Nine Days in New York

I. The First Minute (Days 1–3)

Tab. I

The first thing Dan told us, and he told us it in one form or another for three days, was to feel the weight of the thing we were drawing.

I've paid for enough workshops to have a reflex about sentences like that one, and the reflex is to nod and wait for the instructor to get back to the technical material. It took me most of the first morning to work out that this was the technical material. The outline is a fact about me. The weight is a fact about her.

Move my stool eighteen inches and every contour on the model changes. Not subtly — the whole silhouette reorganizes, because a contour is only the set of places where a volume turns away from the particular point my eye happens to occupy. It's a report on the relationship between us, and the half that moves when I move is the half I contributed. It doesn't belong to her at all.

Her weight doesn't change when I move. Neither does the structure carrying it: which hip is loaded, how the ribcage is turned on the pelvis, where the line of gravity runs down through the stack and into the floor. Those are properties of the thing itself, true from every stool in the room, true from behind her, and still true in the dark.

So it isn't that I should avoid drawing what I see. I'm looking for six hours a day and my eyes ache by lunch. It's that the easiest things to see are mostly information about my own position, and the things worth drawing are harder to see because they aren't visual. Weight has no appearance. There's no wavelength for heavy. I have never once seen a load. I've only felt one, and the whole apparatus I own for knowing about weight is in my inner ear and the soles of my feet and the muscles down the backs of my legs.

Which is why several of the instructions weren't about looking at all.

I can trace a coastline with perfect accuracy and know nothing about the country.

What I was supposed to draw instead was the impulse — and that word is misleading in a specific way. It sounds vague and mystical. It isn't. Impulse means something I can be wrong about, and be told I'm wrong about, in front of people.

The impulse is not the pose. The pose is what the model settles into after about four seconds, when she's found the version she can hold for twenty minutes without her balance failing. The impulse is what she did to get there — the intention that threw the ribcage back and swung the pelvis forward and left her whole weight hanging off one hip like a coat off a hook. It's a verb. The pose is the noun the verb fossilized into. Draw the noun and you get taxidermy.

The question is at the top of my notes from the first morning and again, in worse handwriting, on the second: What is the primary impulse of the pose? Singular. Primary. Not the impulses I might catalogue at leisure — the one. Decide what the body is mainly doing, and decide it before I have any information I'd consider adequate.

Tab. II

The materials, which are a position

Graphite only. Tombow HB, B and 2B, a value range so narrow you could hold all three and not tell from four feet which made which mark. A paper stump. Strathmore 400 Series, a decent toothy sheet that costs actual money — unlike newsprint, which doesn't, and which nobody hesitates before ruining. No charcoal anywhere in the room, and its absence isn't an oversight.

Charcoal is forgiving and theatrical. It goes down black, lifts with a chamois, has a velvet in it, and it will make a mediocre drawing look decisive because the material is decisive and the drawing is borrowing against it. Graphite won't. Graphite is grey, with a slight sheen that catches the overheads and shows every place I went over something twice — a visible record of my hesitations. A 2B pressed as hard as the paper will take isn't very dark. You can't bluff in graphite. Whatever authority the drawing ends up with, I put there on purpose.

The stump is the one soft instrument in the room: wound paper, no binder, no pigment of its own, useless until it's charged with graphite. Its whole function is to take a mark I've already committed to and smear it into a volume. It can't originate anything. It's a tool for turning a decision into a mass.

Ascending

Day one ran one minute, then two, then five.

There's nothing unusual in that shape. Every life room I've drawn in ascends, and I'd assumed for years that I understood why — the short poses are a warm-up. Loosen the arm, calibrate the eye, shake out whatever came in from the street, and the actual drawing begins when the model settles in for twenty.

That's not what these were. The short poses weren't preparation for the work. They were the foundation of it, and it took me most of the first morning to notice the difference.

The point isn't to learn to draw fast. The point is that the first minute of a twenty-minute drawing is the same minute as a one-minute drawing.

The most important decisions of any drawing are made at the beginning of a pose. So you build up. One, two, five. By the end of the second day we were given ten, and then twenty, and by then it was clear that a twenty-minute pose is not twenty minutes of drawing. It's a one-minute drawing and then nineteen minutes of testing it.

