, why we listen that way, and what implications and consequences come with the inclusion or marginalization of certain methods of listening. The interview of nearly three hours conducted in the morning included a “life story” interview that focused on her memories of having engaged in practices that entailed methods and reflection on listening from her childhood through her education and formation as a scholar of literature, computational methods, and the digital humanities. It was followed that same afternoon with a listening practice designed and led by Dr. Clement in which she invited participants to select and play short sound recordings relevant to their own research projects so we could discuss how we listened to them, and practice applying some of the concepts and listening terms explored in her recent book, Dissonant Records: Close Listening to Literary Archives (MIT, 2024). The excerpt from the longer interview selected for this special issue of Amodern focuses on the methodological import of several of those key terms from Dissonant Records discussed as concepts that are useful for critical listening to archival audio, namely, the imperative, “amplify”, and the nouns, “distortion, interference, compression, and reception.”
–Jason Camlot
So, to start, maybe you could say a few words about the book, and then I’ll flag some passages, ask you to read them, and we’ll talk through them. But first, I just want to say — I’m really interested in the metaphors you’ve chosen for thinking through modes of listening. The way you frame listening as a method for attending to sounds we usually ignore is fascinating. In a way, the method you’ve been developing since sixth grade [as discussed in the first part of this interview] is a way of hearing things you hadn’t heard before and appreciating when someone teaches you to notice something you hadn’t noticed before. Your book is a really complex and rich realization of concepts we can use to frame that kind of practice — especially in relation to sound objects in the archives. That would be my summary, based on your life story and having read the book. But maybe let’s hear your quick summary of what Dissonant Records is.
Well, it came from doing research in audio archives — basically, hearing things that scholars weren’t writing about. And I was curious about that. But it also came from my digital humanities work — especially in making audio collections more accessible and discoverable.
Back in 2010, I read an amazing piece, sort of a manifesto, written by the Council on Library and Information Resources (CLIR) and the Library of Congress in the U.S. It was all about endangered audiovisual archives. The main argument was that if people don’t use them, libraries and archives won’t save them — because they can’t save everything. If they don’t believe these recordings are valuable to the public or the historical record, they won’t prioritize them. They just don’t have the resources to do everything. That struck me because, in my experience, people often don’t use these archives because they’re not accessible. And since they’re not accessible, people don’t use them. It’s a chicken-and-egg problem.
That became an important reason to write the book — to say: If I write a book showing that literary audio recordings are essential for understanding famous literary figures — figures people have been writing about for decades, yet haven’t examined their recordings — then maybe that could be an intervention. Maybe people would start taking recordings more seriously in scholarship and teaching. And maybe they’d start noticing the gaps in the record — the things we haven’t been attending to, but should be. And asking: Why haven’t we?
That was the impetus for the book. That’s also why I chose recordings by Zora Neale Hurston, Ralph Ellison, Gloria Anzaldúa, and Anne Sexton — these are all major literary figures. But the recordings I selected aren’t their performances. They’re not stage readings, not necessarily them performing their work. My criteria were that I wanted recordings that were essential to their writing practice. I wanted to argue that these recordings were deeply integrated into why they wrote what they wrote. That was my goal. And I’m sure your next question will be about some of the definitions in the book. Those definitions came from actually going into the archive and trying to listen to things that weren’t made very accessible or discoverable.
I love thinking of this book as a preservation project. We’re both engaged in digital preservation projects in different ways. But thinking of the book as a scholarly intervention that could motivate preservation efforts — is a very useful way to frame it.
And as a side project, through my digital humanities initiative AV Annotate, I’ve created online listening experiences for the recordings. The ones I can make available online accompany the book’s chapters.
I also love that your definitions of dissonance and difficult listening modes take into account not only the sounds within the recordings, but also the contexts in which listening happens — the material and institutional infrastructures and structures, the non-sonic barriers that shape how we listen. So, if I were to ask you — one of the first things I did when reading the book was make a list of the metaphors you use as concepts for methods of listening.
Resonance.
