
Leonardo’s Sixteen Years With the Mona Lisa
Leonardo da Vinci was supposed to paint the Mona Lisa and hand it over. Instead, he kept it for sixteen years. Why he carried one portrait across countries, revised it for years—and never gave it to the family who commissioned it.
In 1503, Leonardo da Vinci began a portrait of a Florentine silk merchant’s wife. Sixteen years later, in another country, the painting was still with him. The usual explanation is that a perfectionist genius could not bring himself to finish. The surviving evidence tells a stranger story.

I. A NOTE IN THE MARGIN
The first surviving sentence anyone wrote about the Mona Lisa was written in the wrong place. Not in a contract, not in a payment ledger, not in a letter from an impatient husband asking when the portrait of his wife might finally be ready. It survives in the margin of a book.
Sometime in October 1503, a Florentine civil servant named Agostino Vespucci was reading a twenty-six-year-old edition of Cicero. Vespucci worked in the chancery of the Florentine Republic, in the world of secretaries, diplomats and government clerks; he was a colleague of Niccolò Machiavelli. And while reading, he reached a passage about the ancient Greek painter Apelles.
Apelles had been dead for nearly eighteen centuries, which for Renaissance humanists made him more useful rather than less. He represented an almost unreachable standard: the painter against whom a living master might be measured if you wanted to flatter him extravagantly. The passage in front of Vespucci described an unfinished Venus. Apelles had brought the head and upper body to extraordinary refinement while the rest remained incomplete.
Vespucci thought of somebody he knew. He took up his pen and wrote a name in the margin: “Apelles the painter.”
Then he wrote that Leonardo da Vinci behaved the same way in his paintings, and he mentioned two works Leonardo had underway: an image of Saint Anne, and the head of Lisa del Giocondo. And then, almost casually, he added something else. They would see, he wrote, what Leonardo managed to do with the great council hall, since he had just reached an agreement with the gonfaloniere.
It is an astonishing little document. In a few cramped lines, written for no audience we know of, a Florentine bureaucrat gives us the first secure contemporary evidence that Leonardo was painting Lisa Gherardini del Giocondo in October 1503. Heidelberg University Library, where the annotated Cicero is now preserved, describes the discovery as the evidence that fixed the portrait to Lisa and to this moment.
But the note tells us something more interesting than the sitter’s name. The Mona Lisa enters recorded history already entangled with incompletion. That is not something a later biographer imposed on the story. It is there at the beginning.
The tone matters too. Vespucci does not discover that Leonardo leaves paintings incomplete and diagnose a defect of character. His first instinct is to reach backward through antiquity and set Leonardo beside Apelles. In the humanist imagination, incompletion could carry its own prestige: the trace of a master whose thought had exceeded the boundaries of a finished object.
The modern question begins somewhere else. Why, we ask, did Leonardo take sixteen years to finish one portrait? Vespucci would have found something strange in that. He had watched the way of working while it was happening, and he did not begin from the assumption that every picture marches toward a clean terminal state called finished, or that failing to arrive there requires a psychological explanation. He was describing a process, not a delay.
There is another reason to linger over his little note. In the same handful of words in which the Mona Lisa first appears, so does the other Leonardo: the Leonardo of grand commitments, republican money, public expectations and spectacular promises. Lisa’s portrait is there. Saint Anne is there. And then Vespucci glances toward the enormous wall Leonardo has promised to paint for the government of Florence and writes, in effect: we shall see.
The entire story is already inside those words.
Leonardo is fifty-one years old. He has just started a small portrait on a wooden panel. He will still have it when he dies sixteen years later, hundreds of kilometres from Florence, in the service of a French king.
To understand why, we have to follow the painting. The difficulty is that, almost immediately, it disappears.
# PART ONE
FLORENCE, 1503–1506
The years he should have been doing something else

II. THE MAN WHO CAME BACK
Florence in 1503 did not need to be introduced to Leonardo da Vinci. It had been waiting for him.
Leonardo had trained there as a young man, in the workshop of Andrea del Verrocchio. Then, in the early 1480s, he left for Milan and stayed away for the better part of seventeen years. By the time he returned, the city had changed and so had he. He was no longer an unusually gifted Florentine painter. He was Leonardo. He had designed machines, court entertainments, weapons and buildings. He had painted The Last Supper. He had acquired a reputation that travelled farther than his paintings did.
That last distinction mattered, because Leonardo returned with a second reputation as well. There were things he had promised to make, and they had a habit of not becoming things.
In 1481, the monks of San Donato a Scopeto commissioned an Adoration of the Magi, to be finished within thirty months. Leonardo began it magnificently, figures emerging out of the panel in an extraordinary storm of horses, ruins, bodies and gestures. Then he left Florence, and the painting stayed behind, unfinished.
In Milan, Ludovico Sforza gave him an assignment so enormous that merely attempting it was an advertisement for the regime: an equestrian monument to Ludovico’s father, Francesco Sforza. Leonardo planned a bronze horse on a scale Italy had never seen. He studied equine anatomy, designed methods for casting it, and produced a colossal clay model. Then war reordered the priorities of Milan, the political world that had commissioned the horse collapsed, and the horse was never cast.
This pattern has often been folded into the same convenient explanation. Leonardo was a genius. Leonardo was a perfectionist. Leonardo became distracted. Leonardo could not finish things. All of these contain enough truth to survive casual inspection, which is precisely why they are dangerous. They turn very different events, among them war, technical experimentation, patronage collapse, contractual conflict and changing interests, into a single temperament.
Once that explanation is installed, the Mona Lisa seems simple. Of course he kept painting it for years. Leonardo was Leonardo.
But it conceals the more revealing pattern, which is not that Leonardo left things unfinished. It is what he did with the unfinished things. The great public projects stayed where they were. The Adoration stayed in Florence. The equestrian monument never escaped Milan. Within a few years another immense work intended for a Florentine wall would remain unrealised. Leonardo moved on.
The small paintings behave differently. We are not there yet.
First, Florence gives him the largest wall it possesses.