Which is not the same as nineteen minutes of protecting it, and the difference is the whole method. The early process is a sequence of statements: put down a position, then relate it to another part of the form and find out whether the first one holds. When a relationship is off, adjust. That's what the eraser is for. What you don't do is treat the founding marks as sacred — you treat them as the claims the rest of the drawing is going to interrogate.

The clearest instance of this was a move of Dan's that I've been using ever since, and it starts before any of the figure exists.

Fill the page. He discouraged us from drawing small. First thing, before anything else, mark a standing pose as close to the top and bottom of the sheet as you can get away with — establish the two extremes, then go find the main line of action.

The reason isn't compositional, or not only. Those two marks are the drawing's fixed frame, and having a fixed frame is what makes the rest of it correctable. Because if you get down the figure and realize she's come out too tall or too short, the instinct is to move the top or the bottom. Don't. Leave them. Narrow or widen the figure instead — and the proportions resolve without the height of the drawing ever changing. You make the figure taller by making her thinner. The frame stays put and the figure adjusts inside it.

That took me a while to believe and it is exactly right. It also means the first marks aren't fragile. They're the fixed points that let everything else move.

So the one-minute pose isn't a sketch. It's the opening of a long drawing, extracted and practiced in isolation, the way a pianist works four bars for an hour. I wasn't being taught to make small drawings quickly. I was being taught to survive the first sixty seconds of a large one.

What nobody warned me about is what the long poses feel like once you're acclimated to the short ones.

I'd assumed the twenty would arrive as a relief. It arrived as vertigo. I'd get to minute two with the thing essentially decided — impulse established, weight sorted, plumb dropped — and then look up and there were eighteen minutes left and the model hadn't moved and wasn't going to. In an ordinary life room the complaint is always that there wasn't enough time. Here I sat in front of a twenty-minute pose holding a drawing that had already said most of what it had to say, with no idea what to do with the rest of the afternoon.

That's a real problem, not a rhetorical one, and I didn't solve it in three days. What you're supposed to do with the surplus is develop: take the founding decision and press it into the specifics, resolve the structural framework into the actual forms, supply what the first minute could only imply. Refine your understanding of what you just felt.

What I was strongly tempted to do instead was fiddle. Add. Render. Drill into the detail long before the drawing had earned any. Find something, anything, to be busy with.

The reason is that the alternative is a state I already know from somewhere else. I meditate most mornings before I paint, and a twenty-minute pose after its second minute is the cushion exactly. You have hold of something open and unfixed and faintly ungraspable, your job is to stay in contact with it, and the mind will not have it. The mind wants a task. It offers the eyelashes. It offers the precise edge of the sternocleidomastoid. It offers anything at all rather than sit still in front of a thing you felt once and are trying to keep feeling, and every one of those offers is a way of leaving the room.

Noodling is the monkey mind with a pencil in its hand.

One sheet of grey paper

We erased between every pose. Every one, same sheet. The 400 Series takes it, more or less, and over a session the paper develops a grey ghost-history — an accumulated fog of everyone who's stood on that stand — and you draw into your own residue.

Erasing itself isn't new to me. I've erased a great deal over a working life and there's nothing unfamiliar about lifting a passage out and putting it back down better.

What was new was walking away from two days with a cloudy grey sheet and nothing else. Practice leaves a record. Every session I've ever put in has produced a drawing of some kind — not necessarily a good one, frequently a bad one, occasionally one I wouldn't show anybody — but something: physical evidence that the hours occurred and that I was the one who spent them. The sheet is the substrate and the drawing is what ends up on it. That arrangement is so basic I'd never thought to check whether it was required.

It isn't.

Tab. III

The eraser measured how much went. I brought a kneaded eraser from home: grey putty with no abrasive in it, which lifts graphite by absorbing it, and which you fold in on itself to bring a clean face back, like working dough. There's a finite amount of folding available. By the middle of the second day mine had taken on so much graphite that it stopped lifting and started depositing — I'd touch it down to clear a line and it left a soft grey smear instead. The eraser had become a stump. It had absorbed enough of my work to be converted into its own opposite, and I went and bought another one at the art supply store in the building.

