SynasthesiaMe-ing the Mona Lisa
A friend tagged me with the SynasthesiaMe-ification of his name, and said I might find it interesting. He was right. This post is the result of the first collision impact of me and that tool. (Thanks, Brett!)
SynasthesiaMe
If you haven’t heard about it yet, the author Bernadette Sheridan took her lifelong grapheme-color synesthesia experience—in which letters in names appear as colors—and created a tool that allows anyone to type their name and see the colors that result. She provides a great overview graphic of this process in her Medium post about it, titled “What Color Is Your Name? A New Synesthesia Tool Will Show You.”
Head to https://synesthesia.me/ and play around with it if you like to get a sense of how it works. Then come back.
Why might this be interesting to me?
For…wow…20+ years now, I’ve been simmering some ideas about semantic noodling systems, or meaning machines, that randomize signs inside of a fixed grammar to produce new meaning. The deep structure plays out in sortilege (a particular form of divination), lots of entertainment examples, and, when applied right, to creativity and design. I spoke about this most recently at UXLX in Lisbon. So if you want an overview, here’s that.
Anyway, SynasthesiaMe is just such a meaning machine. The randomization is the accidents of parental psychology, culture, and language that gave you the particular spelling of your names, and mapping it to a color table. Names in, colors out. Meaning machine. (Compare it freely to Margaret Magnus’ Gods in the Word.)
Having spent so much time thinking about these things (and, I know, not actually publishing them. I’m sorry, I’ve been…busy) I had ideas around how SynasthesiaMe could be extended.
- Could it provide the new meanings explicitly, like “Oh, Christopher, here are your colors, which feel like springtime in the foothills of the Southern Alps,” or the like. Even just a list of adjectives.
- Could it do an internet-wide image search of, say, famous paintings that are similar to your colors? “Christopher, your name connects you to this Jan Brueghel the Elder still life.” Do you want to learn more?
- Could you reverse the engine, and put in images to find out what they should be named?
The Mona Lisa
I had a bit of unexpected free time over lunch, so I ran an experiment for that last idea. Is the Mona Lisa actually “Mona Lisa” in Sheridan’s mapping? Let’s find out.
First, let’s find the colors of the Mona Lisa. That’s easy. Grab the image from Wikimedia Commons, drop it into Photoshop and reduce it to Indexed Colors. “Mona Lisa” has 8 letters, so let’s set the number of colors to 8.
Sadly, these did not map to the https://synesthesia.me/ output of “Mona Lisa,” like, at all.
Which might be the end of it. Or not. I suppose you could go and re-run this experiment across all of the painting’s alternate titles, “Monna Lisa”, “La Gioconda,” or “La Joconde,” and find which one is the “real” name, but that would involve some custom coding to find the “distance” between two .act files, and I was committed to this line of reasoning only for the duration of a lunch.
So instead, I thought, what if we tried to uncover the “actual name” by piping the original image’s colors back through Sheridan’s map? That map was not published in the post, but it’s easy enough to simply type out the alphabet in the tool and screen cap that. (n.b. There are only 26 characters in the map, as diacritics all fall to their base letter. This makes our search easier, if a little more Anglocentric. Poor Benoît.)
Dropping that into the Photoshop file, I used the Color Range tool to find the closest match for each.
Mapped back, the 8 colors became the 8 letters VOGNYUXU. While the pronunciation of this speculative name is its own a challenge (/vɐg-ɲy-kʒʉ/?), we are not stuck with it.
The order of the colors in this “true name,” and therefore the order of the letters, is arbitrary for our purposes, so I ran VOGNYUXU through an anagram server. It didn’t exactly go well.
- Gnu Vox Yu
- Gun Vox Yu
- Vug Nox Yu
- Vug Noy Xu
- Vug Yon Xu
- Vug Un Oxy
- Vug Nu Oxy
- Guv Nox Yu
- Guv Noy Xu
- Guv Yon Xu
- Guv Un Oxy
- Guv Nu Oxy
- Yug Von Xu
- Yug Un Vox
- Yug Nu Vox
- Guy Von Xu
- Guy Un Vox
- Guy Nu Vox
Of that list, the one I can make the most sense of is “Guy Un Vox.” You can read it as a patchwork of English, French, and Latin for “Dude (with) One Voice.” That kind of fits the conspiracy theory that the Mona Lisa is DaVinci’s self-portrait as if he were a woman. (Hers may be a nonbinary smile!)
So finally, I applied SynasthesiaMe’s 8 colors back to the Wikimedia Commons image of the original painting (via a forced color table Index Color mode, for Photoshop nerds) which gives us this lovely version.
Intriguing, isn’t it? Like, I genuinely like looking at it, for all its strangeness. And I guarantee you that had I been set free to reinterpret the image without these Oulipan-esque algorithms, I would never have created anything like this. That’s a nice demonstration of the power of meaning machines.
Anyway, I hope you enjoy this little journey investigating and extending SynasthesiaMe’s color map. The mapping, the meandering steps, the iterative feedback—even pushing the image back through SynasthesiaMe to remap the original image—are all examples of semantic noodling for creativity.
If you want to know more, give that video above a watch. If you want to pressure me to finally put the concept and related patterns to writing, I don’t know, like and share this, yell at me some, and help me find some more free lunches?
Happy noodling, and thanks for the inspiration, Bernadette!