III. THE WALL
There is a wonderful cruelty in the commission.
The government of Florence had built an enormous assembly chamber inside the Palazzo Vecchio, the Hall of the Great Council, and it needed images commensurate with the ambitions of the republic. Leonardo would depict the Florentine victory at Anghiari. Michelangelo Buonarroti, in his twenties and fresh from the David, received the opposing battle commission. Leonardo was more than twenty years older.
For generations afterward, people would speak about the confrontation almost mythologically, two artists who would become perhaps the most famous names in Western art, assigned opposing walls inside a single room. The important thing is that neither finished.
Leonardo’s problem began with the thing that made him Leonardo. He did not simply want to put an image on a wall. He wanted to solve a problem.
Traditional fresco requires speed. Pigment is applied into wet plaster and becomes chemically bound as the plaster sets. The system is durable, but it imposes a timetable, and the wall determines the pace. Leonardo wanted oil-like depth, revision, transparency and shadow—the freedom he had on a panel—so his plans and experiments pushed beyond standard fresco practice.
What actually happened to the Battle of Anghiari is less secure than the famous story makes it sound. Later accounts describe an experimental mixture and an attempt to accelerate drying with heat; some versions have braziers lifted toward the wall, after which sections failed dramatically. But recent documentary and architectural work goes further, arguing that Leonardo never painted the hall wall at all and that the project stopped at preparatory cartoons. The secure point is narrower than the legend: Florence commissioned a monumental painting, Leonardo developed it extensively, and no finished mural resulted.
That uncertainty is worth keeping. Leonardo has suffered enough from anecdotes that became facts because they were irresistible. We do not need paint literally pouring to the floor to understand the scale of the failure. The government had given him a monumental commission. Years of preparation did not produce the public image it had asked for.
There is something almost unbearably Leonardo about this. The safe method was known. He wanted another method because the known one could not produce what he wanted. The same appetite that made him extraordinary could make him professionally impossible.
The remarkable thing is that Vespucci had somehow seen the problem before the wall was painted. Remember his marginal note: the portrait of Lisa, then the government commission, then we shall see what he does. Two projects starting at almost the same moment. One immense, public, politically charged, watched by Florence. The other a poplar panel less than a metre high.
The enormous thing does not survive. The little thing does.
And while Leonardo is failing on the wall, something quieter is happening on the panel.

IV. THE WOMAN
Her name was Lisa Gherardini. For centuries, because of the fame of the thing that happened to her image, it became easy to treat her less as a person than as a container for mystery. Who was she really? Was she secretly someone else? Was the smile hiding something?
The surviving documents make the basic identity far less mysterious than popular culture prefers. Lisa was born in Florence in 1479. In 1495, still a teenager, she married Francesco del Giocondo, a silk merchant considerably older than she was. They had children. They bought things and sold things. They dealt with property and religious institutions and the practical machinery of an ordinary prosperous Florentine household.
This ordinariness matters. The woman in the Louvre did not begin as the most famous face on earth. She began as the wife of a merchant.
Francesco del Giocondo had links with Leonardo’s father, Ser Piero da Vinci, a Florentine notary. The Louvre’s current catalogue notes contacts between the men dating back at least to 1497 and suggests that these social connections may help explain how Francesco encountered Leonardo. That possibility changes the temperature of the commission.
Strip away everything that happened afterward. No bulletproof glass, no Louvre, no theft, no reproductions, no novels, no advertisements, no tourists raising phones above one another’s heads. A Florentine merchant wants a portrait of his wife. His social world overlaps with the father of a famous painter who has recently returned to town. Perhaps an introduction is made. Perhaps an arrangement follows.
We do not possess the contract. That absence will become important.
For now, the panel begins. And before Leonardo paints Lisa’s skin, he alters the chemistry beneath it. In 2023, researchers published an analysis of a tiny sample from the preparatory layer of the Mona Lisa. They found plumbonacrite, a rare lead compound whose presence points toward an alkaline preparation involving lead oxide. The authors argue that Leonardo probably treated his oil with a substantial amount of lead oxide, perhaps to obtain a thicker material suitable for preparing the wooden support.
This is Leonardo before the portrait has properly begun: already experimenting, already altering the recipe, already asking whether the material everyone else uses can be made to behave differently.
Under later imaging, researchers also found evidence that the design was transferred to the panel from preparatory work rather than invented freely on the wood. There was planning before paint.
So the Mona Lisa is not the story of a man who sat down one morning, looked at a woman, and spent sixteen years unable to stop fiddling with her face. There was a design. There was procedure. There was preparation. There was a painting that, at least in the beginning, belonged to the recognizable world of Renaissance portraiture.
And then something happened to its relationship with time. We do not know exactly when.
This is where the story becomes difficult, because Leonardo leaves Florence, and the painting vanishes from the written record with him.
# PART TWO
MILAN, 1506–1513
The years he learns what he is painting

