Cringey Tweets from Connecticut: Twitter and my High School Worldview

Facebook has really been profiling me as a queer person lately and sending me super targeted ads, mostly for lesbian dating websites and LGBTQ health studies. Among other things, it’s gotten me thinking about how much data Facebook has on me that it gets to analyze, and how little access I have to intelligible data about my own online life. There’s so many of my thoughts out there in the universe, including lots of thoughts from years ago that I don’t remember having, that these big companies know I had. I was posting online back when I was still a straight girl who loved field hockey, spoke only one language and had a passion for Hillary Clinton (all now embarrassing facts about my past), and if Twitter and Facebook and Tumblr all know about strange high school me because they have the chops to make sense of all of the huge mass of posts I created, it just seems like I should have access to that information as well.

Weirdly enough, I do have a social media time capsule that starts right when I came out to myself and ends before I left the US in 2016 and started developing a more global awareness and radical politics. I have the timeline from my twitter account @thetinamitchel, which I used from 2013–2016. Who was I? What my worldview in this weird time in a small town in Connecticut before I was out?

Obviously these are questions that I can try to dig back inside myself and answer, but it’s totally hard! It’s hard to connect to and empathize with someone who tweeted “I am a cerealatarian” into the void (me, circa 2014, no likes or replies), and I have to face the truth that most of my memories are clouded by 2021 me judging high school me. Using my twitter data gives me a new way to look at myself, through a more stable set of data than my memories, and computational methods allow me to comb through this data without having to read every single tweet (which would overdose me on cringe).

So I’ve scraped my tweets from this account using the web scraper twint. My account has 67 deleted tweets that I’m not able to recover, so the dataset I’ve created doesn’t include all the things I apparently decided I needed to hide, but if my memory serves me, most of the deletions are from when my iphone cracked and started automatically opening twitter and tweeting without my consent, so the dataset is probably not missing too much critical information about my past.

To get a better sense of how representative my tweets are of the ‘high school Tina’ I’m investigating, I’ve charted them over the 4 years I was tweeting.

Here are my tweets over time:

Nice! Looks like I was active for most of high school, with an early peak in 2013, which makes sense since in my memory I got busier and busier as high school went on and used social media less. But I was active enough in each year that I think this twitter account is pretty representative of things I wanted to shout into the universe throughout high school.

One way to consider my worldview during this time is to understand where I was thinking about. Towards this goal, I ran Named Entity Recognition on the tweets and pulled out location words. Named Entity Recognition is a method that creates tags for set categories of words in the text, identifying people, places, cardinal numbers and more. It was effective for me in quickly pulling out the proper names of cities, states, countries, and a landmark, though I did have to delete a few words that were mistakenly tagged.

Ready? Here’s the list of recognizable places I was tweeting about:

Budapest, Vienna, America, London, the Grand Canyon, New York, Abu Dhabi, China, Connecticut, Spain, Alabama, Fort Lauderdale, Florida.

Many of these are not surprising to me. I’m from Connecticut and I definitely remember a family vacation to Alabama some time in my adolescence. But… why was I tweeting about Abu Dhabi? Or Vienna? Places that I’ve never been? Places I thought I’d never really thought about?

Pulling up the tweets that include ‘Abu Dhabi’ I find, among others, “Is it even possible to not feel a deep desire to attend NYU Abu Dhabi???” and “NYU Abu Dhabi though.” My Abu Dhabi tweets are about NYU! This is a dream I did not remember I had.

What about Vienna? Looks like I tweeted “New mode of procrastination: obsessively looking up train tickets from Budapest to Kraków and Vienna.” OH dear, this is from a very specific time in my life when I thought I would be going to study in Hungary, which then did not happen.

Based on this list and these examples, it looks like my globe was really small (in my 2021 eyes), and yet full of dreams of going beyond my small town.

To get a more comprehensive vision of what my world looked like (and how I was looking at it), I’ve mapped my tweets that mention geographical locations. Follow the link to my interactive map and click on a dot to see one of my tweets about that location!

https://sleepy-allen-273a7a.netlify.app

For me, the biggest thing I notice on this map is how empty East Asia is. In the 5 years since I stopped tweeting, I’ve spent two years in Japan, started studying Korea, and extensively dreamed about taking a trip to the Russian Far East. This seems like an obvious statement, but it needs to be said — in high school, I didn’t know my life would go in these directions!

Looking at this map makes me more empathetic to my high school self. Especially as I eschew her political opinions (which I do, I really do) I feel like I can understand more where they come from. This is the world of knowledge and dreams that I lived in, because it’s what I had access to, and I wanted to go beyond it. Actually, it makes me kind of proud to see how much I believed that institutional education would be my way of accessing the world, and that’s something that really worked out for me, allowing me to study several languages and study in several countries that I couldn’t imagine in high school. Basically, this map says to me that I was pretty stupid, but I had a pretty good concept of how to get un-stupid, and it worked out.

So… HA Twitter, Facebook, and Tumblr, I’ve succeeded in knowing data-driven things about myself. True, they’re things you probably already knew, but I have the context of my memories to make more sense of these facts than you ever could… unless you are tracking me on Medium.com, and now have that context as well, which is a distinct possibility (I’m not very vigilant at reading tracking permissions).

Now that I’ve conquered Twitter, it’s probably time to go back to Facebook and try to figure out what kind of sense I can make of all of my data over there, which spans both high school and college. I wonder if ‘high school Tina’ will seem like the same person as she does on Twitter — maybe I’ll have to readjust my conclusions.