I've thought about that lump of putty more than is reasonable. The first two days didn't vanish. They went into the eraser. Everything I made on those two days is in a ball of grey rubber I kept as a souvenir, and that's the only archive of it there will ever be.

Which is the point. The erasing isn't housekeeping, it's a teaching instrument. It removes, in advance, the possibility of the drawing becoming the thing I'm working on. I can't get precious about a mark I'm going to lift in ninety seconds. I can't bank a good passage and start protecting it. There's nothing to defend, so I attack.

Three boxes

The reduction is more severe than I expected and more useful for being severe. Three masses: head, torso, pelvis. Not eggs. Not buckets. Not the tapering cylinders most figure instruction reaches for the moment things get curved.

Boxes. Rectangular solids with a top, a front and a side, and nothing else permitted.

Nobody's ribcage is a box. Everyone in the room could see it wasn't, and that's the point. A box has three faces and therefore three axes, so the instant I commit to one I've committed to a statement about which way that mass is pointing — rotation, lateral, sagittal, three claims, all falsifiable in four seconds by anyone standing behind me. Two boxes give the relationship between two masses. Three give the pose. Everything else in the figure, every limb, every contour anybody would call beautiful, is downstream of what those three solids are doing to one another.

The abstraction gets closer to the truth than the accuracy does. The contour — the sensitive tracing of the edge that many people believe drawing consists of — makes no claim at all. It can't be wrong, because it hasn't said anything.

And the boxes have to be found from evidence, which produced the most useful sentence of the three days: look to the gladiolus from the front or the thoracic midline from the back, not the shoulders, to determine the rotation of the ribcage.

The shoulders are liars. The scapulae float on the back of the thorax, sliding around like two plates on a barrel, and the clavicles swing, and the whole shoulder girdle can rotate a substantial angle off the ribcage it appears to be attached to. Draw the ribcage from the shoulders and I'll build a box pointing somewhere the ribcage isn't. The gladiolus — the long central blade of the sternum, between the manubrium above and the xiphoid below — is bone, joined to the ribs, and it can't go anywhere the thorax hasn't gone. Find the gladiolus, find the front face of the box.

The sternum, incidentally, is a sword. Manubrium is the handle, gladiolus the little blade, xiphoid the point, from the Greek xiphos. Somebody looked at a breastbone and reached for the nearest object on the wall — which is what I'd been doing all morning by calling a ribcage a box.

Weaving

I blocked in with long straight lines, let them run past where they should stop, let them cross. Don't stop a line unless you have to. Dan calls them his 2x4s, and once he says it you can't see them any other way — at the five-minute mark the drawing looks less like a woman than like the framing of a house, studs run long and lapped past their joints, a structure that will hold a figure but isn't one yet.

It's the right word for them because a 2x4 is dimensional lumber. It is milled to a standard, it is interchangeable, and nobody frames a house out of curved ones. The straight line is the stock you build in.

The word for this is weaving, and it's the right word: the lines interlock, pass over and under, and the figure emerges as the thing they're collectively cinched around. The instruction attached to it — think about how your lines call to and answer one another — is a larger claim than it sounds, because it means the marks have obligations to each other independent of the model. If the marks answer only to the model, the drawing is a measurement. If they answer to each other too, it's a picture.

The straight line is a decision. The curve is a hedge — a curve can always be a slightly different curve and nobody will call you on it, whereas a straight line from the pit of the neck to the far knee is a claim about the whole mechanical situation, checkable with a plumb, falsifiable in one second. Robert Henri got there first: he wrote that good composition works like a suspension bridge, every line adding strength and none subtracting, and that lines merely running into each other isn't composition — there has to be a motive for the connection. He also wrote that a curve doesn't reach full power until you set a straight line against it, which is the first workshop in a sentence.

And in a standing pose the plumb has somewhere specific to land: the core weight travels up and through the ear. Weight-bearing foot, the stacked negotiation of the three boxes, then the ear — not the top of the head, not the nose. The ear. The head's position on the neck isn't a detail of the portrait, it's the last term in a load-bearing equation that starts at the floor.

Shelf life

The negative shape has a shelf life. Use it in the first twenty minutes; the model will shift, and it becomes less reliable as time goes on.

