Thursday, May 08, 2014

Mystery in Aden: V

(Parts I, II, III, IV)

* * *

Brice Poreau, a researcher at the Laboratory of Anatomical Anthropology and Paleopathology at the Claude Bernard University of Lyon 1, has recently published a report on the 1879/80 Aden photograph that was discovered by Jacques Desse and Alban Caussé in 2010. Using biometric similarity—a technique first used in certain areas of forensic science (face recognition in muddled surveillance videos, for example), then later in the realm of art (comparing faces in paintings)—Poreau examined five pertinent Rimbaud depictions and compared them with the mustached man in the Aden photo.

Similarity biometrics is a method of measuring the distance between various points on a face, then noting the sizes and placement of the various features in relation to one another. On each of the five photographs, dozens of such measurements were made, and each was compared down to a hundredth of a millimeter.

"The first step was to establish anthropological facial points," Poreau explains. "Then 2 measurements between two points were taken with a caliper. The third step was to calculate indexes (division between two measures). For each index, we made the difference between the faces. We added these differences and divided the result by the number of index. Finally, we reported the result on a table to get the percentage of similarity." A match of 100% is nearly impossible; it would mean that the faces are identical. The two Carjat photos of Rimbaud, for example, only yield a similarity of 98% (which, Poreau notes, also demonstrates the validity of the method). Because of this, and because of the flaws inherent in comparing paintings with photographs, a match of 80% or more is considered to be a fairly convincing number.

After comparing the five Rimbaud images (shown above) to the man sitting on the terrace of the Universe Hotel, the results ranged from 84% - 92% similarity. The lowest match (84%) was with the Fantin-Latour painting, as one might expect. And the highest (92%) was with the African photo taken by Sheikh-Othman, which also makes sense as it's the only photo depicting Rimbaud as an adult. (The most famous Carjat yielded a 90% similarity.)

Poreau concluded that the man in the 1880 photo has a "very high probability" of being Arthur Rimbaud, which also means that its the first photo to "show the adult traits of the poet." As I noted previously, the identity of the mustached man in the picture will never be known with utter certainty. But this study brings us as close to certainty as we could ever hope to get.

Without forgetting that I'm taking all of this at face value—the validity of the procedure (similarity biometrics) and the expertise of its practitioner (Brice Poreau)—I finally feel like I have the confidence to say, Yes, that's Rimbaud. At the conclusion of my Mystery in Aden IV post, I wrote the following: "Personally I think there's still a decent chance the mustached man is (inexplicably) Rimbaud, but a lot of that has to do with the very unscientific and biased fact that, unlike most, I happen to think that the face on the terrace looks a lot like him (or at least how one might expect him to look)... The droopy eyes, the ear, the puffy bottom lip..." And now I—and those of you who felt the same way—have a bit of science to buttress our intuition.

Here, at last—finally and once more—is a photo of Jean Nicolas Arthur Rimbaud, taken about eleven years before his death.

During an interview, Stéphane Barsacq, author of the recent Rimbaud: Celui-là qui créera Dieu, spoke about the newly "authenticated" 1880 photo (albeit in French): "A desperate look, a dazed man. Exactly what had become of the greatest poet after his departure so long ago... It is known that this picture was controversial for reasons quite absurd. It broke the myth of eternal adolescence that the Carjat photo popularized. However, nothing was more necessary: Rimbaud is not a myth. This is a man of flesh and blood, a being who has suffered, a poet who has paid cash... Let us not forget that the Carjat photo has been retouched, and Rimbaud, as it was, found more in the portraits of his sister than in this print arranged according to a technique that anticipated Photoshop. Rimbaud's Carjat photo flatters the idea that the poet must be as beautiful as his creation."

The famous Carjat photo that Barsacq alludes to is known only from the retouched reproductions made in 1900 (the original is lost and has never been reproduced). Around 1912, however, a reproduction was made for Paul Claudel from the copy kept by the Rimbaud family, which was only recently discovered in the archives of Paul Claudel by Jacques Desse. Published for the first time on May 4th, this small, black and white reproduction is the most faithful rendering of the original Carjat out of any of the versions now circulating.

A third copy of Carjat's less famous Rimbaud portrait—which was said by those who knew him to look much more like the poet than the more famous one above—was also discovered in the Paul Claudel archives. It was apparently bequeathed to Claudel when Isabelle Rimbaud died in 1917.


No comments: