Blog Post

Searching for Meaning in Search Terms

Close-up view of some search keywords from We Are Searching For. Photo by Mimi Onuoha

uk zip code. octet truss. ameliorating. look mum no hands. tumblr br not working. time in edmonton.

These are some of the words that I have entered into my browser as search terms since having moved to London. There’s a story behind every phrase, and I remember each one perfectly. I searched “time in edmonton” while trying to schedule a time to talk with fellow Nat-Geo Fulbrighter Ann Chen; “look mum no hands” is the name of a coffee/bike shop here in London that I used to visit. “Ameliorating” is a word that I always forget the definition of, the line break tag in Tumblr really doesn’t ever work, and zip codes in the UK are much different than zip codes in the States. With me to provide context, each search term becomes a self-contained and easy-to-understand story in itself.

But if I weren’t around to explain, the combinations of those words would seem much more nonsensical. If I had provided even more of my unfiltered search keywords, the results would read as an absurd cacophony of randomness, a gathering of lots of whats with no explanations of why. Even still, they would begin to hint at something.

Data Minus Context

Larger view of We Are Searching For. Photo by Mimi Onuoha. For more images, see here.

My project for the Fulbright-National Geographic fellowship involves using people’s personal data to create maps of relationships that stretch across online and offline space (the project has evolved since I first started, see my other post for more information on what has informed those changes). The project’s underlying presumption is that we can learn previously unknown or unconsidered things about ourselves and our relationships by paying attention to our data trails. In other words, examining digital data can flesh out understandings of ourselves.

But I’ve also seen examples of when that situation seems different. Occasionally the data crumbs that I’ve been examining don’t actually appear to clarify much—at least, not on first glance. The keywords database is a great example of this. Most browsers save and store the queries that users have typed into the search bar, and it is relatively easy to access those results. I took a look at my own search terms from the past few months and found the process to be equally hilarious and disjointing. Just as the first line of this post has shown, the phrases seemed like nothing more than random gibberish. But upon closer inquiry, I began to see that there was something universal about them—they weren’t completely divorced from meaning. Even without explanation the words did hint at my mindset, frustrations, curiosities, and the (often embarrassing) things that I’d been wondering about. They stood on their own as unflinching snapshots of moments in my life.

We Are Searching For

Setting up for the show. Photograph by Mimi Onuoha

I found myself captivated by this ability of search keywords to simultaneously reveal and obfuscate. I also was fascinated by just how easy they were to access—assuming one knew how to access them, and provided that people weren’t regularly deleting their browser histories, it was theoretically possible for anyone to take an unfiltered peek into anyone else’s searches. Whether or not this was a negative or positive thing felt much less pressing than pointing out that it was a possible thing.

So I ended up creating work that did just that. I wrote a script that would collect keywords from various browsers, and using that, I went around to 39 of the Royal College of Art’s (RCA) public computers and gathered all of the words that people had searched for. I aggregated all of those words and presented them as a print called “We Are Searching For”, which I showed in the RCA’s Work in Progress Show.

The print contains 1,353 searches from those computers, and it prompts a few questions: though all of the words that I displayed are publicly available and the people who searched for them are left anonymous, there’s still a sense of voyeurism upon seeing them. This suggests something like a breach in privacy. So what is the line between public and private? Furthermore, all of the terms share in common that they were likely typed by someone at the Royal College of Art. In that sense, they function as a reflection of the community; what would the same aggregation look like for other institutions?

Of course, the terms are also entertaining in a variety of ways—they include phrases range like “why is most of sci fi movies based in future los angeles”, “obamas daughters pregnant”, and “that wasn't jumbo parsley?”.

The WIP show concluded on Sunday, February 8, so the piece is no longer available for viewing, but if you want to learn more about the work and the phrases that people searched for, check it out on my personal site. And if you’re interested in seeing your own keywords, you can either access your keywords using any SQLite database manager, or you can use some code that I've posted online. This gist shows some very simple Python code for accessing keywords on Chrome or Firefox (the code is for Mac OS; if you want similar code for PCs or other browsers, just let me know).

Next week I'm back to working on more explicitly project-specific things, but as always, feel free to reach out to me on Twitter, my project Tumblr, or in the comments below.

Back to Top
About National Geographic Society

The National Geographic Society is a global nonprofit organization that uses the power of science, exploration, education and storytelling to illuminate and protect the wonder of our world. Since 1888, National Geographic has pushed the boundaries of exploration, investing in bold people and transformative ideas, providing more than 15,000 grants for work across all seven continents, reaching 3 million students each year through education offerings, and engaging audiences around the globe through signature experiences, stories and content. To learn more, visit www.nationalgeographic.org or follow us on Instagram, LinkedIn, and Facebook.