A practical observation about a practical technique — measuring the pocket of air between an arm and a ribcage is often easier than measuring the arm. It's also a compressed statement of the entire problem.

The model drifts. Not out of laziness. Out of being alive. Weight settles, a hip creeps, a hand slides a quarter inch, breathing happens, and the shape I measured at minute three is a different shape at minute eighteen. The living body has a shelf life as an object of measurement, because it won't consent to be an object. Everything reliable about it is reliable only briefly. I'm working against a decay curve, and the decay is the life.

I wrote that down as a scheduling note. It turned out to be the thesis.

Tab. IV

The two hard ones

Draw through the eyes, not with the eyes.

Cultivate a sympathetic emotional reaction to the model's impulse.

The first is nearly Nicolaides, who defined the seeing he meant as an observation using as many of the five senses as can reach through the eye at one time. The preposition is his. Dan's version compresses a long sentence into eight words and loses nothing I can identify. Nicolaides taught at the League for fifteen years and died in 1938, so the line was already in the building.

The second is the one I underlined and didn't, at the time, entirely believe. It sounds like the soft, unfalsifiable thing that gets said when the technical material runs out. But it's an instruction with a mechanism: feel, in my own trunk and hip and shoulder, what her trunk and hip and shoulder are doing — run the pose in my body — and draw from that rather than from the retinal information. It's a request to use proprioception as a drawing instrument. I have a lifetime of expert data about what it feels like to hang my weight off one leg, and none of it is available through the eye.

Late on the second day the model went up into a shoulder stand and held it for twenty minutes. I have no lifetime of expert data about that. I know a loaded shoulder and a compressed neck and I have never once put them together upside down, and drawing it worked anyway. Make of that what you will.

The reason I've decided this isn't mysticism is that it produces checkable results. A drawing made this way and a drawing made by measurement don't look the same, and other people in the room can sort them without being told what they're sorting for.

She rests every twenty minutes and puts on a robe and looks at her phone, and the transition is instructive, because the instant she stands and reaches for the robe she does about nine things with her body more interesting than anything she held on the stand. The impulse in the wild. Unposed, unmeasurable, gone in under a second.

The third day

And then on the third day some of it stayed.

Tab. V

There's a version of that sentence that's a redemption arc, and I don't think that's what happened. What happened is duller and more mechanical. By the third day the poses were long enough that the founding minute had somewhere to go afterward, and I'd done enough repetitions of that minute that I was no longer spending it deciding how to spend it. Those two conditions arrived at the same time. A few sheets from that last afternoon are the first ones all week I didn't want to erase. Nobody made me. I brought them home.

There aren't many, they aren't finished, and by the standards I'd apply to my own work they aren't competent. They're fast and thin and they leave most of the figure out. But they have the thing in them, or something adjacent to it, which the careful, rendered, correct drawings I made later that week don't have and structurally can't.

Two days of grey paper. A dead eraser. A few sheets that survived. And a definition I was pleased with, which the following week took apart: the figure is not a shape. It's a shape doing something.

II. The Rare Book Room

Tab. VI

Fifth Avenue and 103rd, far enough up that Central Park has gone quiet and slightly feral, in a building with the overbearing Romanesque confidence of an institution that was very sure of itself in 1926. Upstairs, a reading room. Foam cradles. Snake weights — long fabric tubes of lead shot you lay across an open page to hold it, with the disconcerting warm heft of something living.

The first surprise is that you don't wear gloves. This runs against every instinct and every movie. The reason is good: cotton gloves deaden the fingertips, deadened fingertips fumble, and a fumble is how a 1543 folio gets a new tear. Clean dry bare hands do less damage than clumsy gloved ones. So they hand you a Vesalius and you touch it with your skin.

I will not pretend to have been calm about this. The paper is better than any paper I have ever bought — rag, soft, opaque, four hundred and eighty-odd years old and still supple — and I could see the deboss left by the type from 8 feet away, feel the plate mark with a fingertip, the physical dent where a copperplate was pressed into damp paper by a man operating a screw press with his body. The book isn't a carrier of an image. It's the record of a pressing, and you can feel the pressure.