Yes, I was going to ask you to start with resonance. So, I’ll pass you the book when we get to passages I think are interesting to read. You can decide whether to read them or just talk about them. You can decide whether you want to read it or just talk about it — or read it first, then expand on it. I’d like you to read some of them because, even though you just wrote this book, hearing yourself read it might make you think about it in a new way. The yellow dots mark the beginning and end of the passages I have in mind.
Okay. Yeah. So, this is resonance. [Tanya Clement reads the following passage]:
When I hear someone say that a person, place, thing, or event resonates with them, I understand the term as a placeholder for a sense or a feeling of meaning for which the variables of causation are too numerous or too complex to articulate exactly, but for which the speaker has a presumed sense of matter and of what matters — of significance. Consequently, I use resonance as a capacious theory to help identify properties of meaning-making and how and why the literary media archive registers. (Dissonant Records, p. 7)
What I like about the term resonance is that it is actually a physical property. There are three tenets of resonance that I focus on in the book and find generative. The first is that for something to be resonant, there has to be a materiality to it — it has to ping off something. Whether you’re talking about the ear or some other object, resonance is a physical property. For example, if someone sings at a certain octave and a glass breaks, that’s a physical resonance. Or when a swing moves and reaches that moment of levitation — that too involves a resonant frequency. The resonant frequency of an object determines how much that object responds to movements in the air around it. So, in your ear, you have a resonant frequency when you hear something. But resonance isn’t limited to hearing — it’s something you can feel. That’s why, if a train rumbles by, you can feel the vibrations in a desk. It’s not necessarily the earth moving in that case — it might be — but think about a plane flying overhead. Obviously, the earth isn’t moving, yet the windows shake. That’s resonance.
Yes. Or like that known story of Edison listening to music by biting —
Yes —
the piano. And he’s basically experiencing the resonance through his teeth latched to the wood.
Exactly, exactly. So, that was the first property — that resonance is inherently material. Secondly, resonance is dialogic. For something to resonate with you, it is a two-part experience, whether that’s a material interaction or a conversation. Resonance is inherently dialogic. You aren’t by yourself. The last aspect is entelechy — or intellechy. I don’t even know how to say that word!
I was going to ask you.
I can say it when I see it, but not when I’m not reading it. That term means that all these elements are coming together at once. I use the analogy of being in church. I reference bell hooks, who talks about being a child in the South, sitting in church, listening to the preacher speak. Her experience is shaped not just by the preacher’s voice and the way it resonates in the space, but also by who she is — a young Black woman — and the ways she reacts socially and personally to that moment. So resonance is not only dialogic and material, but also situated in place and time. It has history to it. I use these three aspects of resonance to think through how listening always involves these factors. A listening practice, if it is a form of close listening — which I would argue is an interpretive practice — should account for all these aspects. But it’s almost impossible, right? When you make assertions about meaning-making, and any of these aspects are absent, what exactly are you arguing? What is your basis for making that argument? And yet, we can never fully know all these things. We are always going to fall short in our interpretations.
Yes. And especially challenging in that multifaceted scenario of listening is the question of positionality. I’m thinking of the bell hooks example, but also of how you integrate the idea of agency into the practice of listening. Could you talk a little about agency?
I really enjoyed thinking about agency in this way. One of my reviewers really didn’t like Karen Barad’s work and was upset that I referenced it, but I found Barad’s ideas on agential realism really useful. She explores how agency isn’t just about human control — it’s about how meaning is made through material interactions. Agency isn’t only about the listener or speaker; it could be about the playback machine, the recording itself, or even the way someone listens to themselves. Where does agency lie? Where does the power to shape meaning exist? It’s not always with the scholar. Moments of agency are moments to which we should attend because they offer insight into meaning-making. This connects to digital humanities as well. When you digitize a corpus, you make choices that affect meaning-making down the line. Choices are being made all the time, even if we don’t consciously recognize them — whether it’s how a machine is tuned, how a library structures access to materials, or even how metadata is created. Throughout the book, I argue that these moments require us to think about authority, validity, and the entire process of interpretation. This is why I find this work so engaging — it feels like an open field where we can intervene. Yet, at the same time, resonance occurs in specific ways at specific moments, which constrains that openness.
