An Introduction to Annotating and Presenting Sensitive Audio Using AudiAnnotate
by Bethany Radcliff and Kylie Warkentin
Annotation, an important paratextual genre, has expanded past the typical text-based work in recent years to include things like web-pages and online resources. Further, the rising popularity of social annotation movements in educational practices (especially during the move to more interactive, virtual learning activities as a result of the COVID-19 pandemic), means that understanding and using annotation is more relevant than ever.1 Audio annotation is particularly important when that audio material contains sensitive or traumatic content. This lesson serves as an introduction to annotation as a method for understanding sensitive audiovisual resources, structuring annotations in a machine-readable format, and uploading and presenting audio annotations using the AudiAnnotate application. While much of this lesson will focus on engaging with sensitive archival material in individual or group settings, it may also be read as a general introduction to annotating with AudiAnnotate. Throughout the lesson below, we have included suggestions for modifying this lesson for use in a classroom setting.
In this lesson, you will learn how to create and present annotations using AudiAnnotate, focusing in particular on working with sensitive archival material. This lesson will guide you in thinking through and creating annotations , how to use software like Audacity or a simple excel sheet to structure those annotations for web upload on AudiAnnotate, and how to use AudiAnnotate to collect and present those annotations.
You will need to set up a GitHub account in order to use the AudiAnnotate application. You might prefer to download an open-source application such as Audacity to assist in generating annotations with time-stamps in the steps below, but, alternatively, you can also mark time-stamps manually on a spreadsheet. Our documentation walks you through creating a GitHub account, downloading Audacity, and creating a spreadsheet for use in AudiAnnotate.
Though any publicly-accessible sound can be used for this lesson, this lesson is structured and based on “‘Criminal Syndicalism’ case, McComb, Mississippi,” a sound recording centered around Civil Rights issues in 1964 that is part of the John Beecher Sound Recordings Collection at the Harry Ransom Center. We have chosen this recording not only because of its historical value, but also because it speaks clearly to current reckonings with systemic racism in 2021. Like many historical recordings, this recording contains sensitive content that students and teachers often find difficult to discuss. In this case, this content includes racial slurs and descriptions of the experiences of Black students who were falsely arrested and imprisoned. Please read the full content warning and summary below before listening to the audio.
We have structured this lesson in a way that seeks to be mindful of trauma-informed pedagogy when working with sensitive resources. Trauma-informed pedagogy, developed by Janice Carello and Lisa D. Butler, emphasizes the importance of ensuring the physical and emotional security of learners by emphasizing the importance of contextualizing content and acknowledging and validating any challenges learners face when working with sensitive materials.2 Importantly, the goal of trauma-informed pedagogy is “to remove possible barriers to learning, not to remove traumatic, sensitive, or difficult material from the curriculum.” Likewise, we believe the unique challenge of working with sensitive materials should not deter anyone from learning from them. A trauma-informed approach to working with these materials allows both independent researchers and groups of researchers to empathetically and carefully work with materials others may have difficulties with that are nonetheless important to history. For further discussion of trauma-informed pedagogy and other inclusive practices when working with these sensitive materials (not necessarily in the classroom!), we recommend Emory’s “Inclusive Pedagogy: Discussion and Resources”.
In general, when selecting the audio material with which you’ll be working, you should review the holding institution’s language in describing and categorizing the audio for cues in handling, listening to, and ultimately presenting your chosen audio material. While some holding institutions may not have content warnings or descriptions of sensitive content for individual items, other institutions will have some kind of acknowledgement of troubling content or metadata listed either on the holding institution’s main site or collection page for the item. We recommend you look towards these places for language cues first.3 For example, the Harry Ransom Center contextualizes the Beecher Sound Collection here, but at this time does not include a specific content warning related to the racist language in the tapes. The Ransom Center’s statement on outdated language shares a general warning that users may encounter sensitive, offensive, or outdated information in descriptions of archival materials.