Tab. VII

Dürer, Vier Bücher von menschlicher Proportion, printed in 1528 a few months after he died. Not anatomy in the dissecting sense — Dürer never opened a body — but an attempt to derive the human figure from measurement, from ratio, from a system, page after page of bisected men and women reduced to schedules of thirds and sevenths. The most hopeful book in the room and the least useful, trying to do arithmetically the exact thing I'd spent three days doing with three boxes and no arithmetic.

Tab. VIII

Vesalius, the Fabrica, 1543. Everyone knows the muscle-men. What reproductions don't tell you is that the figures are standing. Flayed to the third or fourth layer, skin removed, musculature peeled and pinned back in swags like drapery, up on their feet in the Euganean hills outside Padua, leaning on things, gesturing, contrapposto, alive. Line the plates up in order and the landscape behind them is continuous — one long panorama the dead men walk across as they're progressively taken apart. By the last plates there's almost nothing left of them and they're still standing.

Tab. IX

Bidloo, 1685, one hundred and five plates drawn by Gérard de Lairesse, and this is the one that took the afternoon away from me.

Bidloo gets described as the anti-Vesalius, and he is, but not the way I'd absorbed secondhand. His bodies aren't standing in Italy. They're lying on a table with rumpled cloth beneath them, pins holding flaps of skin open, a rope hitching a limb out of the way, and in one plate a fly on the specimen, drawn with the care given the specimen. The standard line is that this makes Bidloo the realist: he gives you the mess, the pins, the Tuesday-in-Amsterdam of it.

That isn't what's on the page. These are composed. Lairesse was a working painter of real ambition and he built every plate as a still life: drapery arranged, light raked in low from one side, the specimen placed with the deliberation you'd use placing a lemon, a knife and a facing-away silver cup. The rope isn't documentation. The rope is a diagonal. The cloth isn't the cloth that happened to be there; it's a mass of broken white the composition required in that corner. The fly is a memento mori and a small dark accent in an area that needed one.

The consequence, which I doubt Lairesse minded and the surgeons certainly did: they're less informative than the other books. Substantially less. Where Vesalius or Albinus hand you a muscle in its clearest aspect — isolated, unobstructed, lit for legibility, a diagram with a body around it — Lairesse gives you the same muscle at whatever angle the composition wanted, half in shadow, partly occluded by a fold of skin he kept because the fold is beautiful. As reference material they're measurably worse.

As pictures they're almost unaccountable, and they have a quality of uncanny realness none of the idealizing books come near. The plates feel like they are of something in a way the perfected ones don't. I think the composing does it, which sounds backwards and isn't: an arrangement implies an arranger, and an arranger implies a room, a table, an afternoon, and a hand that put the cloth there and stood back to look.

Lairesse's own life is one I didn't know standing at the table. He had congenital syphilis, which by his forties had visibly deformed his face. Rembrandt painted him in 1665 with what the Met calls uncompromising directness, and that portrait hangs about forty blocks south of where I was sitting — though comparison with the underdrawing shows Rembrandt made the nose smaller than it was. The most unsparing portrait in the room, and the painter still flinched a little. Lairesse went blind around 1690, at forty-nine, and having lost the ability to draw the bodies of the dead he spent his remaining years dictating a systematic treatise on the rules of ideal beauty, in which he dismissed Rembrandt's paint surface as resembling liquid mud on the canvas.

The plates themselves were later bought in bulk by the English surgeon William Cowper, who republished them in 1698 with his own name on the cartouche of Lairesse's frontispiece. Bidloo took it to the Royal Society; the pamphlet war reads like a comment section. Cowper's version is the one that circulated.

Tab. X

Albinus, 1747, engraved by Wandelaar, and here the pendulum swings all the way back. Albinus was after the ideal skeleton — not this skeleton, the skeleton — and to get it he suspended the specimen with cords, interposed two nets, one four feet away and one forty, drew through both, and did this over months. The result has the eerie correctness of something no one has ever seen. Then, because the plate needed a background, Wandelaar put a rhinoceros behind it. An actual rhinoceros, Clara, who was touring Europe at the time. The perfected human skeleton in front of a live celebrity rhinoceros: the most eighteenth-century object I have ever handled.

Lizars (Edinburgh, 1820s, hand-colored, and one doesn't think for very long about where an Edinburgh anatomist in the 1820s was getting his material).

