When I first saw that I occupied a small nodule in the network, I thought it was a badge of pride—I’m an outlier who chose different picks than anyone else. But the explanation isn’t so simple.
When filling out the quiz, I only chose 7 tracks. This happened because I included two non-“musical pieces” in my Golden Record curation: “United Nations Greetings/Whale Songs” and “Sounds of Earth” (though how are whale songs not music, am I right?). The reason I ended up with 7 instead of 8 tracks is because the music titles sometimes had different names—an English description versus their original language on the podcast page. It was only later that I discovered there was a YouTube link with all the Golden Record tracks, and these names corresponded to the ones on the quiz. One track got lost in the mix because I didn’t want to spend time cross-referencing names.
So the only reason I’m an outlier is because me and two other colleagues chose LESS than the required 10 songs, and I chose the LEAST at 7. My opportunities for connections were greatly impacted. I think it’s good that the quiz didn’t force you to choose 10 songs, as that would have misrepresented what I thought should go on the record (I misinterpreted “music” to mean only musical tracks and not tracks in general—but I want whale songs with Louis Armstrong!). But this creates a limitation in the network.
This exemplified the fact that the visualization can fail to capture the reasons behind my different engagement with the task, which has significant political implications. When someone lacks access to information (like me missing the YouTube page) or doesn’t have the cultural proficiency to complete something in the expected way, they appear differently in the data visualization – but this misrepresentation doesn’t actually reflect their true preferences, values, or identity. It’s a powerful observation about how data collection processes can systematically misrepresent people based on access barriers rather than actual differences in perspective or preference.
I was in a 17/17 facet group with Sourabh—a lonely group of just 2 people—and because there were only 2 of us, none of the three songs we shared became nodules or circles. At first I thought you need at least three people sharing a song to get a circle, BUT when I combined our 17/17 group with the 29/29 facet group, Track 14 (Melancholy Blues) only had Isabella and me sharing it, yet it was still a circle. So circles must come about due to the interrelationships and connection density. This visualization choice of which selections become circles further reinforces whose preferences get emphasized and whose remain peripheral, which connects to broader questions of representation.
These visualization patterns reveal deeper power dynamics at work. My choices were really established by the task parameters, yet I wasn’t given much weight in this network. There’s no way the visualization can reveal that it’s built on the hidden assumption that whale songs aren’t considered ‘music,’ but that assumption effectively misrepresented my actual preferences and pushed me to the margins of the network. There’s almost a power dynamic embedded in the assignment that implies whale songs aren’t music and dictates what constitutes a music record (spoken word and whale songs cannot go on a music album; flute and drums can).
Also, while the visualization shows our preferences, it can’t show our motivations. One person might choose Bach for his mathematical precision, another because of his position in the Western canon. This small-scale example gives me insight into a much more serious problem: how marginalized groups might be misrepresented in data visualizations not because their perspectives aren’t valid, but because the data collection methods themselves contain barriers that disproportionately affect certain populations. What you’re measuring isn’t musical preferences but people who completed the task “correctly” from one particular perspective.