The Rosslyn Code Unearthing a hidden melody.By Chris WilsonUpdated Wednesday, May 18, 2011, at 7:01 AM ET
The Rosslyn Chapel's 213 stone cubes were carved when Leonardo da Vinci and Christopher Columbus were schoolboys. Construction on the chapel began in 1456, about 50 years before the printing press arrived in Scotland. The Stewart kings ruled the country at the time, and most historians credit James IV—who took the throne in 1488—for ushering in an era of scholarship and scientific inquiry. By that time, the chapel's founder, William St. Clair, had died, and construction on what was meant to be a much larger structure had come to a halt.
Given the era in which Rosslyn Chapel was built, then, it would be surprising to learn that someone encoded scientifically inspired symbols on the walls. But that's exactly what Tommy and Stuart Mitchell came to believe. The chapel's stone cubes, they were convinced, looked like Chladni patterns, the images that form when musical frequencies vibrate along a two-dimensional surface. Now, they just had to confront the inconvenient fact that Ernst Chladni was not born until 1756.
James IV, who's generally credited as Scotland's first Renaissance king, didn't assume the throne until four years after Rosslyn Chapel's founder died. Click image to expand.James IV, who's generally credited as Scotland's first Renaissance king, didn't assume the throne until four years after Rosslyn Chapel's founder died.
Chladni's date of birth didn't rule out the possibility the Mitchells had a match. Chladni patterns are a natural phenomenon, and it's entirely possible that someone else discovered them hundreds or even thousands of years before the German scientist documented them. All you need to produce these patterns is sand or salt, a flat surface, and some way to make the surface vibrate. You didn't need to be an 18th-century European to make that happen.
Besides, as far as Tommy and Stuart were concerned, they could test their hypothesis without tracking down Chladni's Scottish predecessor. If they could match each of the 12 distinct symbols repeated across the cubes to a Chladni pattern—assigning each symbol to a musical pitch—they would have a 213-note sequence. The proof would be in the melody: If it sounded like something other than an incomprehensible jumble of sound, the Mitchells would have compelling evidence that the chapel's designer had transcribed a song in Rosslyn's stone walls.
But assigning notes to the stone cubes was far from a simple task. Chladni patterns are not universal. They are specific to the shape of the vibrating plate—circular surfaces create different patterns than square ones. So in 2003, Stuart began the tedious process of comparing the designs in the chapel to a variety of Chladni diagrams. As he did so, filling in notes in the sequence piece by piece, a melody began to form in his head.
Creating a melody from scratch is a winnowing process. You have to reduce an infinite number of possibilities into a single line of notes that satisfy an arbitrary sense of what good music sounds like. Of course, the definition of "good music" has changed dramatically over time. If Stuart's melody sounded like Gershwin or Bartók, he could be pretty sure he was not on the right track—no one made those kinds of sounds in 1450. If the Rosslyn melody sounded medieval, there was a better chance he was right.
The “staff angel,” whose tablet resembles blank sheet music. Click image to expand.Top: The "staff angel," whose tablet resembles blank sheet music. Bottom: A close-up of the angel's fingers, which point to the locations for B, A, and C on the staff. The younger Mitchell had one clue to guide him in the process. One of the four angels above what's known as the "Apprentice Pillar" was holding a strange instrument that had puzzled scholars for years. Its design was simple: a series of five strings mounted on a board, like a primitive dulcimer. But if this was a dulcimer, the angel didn't know how to play it. While such instruments sit horizontally, with the strings facing the ceiling, the stone angel holds it toward the viewer, like a guitar. This didn't make much sense, because the other 11 angels all play their instruments quite accurately, down to the violinists' fingering and the bells on the piper's wrist. It occurred to Stuart that maybe this dulcimer-like object wasn't an instrument after all. There's another meaning to five horizontal lines. Perhaps it was a musical staff.
If the staff theory is correct, it's another hint that a melody is concealed in the Rosslyn cubes—and maybe even a hint as to how that melody begins. The index finger on the angel's left hand rests on the center line—a B on a treble-clef staff. The index finger of the right hand stops at the gap between that line and the next one up—a C. The middle finger on the right hand is a bit below the middle line—an A. Even more striking is the fact that the three patterns directly above this angel's head, which are repeated in the same order elsewhere among the cubes, resemble the Chladni patterns for those same three pitches, B, C, and A. If this is a coincidence, it's a tantalizing one.
