The Everyday Life in Middletown project seeks to comment on the collective experience of the everyday through diaries about the particular. They are focused while also being sprawling. We ask diarists to “write about their day” with little to no prompt and the only guidance and examples they’re provided are the diaries which are easily accessible online.
The EDLM diaries are consumed by the particular. A new song a diarist hears on the radio when arriving in Muncie. A diarist runs into a problem at the pharmacy. Diarists eating bagels for breakfast or ordering a salad at dinner. Working on my project, so far, has been about working with this corpus and trying new things to see what methods help me best understand the project using digital tools or working with the text in a visual way.
First, I worked on prepping my corpus, where I ran into several problems. I have access to all of the documents, both original and “scrubbed” (edited for anonymity) that the EDLM team uses. I cross-checked these against the website and had to fill in a couple holes and create one folder on my laptop before I could upload them to Voyant-Tools. Using Voyant, I’ve been working through some questions of method and theory while considering how to do a distance read of a corpus dependent on the particular. We’ve had a lot of these conversations throughout the different iterations of the project, so those conversations are informing my work now—weighing how to think about detail while also saying something about the experience of the everyday and of life in Muncie, Indiana.
My progress has been slow while I’ve tried to work with different configurations of stopwords, where I’m deciding what is and what isn’t important on a textual level to help more meaningful connections rise to the surface. In one approach, I used as few stopwords (20) as possible to see what emerged from that method. With that method, pronouns and words signaling the personal were the most common— my, her, she, he, I’m, etc. While this method did demonstrate how intimate and particular to the person these diaries, are, I wasn’t sure it was the most productive for moving from the personal to something more collective. What does “everyday life” look like? What is the “everyday?” How do people in Muncie, Indiana live their lives differently than in other places, in other cities both more urban and more rural.
I then tried using Voyant’s automatic stopwords, just adding a few to the list (00, like, etc.) based on what the generated word cloud revealed. This was the visual result and my starting point as I move forward.
I’ll be using the following words from the word cloud, both based on their frequency in the corpus and based on my research and familiarity with everyday life theory
- My/ Me
- Time
- Work
- Home
- Day
I’ll be using this set and adding the term “Muncie” to hopefully create a few creative ways to look at the diaries based on these terms. Some of the most useful functions so far are the collocates, to see how people are talking about the same thing but in different ways. Using the collocate function and some of the visual tools on Voyant, which I’m still familiarizing myself with, I will write a corresponding blog post for each of these for EDLM to share, as they see fit. I hope to write three for their site and to include both aspects of the distance reading while also pointing readers to the diaries themselves and encouraging people to take a look at the website. EDLM made a recent effort to encourage community bloggers and encourage this work!
I’m a bit further behind than I hoped to be at this point, slowed down by trying to use Voyant and by wrestling my corpus into something useful. I’m hoping my work will inform future efforts EDLM makes in this endeavor!
This is such an interesting corpus to be working with. I enjoyed the links you included up front in your post. They helped me get a feel for the kinds of things folks are writing about in these diaries.
One of the questions I ran into while working through this is what effects emerge from different authors in the corpus? That is, you have a corpus of posts, but all of those posts are then part of different individual people’s diaries. Another vector through this data is that the posts all occur over time, so you could also approach these documents through a temporal vector. In that context you could start to look at what kinds of trends emerge across the corpus by month.
It’s clear that you are well on you way to getting some intriguing results. It takes a lot of time and effort to get a corpus prepared for this kind of work, and now you are at the point where you can tease out interesting trends or results. The word clouds are visually intriguing, but a lot of the other tools, like the colocates, tend to generate results that take you to more interesting places.
Excited to see more of the results and analysis that you get into with this!
Intriguing, your foray into making sense out of the data provided by many diverse informants, and having to learn a new analytical program in the midst of this endeavor. As a diarist for the project, I am familiar enough with it that I readily imagine it must be daunting. I am not trained as a social scientist. Last night I attended a public discussion of the Everday Life in Middletown Project at which the project founder described the process as creating a collaborative piece of public art. It seems to me that to analyze art with forensic and linguistic tools is an honorable goal and one that may yield insights that enhance understanding of and appreciation for the artwork. At the same time, one must keep in mind that any piece of art is far more than its technique, constituent parts and basic materials.
Best of everything in your continued exploration.
Thank you so much for your comment! It was definitely a challenge, but I love the project and think it was a worthwhile way to spend my time this semester. I appreciate your contemplation on analyzing art with digital tools because we had talked at length in the earliest iteration of EDLM about that very thing! And I think that element is interesting in light of analysis, as you point out. We still want the art to “say something” if it can and to be in conversation with the Muncie community. We’re all working with words in this collaborative, public art endeavor, and an approach that considers that might get us somewhere. I hope that you continue to be engaged with the EDLM project and find it worthwhile. My final blog posts might make it onto their blog (http://bsudsl.org/edlmiddletown/blog/) if you keep an eye out!