Ovid tells the story of Echo and Narcissus in Book III of Metamorphoses. Echo, the mountain nymph, falls in love with the beautiful 16-year-old Narcissus, who cruelly rebuffs her advances. Nemesis, the goddess of vengeance, punishes Narcissus by extending the very character trait that had seen him spurn Echo’s initial advances: self-love. Narcissus stumbles upon a pond, becomes besotted with his own reflection and dies following the realization that, like Echo, he will forever love in vain. It’s an incredibly visual story, the end of which sees Narcissus transformed into a flower ‘with white petals surrounding a yellow heart’, yet one not frequently recounted in the history of art. There are paintings by Nicolas Poussin, JMW Turner, John William Waterhouse, and Benjamin West, all of whom relish the scenery – a lush green forest, a pond, the two youths, a cupid or two. But in the Galleria Nazionale d’Arte Antica in Rome is a painting of Narcissus attributed to Caravaggio, a dark composition wholly given over to a dual image of the boy, leaning over the water to glimpse himself. Is this the moment in which he falls in love, or the moment of recognition that his love is doomed? Maybe the two are the same.
To glimpse your own image in something foreign to you can be exhilarating, curious, painful, infuriating. For Narcissus, it was punishment. Last week, a feature on the Google Cultural Institute’s ‘Google Arts & Culture’ app that uses facial recognition technology to match selfies with portraits from art history went viral, with swarms of social media posts devoted to dual images of users alongside their art doppelgängers. Actress Kristen Bell matched with a portrait of 19th-century conductor August Mann; comedian Kumail Nanjiani with a painting by contemporary Emirati artist Mohamed Al Mazrouei. Is the desire to locate yourself within this art historical archive the result of hubris? Is the resulting match, as with Narcissus, punishing? The punishment in the Cultural Institute’s app seems obvious: in exchange for a diversion, for the opportunity to participate in a meme, you allow Google to mine your facial data.
Facial recognition technology is now embedded into Facebook, which can notify users that a photograph of them has been uploaded without being tagged; the iPhone Photos app, which organizes the camera roll by ‘people’; and the new iPhone X, which can be unlocked with a ‘Face ID’. For years, it has been widely used to scan CCTV images and at border control, but while we have begun to accept its use in day-to-day life, facial recognition systems still feel like a new, foreign and wholly menacing technology of control. The Google Cultural Institute’s app sidesteps these anxieties by deploying the technology so openly, so obviously, that it feels anything but devious. It may also distract users from the fact that Google probably knows what they look like, anyway.
I download the app, find the selfie page, and snap a quick picture. I’m presented with a match of 62 percent accuracy: an 1890 portrait of a little Italian girl by the Danish painter Marie Krøyer, held in the collection of the Skagens Museum in Denmark. The portrait is cropped close to the child’s face. The slight estrangement that I feel upon seeing it and noting how it relates to my own image – big eyes, thin lips: a parcelling of myself into the attributes that the machine detects in my selfie – amplifies when I look up the painting on the Cultural Institute’s site. The uncropped painting shows the girl in a red dress and white bonnet, seated on a tiny chair. I didn’t know Krøyer’s work, nor had I heard of the Skagen Painters, with whom she was affiliated, a group of Scandinavian plein air artists who congregated in the northernmost part of Denmark towards the end of the 19th century. But now, thanks to my young Italian lookalike, I do, which is exactly what the Cultural Institute would say its objective is.
Do we want more people to engage with art? Yes. Do we want Google to orchestrate that engagement? Probably not. The Cultural Institute is a non-profit initiative established in order to make arts and culture available for users whose access may be limited. This goal is mediated with inferior content: authorless listicles relating to old master paintings; artist biographies that hyperlink to Wikipedia entries; introductory slideshows bearing titles such as ‘An Introduction to Caravaggio in 5 Paintings’; irrelevant ‘stories’ that seem to dote on the various institutions that partner with the Cultural Institute. This kind of content is dumbed-down, absent of quality-control and, in the grand scheme of things, unnecessary – a glance at Caravaggio’s Wikipedia page and a quick Google image search would provide a more accurate history of the Baroque master.
The outsized attention the app has received following the launch of the selfie function has, therefore, pushed a product that feigns in-depth engagement but offers little, and this is largely the result of its limited scope. There’s a reason why many of the matches are with 18th-century portraits of subjects you’ve not heard of by painters omitted from art history: the Cultural Institute’s partner list is quite limited. So, to the Instagram user who tried to match Scarlett Johansson as the Girl with a Pearl Earring (1665) with its eponymous portrait, but instead got a contemporaneous Dutch painting: the Mauritshuis that holds Vermeer’s work is not part of the Google Cultural Institute. To the art blog that uploaded a photo of the Mona Lisa (1503): the Louvre isn’t in on the joke. And to whoever matched an image of Donald Trump eating pizza with Goya’s Saturn Devouring His Sons (1823): great Photoshop work, but the Prado isn’t part of it. While the Smithsonian, the Met and the Rijksmuseum have all signed on, the list of participating institutions is far too slight to adequately represent art history’s diversity, which explains why a huge number of users have criticized the app for its limited offerings of images of people of colour, serving them with images that are often stereotypical and hurtful.
The recent increase in online engagement with art has left us with #artselfies in museums (remember Beyoncé and Jay-Z’s Louvre tour?) and, thanks to the Random International’s Rain Room (2012) and Yayoi Kusama’s infinity mirrors, a whole genre of selfie exhibitions. (This reached its logical conclusion in 2016, with the founding of the roving Instagram playground that is the Museum of Ice Cream). As with ‘Doppelgänger Week’ on Facebook, the Cultural Institute app tempts users with form of interaction with the self that is oriented towards sharing. It’s a self-representation that isn’t new: internet users have history with projecting their selves onto gifs of Beyoncé and photos of famous cats, creating a system of meaning that draws on a diversity of shared (largely popular culture) references – a diversity that creates a digital space more representative of the society we live in than that of the Google app.
The app’s results bring to mind Cindy Sherman’s ‘History Portraits’ series (1988-90), which include a number of self-portraits of the artist dressed as figures from old master paintings – Raphael’s La fornarina (1518-20), for instance, or Vermeer’s Girl with a Pearl Earring. Sherman’s self-portraits examine representation, gender and the relationship between the work and the subject it depicts. Conflating the self and art history, Sherman’s series is an example of a relationship between art and the self that goes beyond identification. The selfie function posits a different engagement with art, suggesting that today, instead of looking to art to find out something about the world, we look to it to find out something about ourselves.
Main image: Caravaggio, Narcissus (detail), 1594-96. Courtesy: Wikimedia Commons