V. THE DARK BETWEEN THE DATES
Here is almost everything we can prove about the sixteen years, in two sentences. In October 1503, Agostino Vespucci says Leonardo is working on the head of Lisa del Giocondo. On 10 October 1517, a visitor to Leonardo’s residence in France sees a portrait of a Florentine woman among his paintings.
Leonardo dies in 1519. That is the skeleton, and everything else has to be built around it.
There is no surviving diary that says Tuesday. Added another glaze to Lisa’s left cheek. No courier records the portrait arriving in Milan. No assistant writes that Leonardo unpacked it in Rome. No customs official describes the panel crossing into France. Nobody gives us sixteen years of continuous observation.
The historical record does something much more frustrating. It blinks. 1503. Darkness. 1517. Death.
Even the assertion that Leonardo worked on the painting continuously for sixteen years is stronger than the documents permit. The Louvre dates the object across 1503–1519 and says Leonardo retained it in order to continue its pictorial execution, but that is a scholarly reconstruction, not a day-by-day record. Two fixed points, and inference in between: better to say so plainly than to pretend somebody watched the panel travel.
This is not a weakness in the story. It may be the story’s most important fact.
When people say “Leonardo spent sixteen years painting the Mona Lisa,” the sentence creates an image in the mind. Sixteen years at an easel, sixteen years of adjustments, sixteen years of obsession. But the calendar does not measure brushstrokes. It measures the extraordinary duration of Leonardo’s relationship with an object. Those are different things.
After 1506, Leonardo returned to Milan. Florence expected the departure to be temporary. It was not. The city wanted its famous painter back. There was unfinished business in the Great Council Hall. There had been money, there had been promises, there was correspondence. Leonardo extended his absence, then extended it again.
And the portrait appears to have gone with him.
That last sentence has to be written carefully. We do not have the packing list. We infer the journey because the painting that existed in Florence in 1503 is with Leonardo in France fourteen years later, after Leonardo himself has lived in Milan and Rome. Most tellings smooth over the join. We should not. The gap is part of what makes the picture strange. We follow Leonardo because we can; we follow the panel because, much of the time, we must.
And in Milan, during the years when the portrait is presumed to be somewhere in Leonardo’s working world, Leonardo starts opening bodies.

VI. INSIDE THE FACE
One winter, Leonardo opened an old man.
The man had died peacefully. Leonardo had spoken with him before his death and was struck by the absence of suffering, and afterward he dissected the body, searching for the physical cause of what he described as a gentle death.
This is where the modern separation between “Leonardo the artist” and “Leonardo the scientist” becomes actively misleading. Leonardo did not have two careers that occasionally borrowed from one another. He had questions, and painting, anatomy, optics, mechanics and geometry were the methods of attacking them. What is a body? How does it move? What causes an expression? Where does sight happen? Why does distance turn mountains blue? What is the boundary of a shadow? What does the eye actually see?
Between roughly 1506 and 1513, Leonardo returned intensely to human anatomy. In the winter of 1510–11, probably working around the medical school at Pavia with the anatomist Marcantonio della Torre, he investigated bones and musculature in extraordinary detail. Royal Collection Trust describes this period as the height of his achievement as an artist-anatomist and notes that the surviving sheets analyse the body mechanically rather than merely recording its appearance.
Among those sheets are studies of facial muscles. Not an artist’s shorthand for a face, but the machinery beneath one. The superficial muscles are difficult to dissect because they interact intimately with soft tissue and skin, and yet Leonardo separated and drew the structures responsible for movements around the cheek and upper lip with remarkable accuracy.
Consider what that means in the vicinity of the Mona Lisa. For centuries, writers would turn her mouth into poetry: the unknowable smile, the mysterious smile, the supernatural smile, the smile that knows your secrets. Leonardo, meanwhile, was cutting into dead faces to find out what moves a living mouth.
That does not prove he repainted Lisa’s smile immediately after a particular dissection. We do not possess that sequence. It proves something subtler and more valuable: during the span in which the portrait remained with him, the painter’s knowledge of what a face was changed radically. A commissioned portrait could have left the workshop before that knowledge existed. This one did not.
Anatomy was only half the problem. The other half was sight.
Leonardo had spent years thinking about vision, about shadow and light and the eye and atmosphere and distance and the disappearance of boundaries, and he had become increasingly hostile to the line. Nature, as the eye encounters it, does not outline objects in ink. There is no black contour around a cheek, no drawn border separating a mouth from the skin around it. A line is often a diagram imposed by the mind.
If your goal is merely to identify the objects in a picture, that hardly matters. If your goal is to make a painted face behave like something alive, it matters enormously. You have to make form exist without drawing a fence around it.
The solution was not a single trick. It was accumulation: shadow laid over shadow, transparent or semitransparent paint over earlier paint, edges degraded until the viewer can no longer tell exactly where one region ends and another begins. Smoke. Sfumato.
Five hundred years later, the neuroscientist Margaret Livingstone offered another way of understanding why the mouth behaves so strangely for a viewer. The smile, her research argues, is carried heavily in low spatial frequencies, the broad, coarse visual information to which peripheral vision is particularly sensitive. Look elsewhere in the face and the smile strengthens. Look directly at the mouth and the effect weakens.
This does not prove that Leonardo understood spatial-frequency processing in the modern neurological sense. That claim would turn fascinating evidence into nonsense. What it tells us is that the optical effect has a physical basis. The smile does not merely seem unstable because generations have told us it is mysterious. Its appearance changes with the way we look at it. And the man painting it was, during these years, obsessively interested in precisely that larger category of problem: what the eye receives, what it loses, and what the mind completes.
The Mona Lisa looks different placed beside these investigations. Less like a portrait waiting patiently for its owner. More like a surface on which new knowledge could still be used.