Tab. XI

Tuson, Myology, 1828, which I'd filed in advance as a minor item and would now put near the top.

It's a flap book — lift-up paper dissection, and as it turns out the largest and most complex ever attempted. Seventeen plates carrying better than two hundred movable flaps, every one a separately printed scrap of lithograph, cut out by hand, colored by hand, hinged into position. Every copy. The edition was necessarily small because there's no fast way to make this object. Somebody sat at a table and cut out two hundred paper muscles, set each one where it belonged, got them to lie flat, and then began the next copy — so a student in 1828 with no access to a body could dissect one with his fingers.

Then the detail that stopped me. In most flap books the flaps hinge along the top and lift like a row of little doors. Tuson's don't. Each tab is anchored on a different side, so the flaps interlock — they pass over and under one another, and the stack feels interleaved rather than layered.

I'd spent the previous three days being taught to build a figure out of long straight lines that overrun and cross and cinch a body between them, and here was a man in 1828 doing the same thing in paper, for the same reason. The muscles interlock. Any honest account of them has to interlock too.

It's also the only book on that table where the veil is reversible. Vesalius takes the skin off and it stays off for four hundred years. Tuson lets you take it off and put it back. Close the flap and the man is whole again.

And Maclise, Surgical Anatomy, 1851, with enormous lithographs that are better drawings than most of what's in any museum on that side of the park. The bodies are white and Black — young men mostly, handsome, several of them god-like, drawn with a care that isn't the care of a diagram. And they're composites: Maclise worked from living models he found in the streets of London and Paris and married those figures to dissections done on corpses out of the Paris morgues. The plates are half life drawing.

The Black figures were taken out of the American editions. Philadelphia, 1850 and 1851, on the eve of the war, and it's a decision that can be given no benefit of the doubt. In at least one plate the man wasn't simply dropped but redrawn as a white man, which is worse, and which required somebody to sit down and do the work.

And then, at the far end of the table and four centuries downstream of everything else on it, Vanderpoel — John H., The Human Figure, 1907, which isn't an anatomy book at all and was the most useful thing I touched all afternoon. Vanderpoel was a Dutch-born painter who taught at the Art Institute of Chicago, and the book is his lectures with his own charcoal drawings, going down the body part by part. He never opens anybody. He works entirely from the outside — the planes of the deltoid and where the light breaks across them, the way the mass of the calf hangs lower on the inside than the out, the ledge the clavicle makes. No scalpel, no pins, no rope, no fly. He describes the veil at length and in total seriousness, as though the veil were the subject rather than the obstacle. He taught O'Keeffe, who said for the rest of her life that he was one of the best teachers she ever had, and the book has never really gone out of print — so of everything on that table it's the only one I could have replaced on the walk back downtown for fifteen dollars.

What I took away, and didn't expect to, is that this is a tradition with an argument in it, and the argument is about what you put back.

Vesalius and Albinus put back the life. They stand the corpse up, hand it an attitude and a horizon, and grind the particular skeleton down until it becomes the Platonic one.

Lairesse refuses the standing-up and puts back the frame. He leaves the body flat on the table where it is and composes the table. Nature morte — which is what the French call a still life and which means, with no euphemism available, dead nature.

Vanderpoel puts back nothing, because he never took anything off.

Three strategies. I didn't yet understand they were three answers to the same question, or what the question was.

III. Four Days at Cornell

York Avenue in the high sixties, a stretch of Manhattan that's essentially one continuous hospital with occasional delis. No locker and no coat. I went in wearing what I wore on the subway and came out wearing it, and it came home with me — the smell gets into the weave and it was still in the hotel closet that night and still faintly in that jacket in California.

What I did get was gloves. Nitrile, and this is the joke of the trip, which I didn't see until I was standing there with my hands up: at the library they took the gloves away, on the grounds, I assume, that bare skin is gentler and the object was the fragile thing in the room. Here I put them on. Here the object is not the fragile thing in the room. And I wanted them urgently, within about ninety seconds of the first contact, and the wanting wasn't squeamishness so much as an accurate early instinct about what I was touching and what it intended to leave on me.

Every so often — I never worked out the interval, and I suspect it was by feel — a technician came through with a spray bottle and misted the fleshy specimens to keep them from drying. Formalin. Cut with fabric softener.

Fabric softener. The product from the laundry aisle, formulated to make a towel feel expensive and to smell like a meadow that doesn't exist anywhere. It's an entirely sensible thing to do: the room has to be habitable for eight hours, formalin is brutal on the eyes and sinuses, and if a thing off a shelf makes it bearable then you use the thing off the shelf.

But it means the smell I carried back to California is a compound, and the two halves are doing opposite work. Underneath, the fixative — the smell of the process that converted a person into a specimen. Over the top, reapplied all day, an artificial floral engineered to make the underneath acceptable. A scent whose entire function is to stand between a person and what they're actually smelling.

I didn't have language for why that bothered me until several days later.

I should also say plainly what I looked like in that room. A man in his own street clothes with a paperback of continental aesthetics in his bag and a Pantone deck in his pocket, having Thoughts. The people in this building were medical students who were going to have to do this for real, on the living, with consequences. Every one of those bodies was a person who filled out a form, on purpose, so that others could learn from them, and the form did not say so that a painter can work out his position on Byung-Chul Han. I don't have a resolution to this. I registered it every morning and it didn't get lighter.

There is no life in them. Not a trace. Not a residue. Nothing.

I've killed and butchered a fair number of animals. Deer, hogs, chickens. I've had my hands inside a deer within an hour of its death, up to the wrist in the chest cavity, and the thing about a freshly killed animal is that it's loud with color. Blood nearly black in the body that turns scarlet the moment air hits it. The plum of the liver, the wet aubergine of the kidney, fascia in sheets like mother-of-pearl with an actual iridescence to it, fat that's cream or butter or, in a fall deer, nearly white. Steam. A dead deer is a living animal that has just stopped. Every color in it is a color it had ten minutes ago. Death has occurred but the body hasn't heard about it yet.

A fixed cadaver has heard about it. It has been informed.

They're brown. Not brown like skin. Brown like a piece of furniture, like a wet paper bag, like jerky, like the underside of a mushroom that's gone. Grey where they aren't brown. The fat is a yellow-grey, greasy in a way that transfers to the glove and stays there. Everything is slick, and everything has the same slickness, which is part of the horror of it. In a deer the wet parts are wet differently, the pericardium isn't wet like the diaphragm. But here the fixative has gone through everything and made it all one substance, uniformly slimy and uniformly the color of nothing.

And they're deflated. That's literal, not figurative: the vessels are empty. The body has been drained and perfused and the pressure is gone out of it. A living body — and, for about an hour, a dead one — is under pressure. It's inflated. It presses back. These don't press back. I lifted a flap of tissue and it stayed where I put it, like wet felt.

The deer is dead. The cadaver has been preserved, which is a different category altogether. Death is an event. Preservation arrests the event and holds it open indefinitely, and what it produces isn't a dead person but a specimen, a third thing, neither alive nor dead but converted.

The absence of mystery

I had Saving Beauty in my bag that week, which was either the best or the worst possible piece of luggage.

Han's claim, roughly: the smooth is not beautiful. The smooth — the polished screen, the depilated body, the image that offers no resistance — eliminates negativity. It can't wound you. It solicits the "like" and nothing more. Beauty is inseparable from concealment: beauty is not the object, and not the veil, but the object in its veil. Take the veil off and you don't get a more intense beauty. You get an object. You get information.

The cadaver lab is the veil removed. Completely, methodically, layer by layer, with a scalpel, on purpose. What I kept waiting for — what I think I'd assumed for twenty years would be waiting in there — was some final chamber, some point at which the accumulated peeling would deliver a thing I couldn't have anticipated. It never came. It's muscle, and then more muscle, and then bone. It's all exactly what it says it is. The forearm is a bundle of straps that pull on things. The hip is a mechanical joint with a lubricant. There are no secrets in there, because a secret is a thing being kept, and nothing in a body is being kept from you.

This is the honest version of an experience that gets described dishonestly a lot. People come out of these labs talking about awe and the miracle of the human machine, and I believe some of them, but I didn't have that, and I've come to think the absence is the finding rather than a failure of receptivity on my part. Anatomy is the search for a secret that turns out not to be there. The mystery of a body was never located in its interior. We just assumed it was, because that's where we couldn't see.