About the Beecher “Syndicalism” Recording
This audio reel was recorded on October 19, 1964, likely by John Beecher’s wife, Barbara Beecher, on the portable reel-to-reel machine she operated. At the time, John Beecher, the great-great nephew of Harriet Beecher Stowe and Henry Ward Beecher, was working as a journalist, reporting on the Civil Rights movement in the South. The recording begins with John Beecher and members of Council of Federated Organizations (COFO) meeting with Black high school students and their parents after these students were released from jail in McComb, Mississippi on charges of criminal syndicalism, in which it seems the students have been falsely accused of damaging property by throwing a brick. Beecher speaks with the group about the importance of voting and fighting for equal treatment in the voting process. A white member of COFO discusses how his experience in jail was different than that of the Black students, how he was released without bond, and that his release was sooner. The students discuss the conditions in the jail and the treatment by policemen and guards. The parents of the students discuss how they were discriminated against, given limited interaction with their children, and how officers made visits a challenge.
In this recording, a racial slur is used at 16:06 by a student quoting the language said to them by a police officer while in jail. There is also explicit language used at 15:49, 16:30, and 16:39 by students quoting white police officers. At times in the recording, African American students and their parents are discussing mistreatment and trauma they experienced while the students were imprisoned.
This recording is hosted on the Harry Ransom Center’s CONTENTdm site, but the direct link to the audio (while publicly accessible for listening and downloading) is not supported by AudiAnnotate at this time. The Ransom Center provided a working link to the audio, which you will need when creating your AudiAnnotate project in later lesson steps.
Classrooom Suggestion: For classroom settings, make sure you follow your school’s guidelines for classroom work involving potentially upsetting or controversial material. While we recognize the value and necessity of having difficult conversations, we also recommend discussing class expectations repeatedly and prior to working with these types of materials. You may also choose to modify the lesson introductory materials to add the necessary context for the material your class is working with based on your class’s learning objectives. To minimize student harm and risk When working with sensitive audio, we recommend having a secondary recording for students who express a high level of discomfort to work with that may explore similar themes, but not include explicit or triggering language. The Studs Terkel Radio Archive provides a number of related resources and audio recordings geared at classroom use.
Part 1: Listening
The first step to annotating audio is listening to the whole recording before listening to any selections to understand the context of the recording (though in the classroom setting, we recommend listening to a short portion of the tape. See note below). Listening is a different method than reading or watching and will accordingly require different methods of approach with different demands on time and attention. As you might carefully read and re-read a passage in close-reading, audio requires close-listening.
For audio with sensitive moments, like the one we have chosen for this lesson, listening can present a unique challenge. By pressing play on an audio recording, you become a sort of “witness” to the moments taking place in the audio. Communication and media theorist Amit Pinchevski, writing about audiovisual testimonies of Holocaust survivors at Yale’s Fortunoff Library, acknowledges the place of the listener as witness and says unlike text, the recordings “impose on the audience a complex audio-visual narrative, which calls for the development of collective skills of interpretation and active engagement” (p. 258-259).4 Further (as with other kinds of media), exposing listeners to potentially sensitive or upsetting material with no warning may forcibly shift their focus from the material to their reaction provoked by the difficult content, making any new learning nearly impossible.5
We have structured the lesson in a way that is aware of this mode of listening and witnessing, and acknowledge that the potential of trauma is an important thing to keep in mind for listeners. Note that we have broadly described the Beecher “Syndicalism” case (informed via historical and social context) and provided an audio content warning noting where racial slurs occur and their context prior to any discussions of listening and annotating. You may wish to follow a similar approach if listening to sensitive audio, particularly when listening to it with others. Further, be aware that this audio recording was created in 1964, so some highschoolers recorded on the tape may still be living. Similarly, many archival materials, like the one we have chosen, have connections to the present that can be unsettling or surprising, particularly if the same sorts of trauma described or captured in the recording is something listeners are still experiencing today.
With this in mind, take some time to listen to the recording, or the portion you will be annotating. The steps below will help you think about ways to create and organize annotations.
Classrooom Suggestion: For teaching in a classroom setting, we recommend focusing on a 1-2 minute portion of the audio to ensure students have sufficient time to listen, process, and engage critically with the content. In the Beecher recording, we recommend the clip from 14:07-15:37 for classroom discussion, where the recently released students discuss their limited recreational privileges and lack of sufficient nutrition and have many surprising moments of levity. This particular clip does not contain racist slurs, but due to the overall sensitivity of this recording, students should still read the content warning and summary before listening.
Part 2: Annotating
There is not one particular way to annotate audio, and the methods you use for crafting annotations might vary based on your choice of audio and your own preferences. The steps below can be a helpful guideline as you get started.
After listening to the audio recording and deciding what part of the audio you will focus on, begin to take note of some general features you noticed.
Focus first on understanding the situation, events, and participants in the recording before moving to moments of meaning making. After listening to the audio, it may be helpful to think about the following questions as you begin to understand the events of the recording:
- What characteristics or events did I hear? (People talking, horns honking, clapping, etc.)
- What stuck out to me in this recording? (Moments of laughter, tension, modern relevance, etc.)
Classrooom Suggestion: The above questions could serve as a whole-class discussion starter after the class has read about the audio and listened to the clip from 14:07-15:37.
Pushing a little further, you can make notes, answering the questions below to better understand the specific events in the recording and begin to add your own perspective/interpretation. While we would not recommend getting lost in the granularities of noting the exact time stamp where your responses to any of the following prompts occur in the recording, it might be helpful to note generally where they occur to assist in formatting your annotations in Part 3. As mentioned previously, Audacity is a useful program to concurrently annotate and generate timestamps. Documentation on how to use Audacity for annotation is available here.
- What did I notice about the events and the people in the recording? (People speaking angrily, movement around the room, fluctuation in applause, etc.)
- How would I describe this?
- What did I notice about the recording itself? (Clipping, increases/decreases in volume, etc.)
- What are some topics or themes associated with this recording? (Topics, speakers, etc.)
Any of your responses to these questions and those previous can become annotations. It’s important to realize that annotating is almost always a slow and tedious process, and will likely involve listening and re-listening to your audio sample, perhaps even in multiple sessions. This may not be an easy process with any audio recording, but might be especially challenging with more sensitive audio. Focusing on a portion of the audio as suggested above can make the annotating process more digestible.
You might also consider the challenges of listening to and interpreting the audio as you annotate. For example, you might ask yourself:
- What makes annotating this recording challenging? (volume changes in tape, emotional moments, lack of context, etc.)
- What context am I missing? What am I uncertain about? (location of meeting, time of day, unsure of speakers, challenging to locate context about specific events, unidentifiable noises, etc.)
- What might I research more in-depth as I consider this recording? (the students who were arrested, more information about the events surrounding the meeting, what is criminal syndicalism?, etc.)
A helpful framework for writing annotations can be to think of “categories” for annotation features and can make the annotating process less overwhelming and more generative. As you begin to think about and notice the events and moments you might annotate, you can begin to think of categories that you might annotate. For example, if you notice environmental sounds (like cars horns or birds in the background), you might create a category or list for annotations in the category of environmental sounds. If you are adding annotations that focus on transcription of the audio, you could create a category for transcript. This method will be helpful later when you begin formatting your annotation layers for uploading to AudiAnnotate, as described below.
Classrooom Suggestion: In a classroom setting, students can work in groups to sort general observations and descriptions of audio into “categories” for annotation, deciding on these category titles as a group. Students can be put in groups/breakout rooms to decide on what annotation “categories” to create annotations for and then move to working together to write annotations.
The questions listed throughout the “Annotating” section could serve as a way to press discussion further and allow students to think deeper about how and what they might focus on in annotations. The more specific questions in the last section regarding challenges of annotating that link observation to interpretation could be posed to students as prompts during group work to brainstorm annotations. This step can be combined with Part 3 (below), and students can annotate directly into a formatted Google spreadsheet in their groups.
Part 3: Formatting Audio Annotations for AudiAnnotate
AudiAnnotate accepts structured annotations for upload in .tsv format. There are many methods for formatting your annotations. You can use a software application, such as Audacity, or a simple spreadsheet. Details on how to do so can be found in the AudiAnnotate documentation.