Now that he felt as if he knew the first three notes of the melody, Stuart had a much easier time extrapolating the pitches for the rest of the cubes. When he was finished, he had 213 notes that broke down into 13 short melodies of about 18 notes—one for each of the 13 arches. Stuart postulated that the angel that stands at the base of each arch is an indicator from the mystery composer as to which instrument should play which melody. That left Stuart with the equivalent of a conductor's score, with each instrument's part laid out on the same page.
Is all of this just wishful thinking? To decide for myself whether the Mitchells were on to something or whether they were just hearing what they wanted to hear, I independently coded each carving while in the Rosslyn Chapel. In all but a few cases, I categorized the symbol on the ceiling the same way the Mitchells did. There are perhaps 10 cubes out of 213 that I do not think were correctly coded, though in some cases it's really a judgment call. (Most of the cubes are remarkably preserved, but a few haven't held up as well.) Regardless, these differences aren't significant enough to make more than minor alterations to the melody. Independently matching Chladni patterns to the symbols was trickier work for all the same reasons it was tricky for the Mitchells, but I was still able to make matches for most of the cubes. In short, I don't think Tommy and Stuart have conjured a tune out of thin air.
Stuart Mitchell, too, was satisfied: After months and months of puzzling through the morass of sandstone symbols, he believed he had unlocked the secret music of the Rosslyn Chapel.
To hear piano renditions of five of Stuart Mitchell's melodies, click on the interactive feature below. These melodies (and eight others) form the basis for Mitchell's "Rosslyn Motet." Each melody was derived from a sequence of sandstone cubes that appears above a representation of an angel. The angels are depicted singing (the second melody from the top) or playing an instrument: the lute (top), the recorder (third), the shawm (fourth), and the bagpipes (fifth).
The Rosslyn Code
Is a Scottish Da Vinci responsible for the Rosslyn melodies?By Chris WilsonUpdated Thursday, May 19, 2011, at 7:28 AM ET
Gilbert Hay returned to his native Scotland in 1445 after 20 years on the court of King Charles VII of France—the same king who Joan of Arc helped install to the throne. Now, at age 48, Hay had a new assignment. William St. Clair, a wealthy Scottish lord, had commissioned Hay to translate three volumes on chivalry and warfare from French to Scots, a sort of half-brother to medieval English. Hay set up shop in the St. Clair family castle in the tiny town of Roslin, a few miles south of Edinburgh, and began work on what would become one of the first masterpieces of Scottish literature.
Just as Hay arrived in Roslin, St. Clair was embarking on a project of his own: a massive cathedral, which, by one later account, he wanted to erect as a thank-you note to God for all his good fortunes. (Others think he originally intended to build a university.) The first stone would be laid 10 years later, at a site just a short walk up the hill from the castle. St. Clair would only get one-fourth of his cathedral—the top fork of the intended crucifix-shaped floor plan—before money ran out. The result was Rosslyn Chapel, with all its splendorous figures and curious symbols.
Rosslyn Castle. Click image to expand.Roslin Castle, where Sir Gilbert Hay set up shop in 1445, is now largely in ruins. Tommy and Stuart Mitchell believe the Rosslyn Motet—the piece of music Stuart composed based on the chapel's 213 stone carvings—is proof enough that there's a musical code embedded in the cathedral's walls. Nevertheless, the Mitchells cite Gilbert Hay's presence in Rosslyn as supplementary evidence. At some point during his worldwide travels, they conjecture, Hay learned about Chladni patterns and used them to create a musical code in the chapel that his patron was constructing. Without question, this is the part of the Mitchells' story for which there is scantest evidence.
Hay's education was certainly prodigious. He was among the first graduates of the University of St. Andrews, where he would have studied music as one of the classic disciplines, known as the trivium and the quadrivium. As the chamberlain to King Charles VII, he would have had access to the royal library as well as to the king's eclectic brother-in-law, René d'Anjou. René was a fellow poet and patron of the fine arts (as well as the titular King of Jerusalem). Both men had catholic interests: art, language, literature, governance, and history, to name a few. The St. Clair family moved in circles of people like Hay and d'Anjou. Whether the strange behavior of musical plates ever came up is impossible to say.