VII. THE OTHER MONA LISA
For centuries there seemed to be only one witness inside the room. Then restorers removed black paint from another picture.
The Prado Museum in Madrid owns a Mona Lisa. For a long time it was treated as one among many copies generated by the extraordinary afterlife of Leonardo’s composition. Then the museum studied it. Underneath the dark overpaint, conservators found the landscape. Infrared examination exposed something more important: changes.
The Prado painter had not simply copied the completed surface of Leonardo’s picture. He had followed alterations made during its creation. Lines shifted. The waist changed, and the fingers, and the veil, and the head; cheeks and neck were adjusted. Elements Leonardo drew underneath and later omitted from the visible painting also appeared in the copy.
The Prado put the logical problem beautifully: an ordinary copyist copies what can be seen, and he cannot copy what has been hidden beneath later paint. The shared underlying corrections therefore imply that whoever painted the Prado version had access to Leonardo’s composition while it was developing. The museum concluded that the two works were produced at the same time and in parallel, by someone in Leonardo’s studio.
It is about as close as Renaissance painting comes to giving us a saved earlier version of a file. The Louvre panel shows the state Leonardo ultimately left visible. The Prado panel keeps traces of decisions he subsequently erased.
For one miraculous moment, the painting stops looking timeless. It moves. A hand was somewhere else. A contour was different. A chair existed and then disappeared. A veil changed shape. Someone standing beside Leonardo was watching those changes happen and reproducing them.
That matters because one of the easiest mistakes to make about the Mona Lisa is to imagine its sixteen years as sixteen years of polishing a fixed image. The material evidence says otherwise. The composition evolved.
The caveat is that the Prado evidence does not hand us a calendar for every revision. It cannot tell us that a particular finger moved in 1509, or that the veil changed in 1512. Again the record refuses to become a screenplay. But it proves that the picture possessed states. It was not simply finished slowly. It changed.

VIII. THIRTY MICRONS
The face is built out of distances too small for the eye to resolve.
In 2010, researchers used X-ray fluorescence to examine the painted layers in Leonardo faces at the Louvre, including the Mona Lisa. In some of the glazes, individual layers measured only one or two micrometres. A micrometre is one-thousandth of a millimetre. The total glaze buildup in the analysed areas could be no more than thirty to forty micrometres thick. A human hair can be thicker than the entire system.
This is one reason reproduction betrays Leonardo. A photograph gives us the image. It does not give us the object’s method of becoming visible. The transitions around the eyes and mouth are not colours selected from a palette and placed side by side. They are optical constructions. One translucent film modifies what lies underneath, then another modifies both. The boundary is not drawn. It is accumulated.
Time enters the technique. Oil layers require intervals between operations, and different materials, thicknesses and conditions produce different drying behaviour. Leonardo could return, change, deepen, soften, veil.
Here, though, we should resist another seductive exaggeration. Microscopic glazes do not prove that the Mona Lisa had to take sixteen years. The chemistry does not contain a sixteen-year timer, and thousands of Renaissance paintings involve layered oil technique without requiring a decade and a half. What the glaze measurements prove is more useful: Leonardo’s method was inherently incremental. The object was built through sequences that invited return. And because Leonardo did not release the panel, return remained possible.
That distinction changes the question. Perhaps sixteen years is not the amount of time required to make the picture we see. Perhaps it is the amount of time during which Leonardo continued to allow the picture to change.
A portrait handed to Francesco del Giocondo in 1504 would have frozen Leonardo’s knowledge at 1504. A portrait still in Leonardo’s possession in 1511 could absorb what he had learned by 1511. A portrait still available in 1516 could become something else again.
What mattered most about the Mona Lisa may therefore have been neither incompletion nor perfectionism. It may have been availability. Leonardo could still use it.
And then Milan collapsed around him again. He moved south, and the painting apparently moved with him. In Rome, the question of who the portrait was for becomes stranger.
# PART THREE
ROME, 1513–1516
The years he is overtaken

