2. The Wheat of Discourse and the Chaff of Noise? What Defines a Corpus of Ethnographic Material?


Certain categories of sound contributed directly to my analysis and writing about politics. To begin with, I drew on an essentially discursive analysis of the interviews I recorded during the research. Although most of the conversations and discussions in the field were not recorded, they were reconstructed retrospectively via written notes.


Another important contribution of my acoustic material – for example recordings of union meetings or events in support of territorial claims – was that it expanded the narrative and descriptive writing about ethnographic situations that I was analyzing. In these cases, the writing covered much more than the discursive content, encompassing ambiences, intonations, dissonances, etc. 


A final, and not less important, way in which I made direct use of this varied sound material was to play it back as I was writing in order to “re-immerse” myself, sensorially, through sound, in the specific space-time of the situations. For example, when listening to extracts of union meetings, fatigue, hunger, and weariness can be heard within the language and tone of voice as well as the echoes characteristic of the gymnasiums in which these meetings were held. As Jean Copans (1998) argues, ethnography is not just a corpus to be shaped: the corpus also shapes the ethnologist, sometimes without their knowledge. These “reactivated listenings” (Augoyard 2001) were a necessary stimulus to my writing, all the more important in that a significant part of that writing was done away from Oaxaca, cut off from the social and sensory context of the research. 


However, much of the corpus of sound materials built up over the different phases of the research was not put to use during the analysis and the writing. Excluded material consisted of the previously mentioned everyday sound recordings that were made in the streets, cafés, and markets; multiple voice messages sent and received on WhatsApp (exchanges with my contacts in Oaxaca); as well as to-do lists and impromptu ideas dictated into my phone. 


There are several reasons why they were left out. First, this material did not seem directly linked with my research topic – local politics and conflicts – but more connected with both the logistics and the personal and intimate aspects of my life as an anthropologist (food supplies, time off from the research, poetic or contemplative diversions, etc.). Stored in an “everyday life” subfolder on my computer, I hardly accessed this material except to share it with people familiar with Mexican life and the characteristic sounds of its markets and street bands. Moreover, the lack of care taken in these casual recordings is quite audible, and many of the files in this excluded material are of mediocre quality. Other files, dismissed out of hand, were stored in a general folder without being named. Some of them, although automatically dated thanks to the metadata, link to situations that, ten years later, I cannot characterize precisely enough for them to be useable. 


As for the messages received and sent on WhatsApp (or other messaging platforms) with my contacts in Oaxaca, I did not really see them as useful to my writing or analyses. Nonetheless, they would have had the capacity to convey, in dialogic form, heuristic aspects of ethnographic relationships, with all the accompanying clarifications and misunderstandings: the varied ways, for example, in which I explained the nature of my research work on my interlocutors' messaging systems and how they responded or not, the delays, the omissions and questionings. The messages sent to friends and colleagues, for their part, could have revealed the progress of my research, the emotions and the encounters experienced while being in Oaxaca, as well as my preconceived ideas and analyses. These messages might have evoked the way the research was discussed and presented over time with my peers, revealing, therefore, the undeniable collaborative dimension inherent in social science research. However, I saved none of these messages, and in that respect they are truly lost documents. In fact, they both informed and carried traces of the complex processes that precede most situations reported in ethnographic writing (Crapanzano 1980; Kilani 1994), what we might call an “ethnography of ethnography” (Sanjek 1990: 385).


The whole ethnographic sound corpus mentioned above was, therefore, at the time, subject to exclusion and selection in light of its perceived relevance - a sorting process inherent to any research and writing process. Over the years, however, some of these sound archives, like Millett's whispers, have become relevant for analysis. As Anette Hoffmann points out, a temporal or political distance from the ethnographic archive makes it possible to distinguish “echoes,” “fragmented, acoustic resonances” of “subaltern speaking positions,” which provide access to a new intelligibility (Hoffmann 2023: 10). This is the case with the soundscapes of the Zócalo in Oaxaca, where careful listening makes it possible to identify voices of protest that have now disappeared or background statements in an indigenous language that were unintelligible to me at the time and which I had therefore been unable to take into account. Often, careful re-listening, informed by the passage of time and an acquired familiarity with the sonic and social worlds in question, enables us to gain a better understanding of the recordings, which then acquire a belated relevance.


The rehearsal of a marching band in the rain with handling noises


This example is typical of a daily recording made while walking through the city of Oaxaca. I was in a hurry to get home because of the rain, but I was intrigued by this juxtaposition of banda music and rain, and approached the scene. The musicians, who were rehearsing on a porch, invited me to stay. They suggested that I record them, which I did from the chair where I was sitting, close to them but not daring to move to either take out my headphones or try to get a better take so as not to disturb the rehearsal. I did not invest a lot of effort into the quality of the recorded sound, as evidenced, for example, by the handling noises. In the end, I turned off my recorder and simply listened to the music. This banda music permeates local life in Oaxaca, especially in the villages of the Sierra Norte. These villages often have a community brass band that plays at local festivals and events.