One of the amazing things about your book, and your definition of close listening — quite different from Charles Bernstein’s [in his introduction to Close Listening: Poetry and the Performed Word, 1998] — is your focus on method. In literary studies, we often think of close reading as something we just do, without questioning the methodology behind it. But recently, scholars like John Guillory and Jonathan Kramnik have started unpacking close reading as a method of interpretation rather than an inherent practice, in works such as Professing Criticism (2022) and On Close Reading (2025) by Guillory, and Criticism and Truth: On Method in Literary Studies (2023). Your approach to close listening contributes to that discussion — it challenges the assumption that close reading isn’t a methodological act. Would you say that’s fair?
Yes. In fact, while you were talking, I was thinking about a section in the book where I describe my own convoluted listening practice. I think I actually know what page it’s on. Want me to find it and read it out loud?
Yes.
It’s from the Preface. This is the beginning of it. So, I write:
bell hooks asserts that personal, small details are “often what grabs people,” what “makes theory seem as it does for me, to have concrete application.” Yet listening to details can be a slippery method. Listening is messy, inexact, and trying to listen for meaning — no matter who or where you are, no matter the subject, topic, or medium — is a method of interpretation that requires repeated listenings: stopping the recording, scrubbing back over what’s been heard — if the old CD player even lets you do that — and listening again. Even when discs, tapes, and records are made accessible, their meanings may remain difficult to discern. Recordings can be hard to “read.” Listening for how social and material dissonance resonates in archival records is slow work — especially when these records have historically been dismissed by scholars as unimportant, irrelevant, or non-evidential “hearsay.” Trying to browse or skim while listening is a sure way to miss something important. Spoken words happen in real time, sequentially, and sampling such sounds for meaning is like choosing a random page of a book to reconstruct the synopsis of its plot — the devil of the plot is in which details, not just any details. (pp. xii-xiii)
And so I go on to discuss mishearings, misinterpretations, and mislistenings that I notice — and I’m sure there are many I don’t notice at all. So, to me, listening is a method that is deeply personal and situated in who you are and what you listen for. Other scholars have pointed to the racial implications of listening in certain ways and attributing meaning. What I focus on here is the material situation — listening to something on a CD player versus digital playback, where you can jump to a timestamp. Older CD players used to divide the entire disc into 10 tracks — you couldn’t see where in the recording you were, and you couldn’t navigate between sections. That was my experience listening to Anne Sexton’s recordings at the Schlesinger Library at Harvard. They had placed the recordings on an old CD player. If I misheard something, I couldn’t go back and find it. And so what does that mean for scholarship, where exactitude is considered fundamental to rigor? How do you maintain exactitude when listening isn’t the same as copying a word exactly from a page?
That’s what Kramnik argues in his book — it’s not the last word on the topic, but one of his main points is that close reading is actually a form of writing. His argument is that close reading is enacted in our excerption and integration of quotes into our own language. We’re essentially creating a new work out of it — a fusion of the original text and our own analysis. But the difficulties of navigating recordings make it even harder to perform the interpretive practices we’re used to in literary studies.
Right. Yeah.
Can I have the book back?
Yes, you may.
So, we’re going to go through some of the other categories because they’re all equally fascinating and relevant. If I were to ask you — what is the key disruptor in each of your chapters, in order, could you list them for me?
Yes, I could. The first one is an imperative — Amplify. The rest are nouns: Distortion, Interference, Compression, and Reception.
Right. Let’s go through them — is that okay?
Sure.
I’ve noted a passage to begin with — Amplify. A passage I think is relevant — you decide.
Yes, yes, yes. The reviewers didn’t like that the other chapters were nouns while this one was an imperative. But I did it on purpose.
Right. What didn’t they like about it? The lack of symmetry?
They thought it would be confusing to readers. I don’t care. Readers are probably smarter than that. Or if they have a problem with it, I explain my reasoning in the book. It’s not a mistake. Titling this chapter with a sonic metaphor as an imperative, I argue that Amplify demonstrates why close listening across genres is a necessary and urgent methodology for historical reparative archival work and literary study.
So, how is amplification reparative?