IX. THE OLD MAN IN THE BELVEDERE
Leonardo arrived in Rome in 1513, in his early sixties. The city should have been the final confirmation of everything he had become. Instead, it showed him the future.
Raphael was there. Michelangelo was there. The great rooms had painters. Michelangelo had recently finished the ceiling of the Sistine Chapel, a project so enormous in ambition and physical execution that it already seemed to belong to another order of achievement. Raphael, decades younger than Leonardo, was working at the centre of papal patronage.
Leonardo had status. He had accommodation in the Belvedere under the protection of Giuliano de’ Medici, brother of Pope Leo X. He had ideas. He investigated geometry and mirrors, went on designing machines and hydraulic projects, considered the Pontine Marshes, made plans. But Rome did not give him its defining walls.
The pattern is almost painful. Leonardo had once been the man Florence arranged opposite Michelangelo. Now he could walk through a papal complex in which the greatest monumental work was being done by men young enough to be his sons.
There is a famous Vasari story about this period. Pope Leo X hears that Leonardo is preparing varnishes before beginning a painting and supposedly remarks that this is a man who will never accomplish anything, because he thinks about the end before he has begun.
It is a perfect Leonardo anecdote, and therefore a dangerous one. Vasari adored perfect anecdotes. He was born in 1511 and was a child while these events occurred. His Lives are indispensable evidence for how Leonardo’s reputation developed, but they are not a stenographic record of rooms he never stood in.
Still, even as legend, the line tells us something. By the middle of the sixteenth century, Leonardo had acquired a recognizable character in cultural memory: the man for whom preparation could become the project, the man who thought beyond the object, the man who might understand everything except how to conclude. That legend would eventually swallow the Mona Lisa.
But Rome gives us a more concrete problem: Giuliano de’ Medici. Because when somebody finally sees Leonardo’s portrait in France, he says Giuliano was responsible for it.

X. WHOSE PAINTING IS IT?
The traditional story is wonderfully simple. Francesco del Giocondo commissioned Leonardo to paint his wife Lisa. Leonardo worked on the portrait. Leonardo never delivered it. The end.
Then a man walked into Leonardo’s house in 1517 and left us a sentence that has been irritating art historians ever since.
Antonio de Beatis was secretary to Cardinal Luigi d’Aragona. During a European journey, the cardinal’s party visited Leonardo at Cloux, near Amboise. Unlike Vasari, de Beatis was actually in the room. He saw the paintings. And he described one of them as a portrait from life of a certain Florentine lady, made at the request of the late Giuliano de’ Medici.
Giuliano had been Leonardo’s patron in Rome. He died in 1516. So what are we supposed to do with Francesco del Giocondo?
There is no universally accepted answer that makes the problem disappear. Perhaps de Beatis misunderstood what Leonardo told him. Perhaps Leonardo or someone in his household represented the provenance differently for prestigious visitors. Perhaps the original commission changed purpose over the years, or Giuliano became involved with a painting begun long before his Roman patronage. More radically, some scholars have proposed that two different portraits sit behind the conflicting documents.
There are additional sitter theories tied to Giuliano, including Roberto Zapperi’s proposal involving Pacifica Brandani, the mother of Giuliano’s illegitimate son. But de Beatis explicitly calls the woman Florentine, and Pacifica was from Urbino; the theory solves one problem by creating another. It is a live dispute, not a secret identity waiting to be revealed.
There is no need to turn any of this into a whodunnit. The 1503 Vespucci note is powerful evidence that Leonardo was indeed working on Lisa del Giocondo. The important point is a different one: during the long years in which Leonardo retained the painting, its relationship to its original commission became obscure even to people close enough to ask him about it.
That is extraordinary. A portrait is supposed to be one of the simplest transactions in art. A person, a painter, a patron, an object, a price, delivery. Here, almost every link loosens, and one of them disappears almost completely.
The patron.
Because if Francesco del Giocondo commissioned the portrait in 1503, there is a question we ought to ask far more often. Where is he?
Leonardo was not operating in a world without contracts. The Adoration came with a thirty-month deadline, and the Great Council Hall commission generated surviving records of payment and official expectations. This was not a culture in which patrons necessarily shrugged when painters vanished.
Francesco del Giocondo was not powerless. He was a prosperous merchant. Yet the archival reconstruction assembled by Kemp and Pallanti has produced no lawsuit over the portrait, no furious letter, no formal demand, no surviving complaint.
Absence of evidence is dangerous. Documents disappear and archives are incomplete, and it would be absurd to infer too much certainty from silence. But silence can still be a historical fact worth noticing.
And here, a detail from the beginning returns. Francesco knew Leonardo’s father. In 1497 Ser Piero da Vinci arbitrated a dispute involving the Servite friars and Francesco and Antonio del Giocondo; Leonardo later lodged at Santissima Annunziata, an institution connected to both families. Research by Martin Kemp and Giuseppe Pallanti, reflected in the Louvre catalogue, places the commission inside a surprisingly ordinary web of Florentine social relationships.
Now imagine one possibility. Not a formal transaction between strangers, but a connection, a recommendation, a favour mediated through family networks. A famous son is back in town. A respectable merchant would like his wife painted. The painter’s father knows the household. The arrangement begins. Then Ser Piero dies in 1504, Leonardo drifts away, and the portrait never develops the normal enforcement structure of an arm’s-length contract.
Is that what happened? We do not know. No document says so. It is an inference built from proximity and absence, and there are other possible explanations. Perhaps money changed hands in a way we have lost. Perhaps another object satisfied the family. Perhaps the surviving Louvre picture is not identical in every respect to the commission Francesco expected. Perhaps the documentary hole is simply a hole.
But the speculative possibility is useful, because it exposes how much the conventional story has been pretending to know. “Leonardo refused to deliver the Mona Lisa” sounds like a documented event. It is not. We do not possess the moment of refusal, or the demand, or the quarrel. We possess an object that stayed with the artist. That is all.
And staying with the artist may have become its function.
# PART FOUR
FRANCE, 1516–1519
The years he stops

