Malte Heckelen

Computational Social Science |

Digital Humanities

Day of DH 2020: Bedroom edition

April 29, 2020, 12:12 p.m.

Surprisingly, you can work from your old bedroom that hasn't changed since you were 16 just fine.

In the middle of March, my girlfriend and I decided to relocate: We live in Aachen, which is near one of the Corona epicenters in Germany, Heinsberg. Being able to take walks in the countryside beats shuffling nervously through a crowded city in the times of a pandemic, so we packed our bags and cat and went home to my girlfriends parents. You would think having parents around and sitting in your old kid's room between heaps and heaps of old stuff would make you ineffective...and it does. But I get done a lot more these days than I thought I would. I will update this post as I go about my day.

The desk of a professional.

08:00: Some reading

Especially now, I am lucky to be able to do a research semester: no course preparation, except for a statistics workshop later this year and a (video) guest talk in a colleagues seminar. Also, no meetings today! So, I have a lot of time for my dissertation, which I quite enjoy. Aside from programming and data munging, which I will do later in the day, it's a treat to be able to read some literature again. My dissertation topic is about the evolution of professional networks as well as the character networks in U.S.-American superhero comics. It's a network analytical project that requires quite a bit of data engineering and (soon) analysis. To take a step back and enjoy some decidedly non-DH and non-technical literature contextualizes what I do in a refreshing way

Comic Books Incorporated by Shawna Kidman

One piece of literature I greatly enjoy is the one you see in the picture. I (and many other researchers) view mainstream superhero comics in a "non-romantic" way: There is no central genius controlling the story, it's a narrative the dynamics of which exist as the (often unintentional) result of a struggle of many individuals, with different interests. In this direction, Comic Books Incorporated: How the Business of Comcis Became the Business of Hollywood relates comic book history to the economic interests of entrepreneurs, distributors, marketers and others, highlighting especially the impact transmedia licensing had (and has) on continuity.

10:00: Emails, "cleaning", updating literature

My brain won't seriously start operations before noon, so I try to get as much mundane stuff done as I can before that. I have some literature in a folder called "unstructured" which has been growing and waiting for about half a year now. From time to time, I look into it, get depressed, and close the terminal. Today, I actually manage to do some work and index parts of it in my bibtex file. But I have to marvel at my ability to gather literature that I will never, ever cite...and strangely download it multiple times, nonetheless.

My home directory does not look much better. Take a look at this:


$ ls | wc -l
145

12:00: Public Speaking in times of Covid-19

This semester, I give a guest talk in a seminar of a colleague from American Studies. The talk is about social influence and opinion dynamics online, which is a side interest of mine. The seminar is about Opinion Dynamics during the times of the Founding Fathers, especially "scandalous" publishing (the seminar by Saskia Schabio covers Marcus Daniels "Scandal and Civility"), and so the connection was made. I give a broad overview of Framing and its possible effects, as well as some insight into media usage and information diffusion. It's fun, because the talks is not supposed to be particularly academic and I can work with lots of examples.

However, I have to make it a video. Have you heard yourself speaking, lately? If you teach, you probably have. Looking at the footage I just took, I notice all kinds of quirks and weird phrasings. I guess hearing yourself like this makes you a better presenter, but this gives me a new-found respect for all those people making educational YouTube videos.

Also, did you know that you can "easily" take footage from the command line with ffmpeg? I am wasting way too much time on this right now, and the purpose escapes me, but it makes my brain happy to do it this way. Here's the line I use:


ffmpeg -f x11grab -video_size 2560x1440 -i :0.0 -f alsa -i pulse introduction.mp4

You grab your X11 output with "-f", adjust the videosize afterwards and "-i:0.0" instructs ffmpeg to use the main screen (if you have multiple screens). I have alsa as my main audio driver and the input is over Pulse audio. You can also use ffmpeg ot filter out some background noise, which I might post here as well, later.

14:00: Going for a walk :)

Going for a walk is the highlight of my day.

15:00: Don't you just love old code?

Much like hearing your own voice when recording a lecture, it's fun to look at code you wrote 4 years ago. I am looking at some old data I scraped (with permission) from The Marvel Chronology Project. I have an idea for an entry to a CfP with the enticing name "Make the Superhero Great Again!". The call is interesting, because it deals with the changing perception and conceptualization of the superhero. The data I have is basically a reordering of comic issues per character, and the issue orderings within these lists reflect the chronological appearances of characters.

If you merge this data with a dataset that has publication dates, you can visualize and summarize the bursts of retroactive continuity "corrections" (or retcons) that take place over time: when has Marvel done corrections or added new information to characters that dates back a while (e.g. changing details of Spider-Mans origin for future stories)? You would need to do a lot of data cleaning, because this data is hand entry. Lucky me, I did some of this several years ago, when I first played with this data. However, look at this:

Ugly.

Some of this, I don't even understand anymore. Much of this I can do with less, thanks to the conciseness of the tidyverse (and hopefully, experience helps too):

Better.

So I will basically rewrite a bunch of code I already had. Much of the work will be in the data merge, however, matching very structured, yet faulty hand entry issue names with standardized ones from other sources like Marvel Fandom. Hopefully, it will lead to some insight (which I of course think it will). But even if it doesn't, I can use a lot of this for my PhD tasks.

18:00: Done!

That's it. Just another day in my slice of the DH universe. Now, off to get some potting soil (with mask on!), so I can do something outside for a change.