Below is a set of sample annotations created for the Beecher “Syndicalism” recording project found on AudiAnnotate. The layer titles (“Environment_Kylie,” “Transcription_Bethany”) came from categories (like the ones described in the previous section) described and decided upon by the creators of the project. Note: Each .tsv file is one annotation layer in AudiAnnotate. If you have multiple annotation layers, you will need multiple annotations files.
Classrooom Suggestion: After working together to choose annotation categories students can begin to create or format their annotations. The students can create annotations in a spreadsheet as shown above, using our Google sheets annotation template with formatting instructions. We suggest each student contribute at least one sentence of transcription and one commentary annotation, though we of course encourage further annotation.
Part 4: Uploading and Presenting Audio for AudiAnnotate
After you have created your annotations, you’ll add them to your AudiAnnotate project. To do so, you will sign in to the AudiAnnotate application (after authenticating GitHub) to create a project.
As mentioned above, this recording is hosted on the Harry Ransom Center’s CONTENTdm site, but the direct link to the audio (while publicly accessible for listening and downloading) is not supported by AudiAnnotate at this time. The Ransom Center provided a working link to the audio which you will need when creating your AudiAnnotate project.
Documentation for creating a project, adding a link to your audio file, and uploading your annotation layers to your AudiAnnotate project can be found here.
Once you have created your project page and added annotations, you will now be able to view your project page with the annotations you created alongside the audio. You can continue to add annotations to this presentation page and you can invite other individuals to comment on your annotations using Hypothesis. Instructions to do so can be found in our documentation for collaborative projects.
Because this particular audio is sensitive and has potentially offensive moments, we recommend adding a content warning above your project by editing the introductory text. You can find steps on how to edit that text in our documentation. The introductory text is also a great place to add any other contextual material, commentary, or provenance necessary to introduce your audio project. Remember: context is important when exposing large, varied researchers to sensitive material.
Classrooom Suggestion: Because AudiAnnotate presents audio and associated annotations publicly on the web, there are a few different ways to implement this final step in a classroom setting.
First, the teacher could manage project creation and focus the student lesson on generating annotations. Students could then use Hypothesis to comment on each other’s work, facilitating further discussion of the audio recording and the annotation process (see Chapter 6 of Annotation, by Remi Kalir and Antero Garcia for further discussion of learning extensions with annotation).6
A second approach to using this lesson would require that students have more time in class to grow familiar with working with AudiAnnotate and GitHub. Here, the teacher could create a collaborative project on AudiAnnotate by following the steps in our documentation. In this scenario, students would have more autonomy in adding layers and annotation to the final product (vs. the teacher as the final reviewer in the previous scenario). Here, too, Hypothesis would be a valuable tool in adding comments and suggestions to other projects. The teacher may even choose to use Hypothesis as an evaluative tool, adding comments, supplementary material, or grades as necessary.7
1: Kalir, J., Cantrill, C., Dean, J., & Dillon, J. (2020). Iterating the marginal syllabus: Social reading and annotation while social distancing. Journal of Technology and Teacher Education, 28(2), 463-471.
2: Carello, J., & Butler, L. D. (2015). Practicing What We Teach: Trauma-informed Educational Practice. Journal of Teaching in Social Work, 35(3), 262-278. doi:10.1080/08841233.2015.1030059
4: Pinchevski A. (2011) Archive, Media, Trauma. In: Neiger M., Meyers O., Zandberg E. (eds) On Media Memory. Palgrave Macmillan Memory Studies. Palgrave Macmillan, London. https://doi.org/10.1057/9780230307070_19
5: Hammond, Z. (2015). Culturally Responsive Teaching and The Brain: Promoting Authentic Engagement and Rigor Among Culturally and Linguistically Diverse Students. Corwin, a SAGE company.
6: Kalir, R., & Garcia, A. (2019). Chapter 6. In Annotation. Retrieved from https://mitpressonpubpub.mitpress.mit.edu/pub/451exgdi