XI. OVER THE ALPS
Giuliano de’ Medici died in March 1516, and Leonardo’s Roman protection weakened. Then France opened. The young king François I invited him north. Leonardo was sixty-four. He left Italy.
The journey has accumulated picturesque details in biographies, because picturesque details are difficult to resist. A mule. A mountain pass. An elderly master wrapped against the cold, cases of manuscripts strapped beside him. The Mona Lisa somehow travelling through the snow.
The record does not cooperate. Accounts disagree about how Leonardo crossed and about the route, and they differ even over the precise timing of his arrival. The mule is possible. It is not secure enough to saddle as fact. The compelling thing was never the mode of transport: a sixty-four-year-old man crossed the Alps carrying his intellectual household and his late paintings, and the details of the crossing remain obscure.
That obscurity returns us to the panel, which is doing something remarkable now. It is leaving the country in which it was supposedly commissioned. That should feel stranger than it usually does. A portrait made for a Florentine merchant is no longer in Florence. It is not even in Italy.
Leonardo is carrying his world with him. Books, papers, designs, assistants, ideas. And paintings.
At Cloux, today Clos Lucé, near the royal château of Amboise, François I gave Leonardo rank, security and proximity to the court. Leonardo became the king’s First Painter, Engineer and Architect. At an age when many Renaissance lives were already over, he was still designing: ideas for a royal complex at Romorantin, contributions to court spectacles, thinking about architecture and water. The French king seems to have valued Leonardo less as a supplier of pictures than as Leonardo in total, an artist, engineer, philosopher, conversationalist, designer, living repository of knowledge.
And perhaps that is exactly the environment in which the Mona Lisa finally becomes legible. Nobody in France is waiting for Lisa del Giocondo’s portrait. Nobody there needs Leonardo to complete a Florentine domestic transaction. The panel has been detached from the ordinary life it may once have had. It belongs to Leonardo’s working world now.
And it is not alone.

XII. THE ROOM
On 10 October 1517, Antonio de Beatis entered Leonardo’s residence. Consider how much of the story we owe to the fact that this particular secretary wrote things down. Without Vespucci, the beginning blurs. Without de Beatis, the end does.
Leonardo showed the visitors manuscripts. He showed them inventions. He showed them paintings, and three are recorded: a Florentine woman, a young Saint John the Baptist, and the Virgin and Child with Saint Anne.
There they are. Not one mysteriously retained masterpiece. Three late paintings in the possession of the same old man.
This is the finding that changes the story. The Louvre dates the Mona Lisa across 1503–1519 and describes Leonardo as retaining it until his death while continuing its execution. It similarly dates the Saint Anne into 1519 and says he kept working on it until death. The museum states that Saint John the Baptist was also retained for continued work and remained unfinished in 1519. The three pictures de Beatis encountered are not random decoration. They are a cluster of late panels Leonardo kept with him rather than delivered away.
The word unfinished needs care here. It does not mean that de Beatis walked into a storeroom and found three visibly abandoned sketches. These were pictures Leonardo chose to show a distinguished visitor, and the visitor described them as notable works. The Louvre's judgement concerns their terminal state: evidence that Leonardo retained them for continued execution and died before declaring, or being forced into, an end. A painting can be coherent enough to display, accomplished enough to astonish, and still open in its maker's hands. That distinction is central to the room. If unfinished meant merely rough or defective, the three panels would be three failures. If it means highly resolved but not closed against further change, they become evidence for the kind of object the Mona Lisa had become.
Museums have to assign objects a state. Workshops do not always oblige them by producing a documented last stroke. In this case there is no signed declaration of completion to set against the catalogue label. There is possession, evidence of prolonged execution, a visitor's encounter and then death. “Unfinished” names the open end reconstructed from those facts; it is not a verdict on whether the picture had already achieved the force that made somebody want to see it.
This is where the theory of a singular obsession begins to collapse. If only the Mona Lisa had behaved this way, we might be forced to explain its retention through something uniquely emotional about this painting. He loved it. He loved her. He could not perfect the smile. He saw himself in it. Pick your psychology.
But the room contains three paintings. One of them, the Saint Anne, also reaches back toward the beginning of the century. Another, the Saint John, belongs to Leonardo’s later development and pushes sfumato into something almost hallucinatory, the body rising out of darkness. All three stayed close. All three crossed into France.
The anomaly is no longer one portrait. The unit of analysis has changed. We are looking at a way of working.
De Beatis noticed something else. Leonardo’s right hand was impaired. The secretary writes that paralysis had affected it and that one could no longer expect the kind of fine painting Leonardo had once produced, although he could still draw and instruct others. The precise medical diagnosis remains disputed.
There is a complication. Leonardo was famously left-handed, and his surviving notes run in mirror script from right to left, a habit the medical-historical review connects naturally to his left-handedness. So de Beatis’s conclusion is not medically transparent: a palsied right hand does not automatically explain the end of a left-handed artist’s painting career. Age may. Illness may. General physical decline may. The detailed motor demands of painting may have been affected in ways a visitor could see without describing accurately. Once again, the witness gives us evidence, not omniscience.
What matters is the narrowing. By 1517 Leonardo is an old man in Renaissance terms. He can still draw, still teach, still think. But the years in which a picture can be endlessly returned to are running out.
He makes his will on 23 April 1519. Nine days later, on 2 May, he dies.
Leonardo left Francesco Melzi his books and the materials of his intellectual work. The paintings followed a less transparent route. A French payment record from 1518 says Salaì supplied the king with unnamed painted panels, and the Louvre therefore describes François I’s acquisition of the Mona Lisa that year as highly probable rather than presenting every step as settled.
That is fitting. The picture arrived in history through a marginal note. It crossed sixteen years of documentary darkness. And even at the end, the paperwork refuses to close neatly around it.
But Leonardo’s death gives us one kind of ending no contract ever did. He can no longer return to the panel. The sixteen years stop.