-
A few months after I returned home from Scotland, I called the cellist Yo-Yo Ma to get his take on the Rosslyn story. I had interviewed Ma a few years ago for a short profile, and recalled that his Silk Road Project has made him an expert on the thousands of musical instruments that have appeared and disappeared over the ages. I was happy to learn that he also knew a great deal about Chladni patterns—also known as "cymatics"— having seen violinmakers study the unique patterns that form on the board of the instrument as it's played.
Ma believes that Chladni patterns must have been discovered before the 18th century. Millennia ago, he points out, the Chinese had already mastered the art of the "singing bowl," a delicate bell made from a specific bronze alloy that could produce multiple tones at the same time. While there's no documentation that Chladni patterns were used to measure the correct ingredients for the metallurgy, Ma hypothesizes that the instruments' makers knew about the phenomenon. There's some evidence, for example, that Galileo noticed the effect while scraping a brass plate about 100 years after the Rosslyn Chapel was built. Whether this knowledge found its way to Hay, either through his travels or someone else's, will probably never be known. But history dangles the possibility just out of reach.
Even if Gilbert Hay knew of Chladni patterns, an overarching mystery remains: Why would he or anyone else go to so much trouble to conceal music in this obscure format? What would compel someone to carve a secret composition into the Rosslyn Chapel's walls?
Ma says he understands the impulse. "A composer is encoding some of the most intimate things he or she wants to say, and often can't say in a social setting," he says. As a performer, he explains, his "job is to try and do forensic musical analysis … to look at this encoding and say, well, who is this guy? What makes this important to him or her? What gives this meaning?"
When Ma talks about "forensic musical analysis," he means everything from seeking out explicit messages to hunting for what the composer wanted to communicate on purely musical terms. Composers do, on occasion, smuggle jokes and messages into their work. Johann Sebastian Bach liked to slip his initials into various motifs by using the sequence B-flat, A, C, B-natural, which is described in German notation as "B-A-C-H." Each variation of Edward Elgar's "Enigma Variations," likewise, is encoded with the initials or name of a friend. (The first movement contains repeated references to "CAE," his wife, Caroline Alice Elgar.) And, like Bach, Russian composer Dmitri Shostakovich frequently slipped a four-note sequence into his pieces that, when represented in the German notation, read "D, ES, C, H"—that is, "D. Sch."
The Russian composer Dmitri Shostakovich. Click image to expand.The Russian composer Dmitri Shostakovich was fond of sneaking his initials into his work in the form of four notes.The study of musical forensics, though, also requires a subtler unraveling of a composer's intent. Here, Dmitri Shostakovich is again instructive. The Soviet-era composer was perpetually in and out of favor with the Stalinists, and he frequently seeded his music with distaste for his nation's regime. Scholars quarrel over exactly where the composer's political feelings rested, but his work is riddled with sarcastic variations on patriotic marches, clamorous imitations of military exercises, and sorrowful eulogies for the motherland. Putting those sentiments in words would have been a death wish; concealing them in music provided him with protective coloration. (The camouflage wasn't perfect—the government banned his work on several occasions.)
The Mitchells are not arguing that a written message lurks inside the Rosslyn Code. Perhaps, like Shostakovich centuries later, Rosslyn's mystery composer felt the need to hide his musical impulses. Doing a bit of freelance composing in a house of God would not have sat well with the 15th-century Catholic Church. The angels and cubes could have been a way of hiding melodies in plain sight.
I prefer a different explanation. In the 1400s, it was impossible to know whether musical notation would still look the same in 100 years, much less 500. In the pre-light of the Renaissance, new cultural ideas were germinating across Europe and rapidly restructuring the arts. By using Chladni patterns—nature's musical notation rather than man's—a composer could ensure that future civilizations (or even an alien one) would be able to decode his notes.
Or maybe Rosslyn's secret sheet music was just a little inside joke among the chapel's designers. Stonemasons were always doing that—endowing an angel with the face of a favorite aunt, or a skeleton with the face of a mother-in-law. They didn't necessarily care if anyone ever made the connection. It was enough to leave a little part of oneself behind.
-
The 13 melodies that Stuart Mitchell derived from the chapel's 213 stone cubes are attractive and intriguing, but they don't form a freestanding piece of music. For the Rosslyn code to become anything other than a curiosity, Stuart had to arrange the melodies into a fleshed-out piece of chamber music.
Mitchell originally experimented with combining melodies in counterpoint—two voices playing at the same time, complementing one another. His fascination with DNA gave him the idea that the melodies from crisscrossing arches might intertwine like a double helix. (This would be the second-most interesting DNA-related discovery to come out of Roslin: Dolly the sheep was cloned down the street.) The result, though, was cacophony, so he scrapped the idea in favor of a simpler one-melody-at-a-time approach.
Arranging and recombining music is a standard part of composition, and the role of the arranger varies widely. In an extreme case, a composer takes a simple theme and builds a whole new piece around it. One of Bach's most-popular works, his "Musical Offering," begins with a 21-note melody provided by King Frederick II of Prussia and grows into a piece that is unmistakably Bach's. At the opposite end of the spectrum, an arranger might take a complex piano piece—say, one of Franz Liszt's "Hungarian Rhapsodies"—and adapt it for an entire orchestra. Stuart Mitchell's composition, a four-part piece called the Rosslyn Motet, rests somewhere between these poles. The music is closely structured around the Rosslyn melodies, but the harmonizing voices and transitional melodies are his own.
"What he's given us is a basic melody to work with," Stuart said one night over dinner, alluding to Rosslyn's unidentified composer. We were eating at a restaurant called The Shore in upper Edinburgh, on the edge of the North Sea—a place he chose because it has a piano. "Some of the exercises I used to get in university and college were where you'd get about eight bars of Bach or Purcell, and underneath it was all blank. And you would add harmonies to this as Bach would have done or as Purcell would have done." Forming a listenable work out of the Rosslyn Code took a great deal of knowledge of musical history, but it was more technical than creative. You could think of it as Stuart taking an ancient strand of dinosaur DNA and mixing it with modern frog DNA to make a Stegosaurus.
Stuart Mitchell's Rosslyn Motet trades off between the instruments carved in stone in the chapel: lute, fiddle, organetto, bagpipes, and shawm (like an oboe). (The motet also includes several singers—a few of the angels are depicted singing from hymnals.) The melodies are difficult to categorize, but they generally resolve to a version of the A-minor scale. There are melodies that repeat the same note four or five times in a row and others that never resolve. The bagpipe line—one of the most moving melodies in the suite—ends on a D, which sounds tentative and incomplete until the next voice takes up the mantle. A G-sharp makes several guest appearances, giving parts of the piece the Halloween sound of a harmonic minor scale. Taken together, it's an elegantly structured sort of musical prayer, with each melody fitting neatly into an overarching voice that is solemn and arresting.
After dinner, we retired to the keyboard, where Stuart launched into an improvisation on the motet. He began soberly, running through the major themes. From there, his left hand propelled the piece into an extended blues solo over the Rosslyn chords, occasionally drifting back to motif. It was like "Fly Me to the Moon" in the hands of Oscar Peterson. From there he lands on "All the Things You Are," which later turned into the fantastically difficult first movement of Rachmaninoff's second piano concerto, followed by "Nice Work if You Can Get It"—an appropriate characterization of the Rosslyn composition, I think.
In his hands, the music of the cubes—whether an actual relic of the 15th century or an elaborate misreading of some wall decorations—sounds like the most natural thing in the world.
To hear piano renditions of five of Stuart Mitchell's melodies, click "Play Notes" on the interactive feature below. To hear the corresponding melody from Mitchell's finished product, "The Rosslyn Motet," click "Hear Motet." Mitchell took a few small liberties when arranging the melodies into a piece of chamber music. For example, he replaced the first note of the fourth melody with a B to match the opening notes of the first three melodies. He also replaced the last note of the fifth melody with a D, instead of an E, to lead more naturally into the next melody.