In that chapter, I focus on the Tulsa Race Massacre of 1921. It’s a horrific event, but also a fascinating case study in how historical records operate. Originally called the Tulsa Race Riot, the dominant narrative was shaped by white historians. The official story was that the Black residents of the Greenwood district rioted, and in response, white residents burned down their community. The truth is, Greenwood was a thriving Black-owned community — one of the most prosperous at the time — and white mobs destroyed it. There were even planes dropping bombs.The event was systematically erased from historical records — newspapers ignored it, documents were destroyed, and for decades, libraries contained no references to it.
It wasn’t until the 1970s, when amateur historians began recording oral histories from survivors, that we regained insight into what happened. Today, much of our understanding comes from these recordings. The chapter argues that we must amplify these original oral histories to understand contemporary representations — whether in literature, TV, or media. Without close listening, we risk distorting history again and what is happening in those recordings, which is kind of interesting. But I’ll let people read the chapter to understand how we are representing that event today. And so it sets up the methodology for the book in a way to say literature is an important part of that. The literature that people are writing or the stories that they’re making in the TV show are important reflections of what is coming out of these historical writings that are supposed to be fact, but are also based on these oral histories that weren’t necessarily very factual in the first place, because you had survivors in the 1970s, when Tulsa was still very segregated, being interviewed by white interviewers. And those were the historical documents that were really feeding much of the current understandings of what happened.
What is it about documentary recordings, like AV documentation, that lends itself to particularly strong kinds of amplification in the way you’re describing it?
Well, I think that all the things that we’ve already mentioned, there’s, you know, in that example — and, like any good humanities scholar, I’m picking examples where I can highlight the things I’m already interested in — but in that particular example, those recordings, you have this material resistance, and that’s where this dissonance comes in. That’s why the book is called Dissonant Records, because it’s about how these records, these AV artifacts that serve as evidence of some sort, create moments of dissonance. And my definition of dissonance isn’t necessarily unpleasant — it’s a moment of reflection. Dissonance doesn’t mean something is wrong or bad. It means something is different than your expectation. When you perceive a sound as dissonant, it’s because you were expecting a different sound. Oftentimes, especially in rock history, dissonance serves a purpose. It’s not always unpleasant — if you’re from a different culture, certain sounds may not sound dissonant at all because they’re part of your musical tradition. So dissonance is fundamentally about perspective and expectation. And all of these factors come into play with those particular recordings — the material format, the way you access them, their historical context. One of the recordings is between two women — one is a white amateur historian, the other is an African American woman who was a child at the time of the Tulsa Massacre. Their conversation is polite, but deeply layered. At one point, a train comes by, and for anyone familiar with Tulsa’s history, the train tracks were the physical and symbolic divider between Black and white neighborhoods. So, suddenly, you’re thrown into the historical and racial dynamics of the interview. Even though the interview itself is polite and uncontroversial, the situation — the history of segregation, the fact that a white historian is leading the interview — becomes significant. And of course, this is my interpretation, my filter that I’m applying to the recording. That’s why this particular chapter helps me unpack the key issues I explore in the rest of the book.
It’s definitely there in every chapter. Let’s continue to another concept. I would move to Distortion. Is that okay? I think this is a very interesting passage.
[Tanya Clement Reads]:
Distortions are a twisting awry or out of shape. They are what Zora Neale Hurston would identify as “off-key” from an expectation. Figuratively, distortion is a word misapplication, “a perversion of words so as to give them a different sense”; in sound, it is “a change in the waveform of a signal by an electronic device such as an amplifier or during transmission from one point to another, usually impairing the quality of its reproduction.” Distortion happens during the process and production of use, but distortion is also a perception. An audio engineer listening for a certain audio quality might describe distortion as having “a little sense of unease” or a feeling that “things aren’t quite right” while an untrained listener might not notice an aberration at all. Similarly, a person with a limited vocabulary might not perceive that a word has been used incorrectly. Besides, distortions are not without their pleasures. Sound distortions in music can be expressive, just as linguistic malapropisms can be illuminating…A diverted expectation can be an opportunity for reflection and for reconsidering the processes by which perceptions of what is “right” occur. (p. 53)
So, distortion is a very interesting kind of dissonance that you talk about in this book.
Yes, it’s an invitation to think about distortions of different kinds.
When I read that example about mispronunciation, it brought me back to when I was a junior faculty member. There was a job talk for a position in the department, and I knew one of the candidates — he was from Montreal, a working-class guy. He mispronounced an author’s name in a way that I can’t even remember now — maybe I blocked it out because it was so “horrifying” in that performative context. But I saw immediately how that one distortion sunk him completely. In every other way, he was probably the best candidate. But that distortion was unforgivable to my colleagues at the time.
Often, distortions are social in form. And that’s what I talk about in that chapter with Zora Neale Hurston — distortions aren’t just technical. I mean, those are old recordings. They’re from 1935. So there are actual distortions that make it difficult for me to hear her voice. But then there are also these, I would call them, social distortions that make it difficult for me to hear her “voice,” in a broader sense. And so you kind of have to think about those two things at once in order to understand what it means to make meaning from her recordings.
If we were to take a step back for a second — okay, you’ve just fleshed out distortion conceptually, how would you describe it as a part of a listening method? What is distortion as a factor of your listening method?
I think in some ways, these are all moments of dissonance. All the concepts I describe are moments of dissonance and moments of reflection. If you sense a distortion — like someone mispronouncing a name — it’s not necessarily a distortion of that name itself. It’s a distortion of your expectation of what that person should have said. And these are moments of reflection. Whether you think, this person isn’t right for a job because they mispronounced a name, or whether you realize, oh no, my colleagues are going to judge them unfairly for this mistake. Whatever the motivation behind that reflection, it’s about recognizing those moments. It’s the same principle you teach students when analyzing poetry. Or in my case, when I tell my daughters: if you don’t feel safe in a situation, walk out. It’s about those moments where something doesn’t feel quite right. And those are always generative moments — moments where you should stop and pay attention to what isn’t quite right. Is it your expectation that’s off? And often, it is your expectation that determines what seems right or wrong. So I find distortion a useful term for thinking through these moments. Because listening practice can be so many things, I tried to introduce these metaphors as hooks — ways to stop and reflect in the interpretive process.
And there are particular formulations of specific kinds of dissonance. So it’s almost an anatomy of dissonances — not necessarily conclusions, but ways to identify different kinds of disruptions.
Right, right.
That’s one of the unique contributions of this book — you’re offering a framework, a kind of anatomy of different issues that trigger these reflective moments. Let’s move to the next one.
Well, and I also — while you’re finding that one — I wanted to bring up Veit Erlmann’s book Reason and Resonance: A History of Modern Aurality (2010). He makes a comparison between Enlightenment thought — typically seen as a very visual movement — and asks, what if we thought about that period through sound metaphors instead? That idea was useful to me. I wanted to choose metaphors that were unusual. Because I thought they might help us see things differently. I wanted to elevate these terms — dissonance, distortion, resonance — as levers for interpretive study.
Yes. And even with all the qualifications Jonathan Sterne has made about the audio-visual binary and his critique of Walter J. Ong — one of the most powerful things about sound studies is the way it disrupts the dominant visual interpretive modes in literary studies.
Right. Yeah, that’s great.
Distortions. I’m just going to read another passage I forgot to ask you to read. It says, “But distortions on the recordings may result in sound that is missing or garbled, or may introduce noises that audio engineers decide to smooth over or delete” (p. 57). I like that passage because it highlights how we need to attend to the ways distortions get erased — sometimes deliberately.
And those deletions matter. In the Zora Neale Hurston chapter, I reflect on what was deleted from the historical record because it was seen as a distortion. But maybe those were the moments where she was actually speaking in a way that revealed her agency.
Yes. I’ve seen this in archival recordings too. Sometimes, the before and after of a literary reading gets cut, but those moments are revealing. But in some collections, we find unedited versions — like recordings that capture the informal conversations before and after an event. Usually, when you digitize an analog tape, you don’t get access to the parts that were physically cut. But in digital-born collections, those fragments sometimes still exist.
Yeah. Like the bell hooks example — where she talks about snacks and Sprite. There’s even a third voice in the background. And I ask, who is that person?
Next is Interference.