XIII. WHY SIXTEEN YEARS?
Now we can finally ask the question. Why did Leonardo spend sixteen years on the Mona Lisa?
The first answer is that the sentence itself is deceptive. We do not know that he painted it continuously for sixteen years. We know he began a portrait of Lisa del Giocondo around 1503. We know the Louvre panel underwent changes. We know he developed painting techniques capable of extraordinarily fine layered effects. We know his investigations of anatomy and vision intensified during the years in between. We know the painting ultimately remained with him rather than with the Giocondo household, and that it was present in his orbit at the end of his life.
The sixteen years measure a relationship, not a continuous session.
So forget, for a moment, the image of Leonardo standing in front of the same easel every morning for a decade and a half, adjusting one corner of a smile. Look instead at the things he left. The Adoration of the Magi, left in Florence. The Sforza monument, left unrealized in Milan. The Battle of Anghiari, intended for a Florentine wall and never completed. Again and again, Leonardo’s largest obligations were attached to places, institutions, patrons and political systems, and when the relationship broke, the project remained behind.
Now look at the room in France. The Mona Lisa. The Saint Anne. The Saint John. Small enough to move and still within reach for further work.
The contrast is not a law of Leonardo's career. His work is too complicated for a neat binary between public failures and private experiments. Some commissions were delivered. Some disputes ended in works. Political catastrophe rather than temperament killed certain projects, and the three late panels were not necessarily free of obligations merely because they could travel. The comparison establishes something narrower. A wall commission was bound to its wall; a clay monument depended on the regime, metal and site that were supposed to turn it into bronze. When those relationships broke, the projects could not simply accompany Leonardo into another court. A wooden panel could. Portability did not cause retention, but it made continued access physically possible.
That narrower late pattern is real enough to matter. The three paintings in France suggest that the Mona Lisa was not unique because Leonardo alone could not bear to release it. It belonged to a small class of paintings he went on retaining. The evidence does not let us turn that group into a theory of everything he made. It does let us stop treating one portrait's survival in his household as an isolated psychological symptom.
That transforms the psychological question. Maybe the defining act was not a refusal to finish. Maybe it was a refusal, or simply a failure, to close access. The panel remained open.
Think back to its physical history. Before the face existed, Leonardo experimented with the ground. He transferred a designed composition. Then the design moved, and we know it moved because another painter working in his orbit followed changes now buried beneath the Louvre surface. His understanding of facial anatomy deepened. His study of optics deepened. His treatment of shadow became almost perversely subtle. One- and two-micrometre glazes accumulated until visible boundaries dissolved. The landscape gathered distance behind the sitter. The face became less like a Florentine portrait and more like a demonstration of what painting could do to the experience of life.
The Louvre itself now describes something like this transformation, suggesting that in the course of its prolonged execution the work probably became more than a simple portrait of Lisa: the identity of the sitter receded in importance as the picture became an extraordinary demonstration of painting’s ability to recreate living presence.
That may be as close as we can come. The Mona Lisa took sixteen years because after some point, “taking” was no longer what the picture was doing. It was not sitting in a queue waiting to be completed. It was available. It could receive whatever Leonardo had discovered since the last time he touched it: a mouth understood differently after anatomy, an edge understood differently after optics, a shadow altered after another experiment with glaze, a landscape capable of holding his observations about water, erosion, atmosphere and distance.
This does not mean every scientific notebook entry migrated onto Lisa’s face. It means the category of the object changed.
A normal commissioned portrait is supposed to leave. The patron pays, the painter finishes, the object crosses a threshold, possession moves from workshop to household. At that moment, change becomes difficult. Leonardo never crossed that threshold with this picture, and an object that never leaves can remain provisional long after it is already magnificent.
This is why perfectionism is such an impoverished explanation. Perfectionism assumes Leonardo could see a destination and was incapable of accepting that he had reached it. The evidence points toward something stranger: that he may not have thought of his most important late objects as having the kind of destination we keep imposing on them.
His notebooks behave this way too. Thousands of Leonardo manuscript pages survive. He returned to subjects, reorganized them, copied them, cross-referenced them, planned treatises. He published none of the great synthetic works we now wish he had completed. After his death, Melzi organized material on painting into what became the Treatise on Painting, and a printed version did not appear until 1651.
We do not usually say Leonardo spent decades failing to finish a notebook. The category protects him from the accusation, because a notebook is allowed to remain alive. You reopen it because something else has occurred to you. You turn back three pages. You add a line. You contradict yourself. You leave gaps. You carry it when you move. You do not ask which day it was finished, because finished is not necessarily what that object is for.
The Mona Lisa was not literally a notebook. It had a sitter. It seems to have had a patron. It existed within an economic system that notebooks did not. Push the analogy too far and it breaks. But by the final years of Leonardo’s life, the painting appears to have acquired something notebook-like: the right to remain open.
That may be what Francesco del Giocondo never received. Not a finished painting withheld from him, but a painting that had ceased to behave like something awaiting delivery. The distinction is small enough to disappear in a museum label and large enough to change the story.
Leonardo did not spend sixteen years trying to make one perfect face. He spent sixteen years living with a surface he had not surrendered, and because he had not surrendered it, it could change whenever he did.
That is why the most revealing verb in the story is not finish. It is keep.