The Rosslyn Code
Weighing the evidence—is the chapel's code real or fake?By Chris WilsonUpdated Friday, May 20, 2011, at 6:39 AM ET
The Rosslyn Motet emanates from the crypt as you walk into the chapel. It's playing from a modest little boom box, but the acoustics of the underground chambers are so magnificent that you can hear it clearly from anywhere inside. The already solemn tones sound especially ghostly when the chapel is deserted and you can see your breath, even under the heat lamps.
In the five years since the Mitchells announced to the world that they had solved the Rosslyn code, they've received modest recognition in the media and respectable sales of a professional recording of the work. They shrewdly timed their announcement of the solution to the release of the movie adaptation of The Da Vinci Code, giving reporters and producers a ready-made news peg. Over time, however, their story has faded to a curiosity. Interest in whether it is real has waned.
Tom Hanks and Audrey Tautou in The Da Vinci Code. Click image to expand.Tom Hanks and Audrey Tautou in The Da Vinci Code. The Rosslyn Chapel scenes were filmed on location.
Stuart and Tommy Mitchell say they don't much care if people believe them. On my last day in Roslin, I met Stuart and his friend Ian Robertson at the chapel. Robertson and his co-author Mark Oxbrow wrote by far the best modern book on Rosslyn Chapel, Rosslyn and the Grail. Mostly, Robertson and Oxbrow debunk the various rumors and legends that have attached themselves to the chapel, including the crazy ones about the Holy Grail and Jesus' head. But when it comes to the Rosslyn Motet, they're more circumspect. "One 'code' that may actually exist within Rosslyn," Robertson and Oxbrow write, "is the mysterious carved cubes that ornament some of the arches in the chapel."
Robertson and Stuart Mitchell became friends after Rosslyn and the Grail went to press, but Robertson remains ambivalent about his buddy's theory. When we met at the chapel, the first thing Robertson asked me was whether I thought the code was real. I said I didn't know, but that it at least seemed plausible.
That's the maddening thing about the Rosslyn code. When you're standing under the cubes, watched over by the 13-piece angelic band, it all seems so obvious that you wonder why it took 500 years for someone to unearth the music. But when you go back to assemble the evidence, you realize that the Rosslyn Motet sits on top of a mountain of shaky assumptions. Is the resemblance between the cubes and Chladni patterns a coincidence? Could any random set of notes start sounding human-made when you hear them hundreds of times? Was it really feasible for these melodies to be produced in the 15th century.
Against these doubts, only the music stands as counter-evidence. The Rosslyn code's highest-profile critic is probably Warwick Edwards, a music professor at the University of Glasgow. Edwards, who attended the first public performance of the Rosslyn Motet in the chapel itself, judged that the final product didn't resemble any medieval music he'd ever heard. To that point, however, I don't believe you can judge the Mitchells' creation based on how medieval it sounds—this is more an issue of Stuart Mitchell's orchestration than it is with how the original melodies have been transcribed.
The code has several online doubters, the most prolific of which is the anonymous author of the blog "The BS Historian," who has written skeptically about the Mitchells nearly a dozen times in the past few years. The author's criticisms cover the gamut of the Mitchells' story, from the historical holes to the matching of Rosslyn's symbols to documented Chladni patterns. Others have parroted this skepticism. A commenter on the website of high-profile skeptic James Randi sums up a common reaction quite concisely: "I strongly suspect that this is yet another case of seeing what one wants to see, with heavy confirmation bias."
If the Mitchells' theory is indeed fabricated, it's still fascinating that they coaxed compelling, nonrandom-sounding melodies out of the Rosslyn Chapel's stone cubes. Whether any random series of notes can sound beautiful in the hands of an expert arranger is a tough question to answer, though one footnote to the Rosslyn story tilts me a bit towards the Mitchells' side.
On my first morning at the chapel, before I'd met Stuart, a tour guide named Roger told me he'd heard that many of the cubes had fallen from the ceiling over the years. A stonemason, Roger believed, had replaced them on the ceiling at random as part of a restoration a few decades prior. Uh oh, I thought—I've crossed the Atlantic to investigate a centuries-old song that's younger than I am.
As it turns out, the stonemason, an elderly man name Joe Lang, still lives a few miles from the chapel. Lang later told me the real story: Just a few of the cubes at the chapel's northern end had fallen—"no more than seven or eight" out of 213. But I didn't know this when I talked to Stuart over dinner that first night. "I absolutely hate the end of the music, because it's the only part that doesn't make any sense," he told me. If, in fact, the last few cubes are out of place, this might explain why.