So, interference is a little cheeky. It’s a very technical term. And I had a lot more about it, but one of the reviewers was like, it’s too much. So I figured if I get just the part I have in here mostly right, other people could look into interference as a technical term and all the things that it introduces. So, here I describe it like this:
A phenomenon described in acoustical theory, interference occurs when two sound waves occur simultaneously. If both sounds are peaking, they are in phase, and their wave amplitudes will add together, making the wave peak higher. This kind of interference is considered constructive interference and leads to “live spots” in a listening environment. In contrast, destructive interference occurs when sound waves are “out of phase,” when their superimposition in time includes the trough of one wave and the peak of the other, cancelling out both sounds. Instead of an amplification, destructive interference results in “dead spots.” where signal blocks out signal, creating perceived silences. (p. 78)
Do you want to give one just sort of quick example of that as a useful term for describing something you’re hearing?
That chapter is about Ralph Ellison. He was on the speaking tour in the 1950s after he’d gotten the National Book Award for Invisible Man. In the United States at the time, it was the beginning of the civil rights movement, and there were court cases about desegregating schools, a lot of politics around race. The recordings I talk about in that chapter are recordings that Harvard has just recently located, where Ellison was invited to give a lecture at Harvard in 1953, the same year he won the National Book Award. The curator of that collection and some other evidence indicate that he was probably the only African American participant in the conference and possibly one of the few in the room when he gave his lecture. This chapter is about not only the lecture he gives — some of which ends up in essays he would later write — but also the extent to which his participation in these conversations around his lecture influenced those later essays. That’s the gist of the chapter. But in this context, I use interference to open up moments of silence. There are a few moments on these tapes or in his lecture where I argue that he’s deliberately avoiding certain topics, particularly the politics of the time. Although he does include one anecdote that doesn’t show up in any of his essays, which I think is fascinating. He talks about walking into a room in the South that had only recently been desegregated, and a white woman was physically horrified by his presence. He ameliorates it for his audience, who I argue were probably mostly white, by saying, “It’s not her fault, this is how she was raised.” He takes a very impersonal stance on it. I don’t know — I’m not Ralph Ellison, I’ve never been Ralph Ellison — but I think there were many ways in which he had to navigate these kinds of situations. So, I talk about the silences in his lecture and the way he employs interference, where moments that could have amplified racial politics are instead muted. There’s another moment on the panel when an audience member challenges a remark made about a character in Invisible Man and basically accuses the scholar of making a racist comment. The scholar fumbles, trying to joke and cover up the situation. Then he turns to Ellison, possibly trying to draw him in for support or to diffuse the moment. But Ellison either A) doesn’t respond, or B) his response is drowned out by the laughter of the audience reacting to the scholar’s defensive remarks. In that moment, I use interference to highlight how Ellison either chooses silence as a form of response — perhaps to avoid escalating the situation — or he does respond, but his voice is literally covered up by the noise of the audience. This kind of interference, where silence or competing sounds obscure meaning, is key to understanding the dynamics of that conversation.
There’s another case where Ellison speaks to a primarily Black audience and references Tulsa — something very few of his biographers even acknowledge. He had visited Tulsa before and after the massacre, and it deeply affected him. In this speech, he only has to say something like “Tulsa, 1921,” and the audience understands. He doesn’t have to explain it — there’s an unspoken understanding in that moment, unlike his Harvard talk, where he seems to be navigating a very different kind of audience and expectation. And so there’s this way in which I can read into it as a moment of using silence in a more powerful way, because to go into it would be stupid for his audience. I mean, it would kind of diminish the impact of it for his audience. But to not talk about it and just be like, “We all know what that is,” actually amplifies it in a way. That’s a case where you have a peak and a trough — it’s amplified, or there are two troughs and it’s amplified, but it’s the peak and the trough that actually create interference.
Your description of your approach in this chapter and your other chapters spells out something that’s kind of unique in these documentary recordings that you’re listening to. As you said, they really do sound the sociality of the events. And your use of these metaphors, to read the details, it’s like you’re technically breaking down the social communication patterns as though they were sonic processes. It’s a very powerful method for talking about the sound of the social in documentary recordings.
I think — well, we can go to compression — but skipping forward to the last one, reception, where I really talk about my own listening practices quite a bit, I try to break that down by saying: is any of that necessarily true, just because I’m saying it? Even though I’m doing it in a systematic way that maybe makes my readers think it’s true that this is what happened — I have no idea if this is why or what Ellison did. Or why he did it or what he was thinking — I have no idea.