XIV. THE FIRST PERSON WHO CALLED IT UNFINISHED
There is a final irony in returning to Vespucci.
October 1503. The painting is new. Leonardo is fifty-one. Lisa may still be sitting for him. The Great Council Hall is ahead, and Milan, and the bodies, and the years with Marcantonio della Torre, and Rome. Giuliano is alive. François I is a child. The Alps are thirteen years away. Leonardo has no idea that he will die in France.
Vespucci opens Cicero and encounters Apelles. A masterpiece. A portion refined. A portion left incomplete. He thinks of Leonardo.
Not sixteen years later. Immediately.
The first surviving observer of the Mona Lisa does not wait for Leonardo to become old before noticing this characteristic. He sees it at the beginning: this is how Leonardo does his paintings.
That is the clue. Perhaps the sixteen years were never an aberration. Perhaps the painting did not become trapped by Leonardo’s personality, but was born into his method. The later legend of Leonardo the incapable finisher reverses the direction of the evidence, imagining a man trying and failing to terminate his work. Vespucci sees something else: a way of making in which completion is already unstable.
And maybe this is why the Mona Lisa can survive the scrutiny we have imposed on it. People have spent centuries trying to solve the painting because they assume it contains a solution. A secret sitter. A secret code. A hidden portrait. A mathematical key. A buried confession. A definitive smile.
Leonardo’s working life suggests a less satisfying possibility. There may be no locked box. There may only be layers: question laid over question, observation over observation. A face whose boundaries were softened because Leonardo distrusted boundaries. A mouth made unstable because vision itself is unstable. A landscape in which solid earth dissolves into atmosphere. A picture that refuses the clean ending modern viewers keep asking it to provide, not because Leonardo failed to reach one, but because he kept finding somewhere else to go.
# CODA

THE EMPTY SPACE
The Mona Lisa did not become the most famous painting on earth in Leonardo’s lifetime. It was admired. Artists knew it. Raphael appears to have responded to it. Vasari helped establish its legend in the sixteenth century. Nineteenth-century writers piled new fantasies onto the face, turning Lisa into seductress, sphinx, fatal woman and secret. It was famous before 1911.
But fame has scales.
Then, on 21 August 1911, the painting disappeared from the Louvre. Vincenzo Peruggia, an Italian worker who had been employed around the museum, removed it and carried it away. For more than two years it was missing. Newspapers followed the hunt. Suspicions multiplied. Crowds came to the Louvre.
And something peculiar happened. People came to look at where the painting was not. An empty space became an attraction.
When the picture was recovered in Italy in 1913, it returned not merely as a canonical Renaissance painting but as a mass-media celebrity. The Louvre itself now credits the theft with giving the Mona Lisa a new global and popular level of fame.
The sixteen years had almost nothing to do with this. Leonardo could not have imagined the newspaper photographs, the reproductions, Duchamp’s moustache, the advertisements, the postcards, the barricades, the climate-controlled glass, the forest of phones. The most famous thing about the painting became, eventually, that it was famous.
And yet the strange logic of possession never entirely disappeared. In January 2025, France announced a vast redevelopment of the Louvre intended to be completed around 2031. Among the plans is a dedicated space for the Mona Lisa, reachable independently of the rest of the museum and with its own access ticket. Five centuries after Leonardo carried a Florentine portrait out of the world for which it had apparently been made, one of the great museums of the world is reorganizing itself around the problem of how to let people reach it.
There is a temptation to end with the smile. Everybody does. The smile has endured enough interpretation. End instead with the object.
A piece of poplar.
Roughly eighty centimetres high.
A strange lead preparation beneath the paint.
A transferred drawing.
Changes hidden under changes.
Glazes thin enough to measure in micrometres.
A merchant’s wife.
A dead notary.
A ruined wall.
A dissected face.
An assistant painting beside the master.
Milan.
Rome.
France.
A visitor in 1517.
Three pictures in a room.
A will.
Death.
Then kings.
Revolution.
A museum.
A theft.
An empty hook.
A crowd.
Another crowd.
Another.
For sixteen years, Leonardo kept the panel close enough to change. That is the fact underneath the legend.
Perhaps there was no single moment when the Mona Lisa became finished. There was only a last moment when Leonardo was capable of touching it. Then the opportunities to return ended, and history took the brush away.
The painting left the category of work and entered the category of masterpiece. We have been calling that transition “completion” ever since.
Leonardo never did. And the first man who saw him working on it may have understood why.