-
After three days in Roslin, I still wasn't totally convinced by the Mitchells' story. In search of more evidence, I paid a visit to the Scottish National Library in Edinburgh. I wanted to see the manuscript on chivalry and warfare that William St. Clair, the chapel's founder, commissioned from Sir Gilbert Hay—the man Tommy and Stuart Mitchell believe might be responsible for Rosslyn Chapel's carvings. The Hay manuscript is well guarded. To see one of Scotland's most-treasured artifacts, you have to get two senior curators to sign off on the request; at that point, the manuscript must be delivered from a separate location.
Hay's 500-year-old words are penned in the elegant handwriting of a professional scribe, and are more or less impossible for a lay reader to decipher. In many places, the book has been idly vandalized. The Curroy family, later owners of the manuscript, took the liberty of doodling in the margins, practicing their signatures and drawing little hands with pointing fingers.
The very first page in the tome is a half-size sheet with some incomprehensible script on the front. When I turned it over, however, I saw something flabbergasting: a nine-note melody sketched out on a familiar five-line staff.
Was this incredible proof that Stuart and Tommy Mitchell had been right all along? I compared these nine notes to the score of the Rosslyn Motet, however, and I didn't turn up a match. I sent the image to Stuart, and he pointed out that there was a link—the notes in the manuscript pass through the same scale as most of the melodies he decoded. It's not a strong connection, but it's something.
Alas, Oxford University's Sally Mapstone—who wrote her dissertation on Hay's work—told me this was all wishful thinking. The manuscript in the Scottish National Library did not belong to Sir Gilbert Hay; it was copied about 40 years later for a St. Clair descendant. Mapstone identified the script above the music as 18th-century. Unless another detective found himself on the same trail, it's probably just a tantalizing coincidence.
-
More than 500 years after the Rosslyn Chapel was built, codes continue to fascinate and confound us. Back in November, the New York Times ran a profile of Jim Sanborn, the artist responsible for one of the most famous unsolved codes of modern times. It is called "Kryptos," and it lives on a copper sculpture at CIA headquarters in Langley, Va. The serpentine installation contains a long series of letters broken into four parts, with each section encrypted using a different method. The first three sections have been solved, but the fourth has proved to be so confounding that Sanborn felt compelled to give the Times a six-letter clue in the hope that it would hurry along the process. He's sick of dealing with it.
Kryptos code. Click image to expand.Kryptos consists of four different substitution codes, in which each letter represents one letter in the solution. One of the four remains unsolved.
When I got back from Scotland, I showed Sanborn some images of the cubes. He seemed to think they might be the work of a cryptographer. The trouble with the chapel's cipher, he said, was that the act of solving it doesn't prove it was a cipher in the first place. It's virtually impossible to crack the fourth part of Kryptos and get the wrong answer—the odds that the same code could produce two different coherent passages of text is functionally zero. (This hasn't prevented people from confronting Sanborn with outlandish solutions, sometimes on his lawn.) While Sanborn told me that Kryptos has meaning beyond the passages, he verified that—unlike the alleged Rosslyn Code—you don't need any information outside the letters themselves to crack it.
Since there are only 12 unique symbols, it's unlikely someone else will find a written message in the Rosslyn cubes. Anything is possible, but I think the book on the stone cubes is closed. Either the Rosslyn Motet is a 500-year-old musical message, or the cubes mean nothing.
But perhaps that's not the point. The chapel's designer, I'm sure, would be pleased to know that we're still engaged by his creation—pondering what it means and what he wanted us to see. By happy coincidence, I had just finished reading an exquisite little novel by the British author J. L. Carr, A Month in the Country. The narrator, a World War I veteran, is a restorer of medieval paintings that had been whitewashed over by a disapproving churchman. This line of work takes him to a tiny settlement in northern England, where he slowly chips away at a canvas. As he persists, the original artwork—a masterful depiction of the Judgment—comes into focus:
Here I was, face to face with a nameless painter reaching from the dark to show me what he could do, saying to me as clear as words, "If any part of me survives from time's corruption, let it be this. For this was the sort of man I was."