No. And reception in that last chapter emerges from your reading of Anzaldúa as a kind of medium, a receptive medium herself, or something like that.
Yes.
But I think it’s also a core chapter for thinking about your positionality within your method. I think we should move toward wrapping up because we need to make time for lunch before you lead us in a listening practice. But there is a passage — I do want to skip over compression completely, partly because it draws on Jonathan Sterne’s work as well, and I think you are making a very interesting use of it. I’ve actually used it metaphorically for interpreting other sound recordings. I think it’s a very powerful metaphor for thinking through certain kinds of sounds and recordings.
Well, I think a lot of what Jonathan has done has been a model for me. I just — I love his work. And his use of compression in that article is what gave me the idea: “Oh, maybe I could do that with each chapter and pick something else.” So, it was important to me to include compression. And obviously, poetry is a form of compression. And that’s why it seemed important for thinking through Anne Sexton’s work. So, I write: “In Communication and Media Studies, compression has been defined both as a condensation of thought in language and as a technical term for data compaction or dynamic range in audio” (p. 91). And that’s a quote from Sterne’s essay “Compression: A Loose History,” some of that. I continue, “Most importantly for discussion of Sexton’s therapy text, Jonathan Sterne has associated compression practices with a modern anxiety over a loss of meaning based in assumptions about technology and its ability to produce verisimilitude” (pp. 91-92). And then I go into Sexton in particular. “So, as this chapter demonstrates, Sexton was deeply motivated by the relationship between the compression practices that produced her therapy text and her ability to create meaning through poetry” (92). And I should define the therapy text for just a second. So many people know this, but Anne Sexton’s therapy sessions were recorded for many years, which a lot of people have written about. Usually, they write about the things that she was talking about. But what I was mostly interested in was this practice that she created around these recorded therapy sessions. So, her therapist required her to — she would go there twice a week. There are about three years of recordings, all at the Schlesinger at Harvard. She would go there twice a week, and her therapist had decided that she wasn’t getting any better because she couldn’t remember what she’d said session to session. So, he wanted her to record the sessions, and then, in between sessions, she was supposed to go home and write about what she heard in a journal. Then she would go back to the next session and talk about what she had written about, what she had heard in the journal. It was all very circular and meta. And in this three-year journal period, she also wrote one of her Pulitzer Prize-winning collections of poetry, Live or Die (1966). I was interested in how this process of listening to herself listening to herself might influence her poetry. And I thought it was fascinating that no one had thought about that. No one had written about that. And so that’s what I’m trying to suss out in that chapter.
Yes, great. I want to jump to one last quotation. Wait, let me find it.
Is it about reception?
No. I think we could talk more about reception, but I think you’ve already talked about it in a lot of ways throughout. Here, it’s this one. And this is more about you. Starting with “I have become…”
Oh, yes, this is in the coda. Should I set up the coda?
Yeah, okay. Set up the coda.
Okay. So, the coda came from a reader — I actually know who it was, but I won’t say — who asked why this book isn’t about digital humanities at all and wanted me to write a chapter about how the work I had done listening to audio archives impacted my understanding of digital humanities in the context of sound studies and audio archives.
That kind of annoyed you, right?
And it was really — I found it really annoying because I said to myself, “Well, this book isn’t about digital humanities. I’m doing something else.” But in retrospect, it was good because it offered me an opportunity to write a forward-looking chapter that ties this work into that field. What I write in this coda is:
I have become more interested in the so what? of computational analysis with historical spoken word text in the archives. To the reviewer’s query, then, Dissonant Records as a whole is my response: interpretive practices with audio are difficult and subjective and always already entangled with the personal, cultural, sociopolitical, and technological context of listening as an agential practice….I find myself interested in distant listening insofar as its pursuit has resonances with larger questions about epistemologies, ontologies, and other areas of study concerned with better understanding dissonances in the cultural register and imaginary. Distant listening is interesting to me to the extent that its processes hold traces of these entanglements. (p. 151)
So distant listening, in a way, suggests that doing DH [digital humanities] projects — big DH projects, which are extremely time-consuming, expensive, technical, and require a lot of collaboration — isn’t necessarily about achieving some grand result or truth, but rather about learning little things about your own methods along the way. Which is research.
Right? I mean, that’s what we’re paid to do. That’s why we get these paychecks — because we are meant to try out these things, think them through, and consider their impacts. And so, its in this way, among others, that my work in large-scale computational analysis with audio archives has had an influence. I’m glad I wrote the coda because it helped me think through small details. To go back to what bell hooks was saying at the beginning, those small details are what resonate with people, what shape meaning. And I find that the most generative digital humanities projects are the ones that require me to go back to those details and think through, even in more traditional methods, why I am making meaning of these small things as opposed to other things. Because when you’re doing machine learning, you have to learn what the machine can perceive; otherwise, you can’t ask it questions. And so in that context, I’ve had a lot of experience with humanities scholars who want the machine to do very human things. And you have to explain to them, “That’s not how a machine works.” Machines count. And if you can’t count the thing, the machine is going to have a really hard time replicating or representing that thing.
Yes. And often those assumptions — those misunderstandings about machine learning — reflect an ideal they’re hoping to realize. Like, “I know what I want to happen in my interpretation…”
Yes.
“Can the machine make it happen?”
“Can it verify what I already think?”
Exactly.
Yes. And, you know, this is an old Jerry [Jerome] McGann quote that I always come back to: “You don’t know what you don’t know.” With machines, if you don’t already have a hypothesis — if you don’t already know what details you want to explore — it’s hard to get the machine to find them. You have to go through a discovery process like I did with The Making of Americans by Gertrude Stein. I threw it at a whole bunch of different algorithms before I was able to see patterns that I understood. But I’m sure there were patterns I didn’t recognize because I didn’t have the capacity to understand them as patterns.
So, one thesis from our discussion might be that DH has been extremely generative for you — not just for thinking about works in new ways and seeing what others have missed, but also for making you self-conscious about the assumptions shaping your discoveries in the first place. DH has made you highly aware of “method” throughout your work, which is somewhat unique in literary studies.
Well, that reminds me of the beginning of our conversation — how [as a child] I was always aware of learning because it never seemed to fit me quite right. And in the ’80s, if you didn’t think the way others thought, you just fell behind. But I wasn’t willing to do that. So, I was constantly asking myself, “Why don’t I understand this? How can I approach this differently?” So, that hypersensitivity to how interpretation and understanding happen has been a common thread throughout all of my experiences.
Yes, absolutely. The coda also seems to signal a farewell, in a way, to DH methods. I don’t know, maybe I’m reading too much into it, but I’m curious — if you were to decide to enter a new phase and maybe not put all your time, as you’ve been doing for many years, into large-scale digital humanities approaches, what would you do? What are some of the approaches you’d be interested in pursuing if you weren’t doing that?
I think digital humanities remains very interesting. I think it’s going to change significantly in the context of generative AI. And I think with the way machine learning is developing now — where learning models are so big and so far beyond human comprehension in a way — I think that makes them less approachable for an individual humanities scholar. So, it does require these big collaborative projects, and those take a lot of work and energy. I’m happy to support other people doing that work because I do think it’s important. But I wouldn’t mind — and I’ve started to talk to my own department about this — going back to a more traditional literary classroom where we’re talking about scholarship and creative writing from more situated practices. Because I do find that computational models tend to take that out of the equation — the situatedness of the creative endeavor. And I’d like to dive back into that.
Right. I think, in some ways, for distance-oriented computational models to work, you have to take a particular thing and transform it into an entity, a much abstracted, and also reduced, form.
Yes, there’s this interesting quote that I use at the beginning of the coda from Hurston, where she talks about folklore being “the boiled-down juices of life.” And I think about that as a metaphor for machine learning, but it still doesn’t quite work.
No, because “boiled-down juices” suggests all the richness that machine learning may actually leave out.
Exactly, exactly.
Well, thanks, Tanya. This was a great conversation. I have learned a lot listening to you reflect on how we may listen in the archives.
Article: Creative Commons NonCommerical 4.0 International License.
Article issue images produced by the media installation All We'd Ever Need Is One Another (2018) by Adam Basanta.