Tab. XII

The femur

I'd brought a Pantone fan deck and a blade for the express purpose, which sounds insane and produced a certain amount of comment.

What drew me to the femur was that it was the wrong color.

Every other bone in the room was bleached — the uniform, chalky, sanitized non-color of teaching-collection bone, which is the color people think bone is because it's the only bone most people ever see. This one had been somewhere. It had the look of something that had spent time in the ground, or had at any rate been exposed to some process of decomposition the others were spared, and it had come back stained and mottled and various. I recognized it immediately and it took me a minute to place from where: the catacombs under Paris, where the bone is stacked by the mile and no two femurs are the same color, because they've all been quietly taking on the character of the rock and the damp for two hundred years.

It was the one object in the room with a history written on its surface, and the only thing I saw there I'd describe as beautiful.

Nothing on it is white. The shaft is a warm greenish-grey, roughly the color of an old piano key or a fingernail, and it isn't one greenish-grey but a drift of them. Down at the condyles it goes translucent and horn-colored and much warmer, an amber that reads almost orange against the shaft. The linea aspera — the roughened ridge running down the back where the adductors and the vastus muscles anchor — is darker and browner, stained, essentially soiled, like the grip of a tool that's been used by somebody. The nutrient foramen has a dark halo around it. There are patches near one end that are frankly ochre and one small area I could only call a dirty violet. The whole thing is warmer than any painter would guess and lower in chroma than any painter would dare.

I drew it first — a straight sketch, no color, just the form. Then I cut chips to match, approximately, and pasted them onto the page around the drawing, each with a ruled line running to the place on the bone it belonged to.

Which is, I notice now and didn't notice then, an anatomical plate. Leader lines to labels. I'd spent the morning irritated by exactly that convention and by lunchtime I was using it, except that my labels weren't names.

The chips are a preposterous instrument for this. A chip is a fixed, industrially reproducible, absolutely flat quantity of color, and a femur is none of those things. The color changes with the angle you hold it at, because the surface is semi-translucent and partly specular, and the color I was matching was really a relation between the bone, the fluorescent tubes overhead, my grey glove and the stainless table. Every match was correct for about one second. Then I'd move my head and it was wrong. I ended up with fifteen chips that are individually all slightly wrong and collectively, as a run, sort of true — the same result as the three boxes in week one, arrived at from the opposite direction. The wrong thing, committed to, gets you closer than the right thing, hedged.

The names

We were taught the naming system for the muscles of the arms and legs, and it's one of the genuinely delightful parts of the week, because it isn't a system of labels. It's a system of sentences.

Every name is built from some combination of shape (deltoid, a delta; teres, round; gracilis, slender; piriformis, pear-shaped), size (maximus, longus, brevis, vastus), number of heads (biceps, triceps, quadriceps), fiber direction (rectus, straight; obliquus; transversus), location (brachii, femoris, profundus), attachment (sternocleidomastoid, brachioradialis), or action (flexor, extensor, supinator, levator).

So flexor digitorum profundus isn't a name, it's an instruction: the deep bender of the fingers. Extensor carpi radialis longus: the long straightener of the wrist, on the thumb side. I wasn't memorizing four hundred arbitrary tokens; I was reading four hundred very short, very old, extremely compressed descriptions written by people who had to remember all of this without photographs.

My favorite is sartorius — the long strap running diagonally across the thigh from the outside of the hip to the inside of the knee. From sartor: tailor. Named after a job, because it's the muscle that crosses your leg the way a tailor sits. The others in that vein: soleus, from solea, a sole, the flatfish. Pectineus, from pecten, a comb. Whoever named these was looking at objects and reaching for the nearest household comparison, which is exactly what I'd been doing with the ribcage-as-box three days earlier, and what the person who named the sternum after a sword was doing two thousand years before either of us.

That's the system working, and I don't want to undersell how good it is.

What happened over three days is that the system stopped being a way of thinking and became a thing to be recited. There's a mode that takes over a room like that, and I have no doubt it takes over medical schools too: an escalating enthusiasm for naming, in which every structure on every specimen and eventually on the living model gets pointed at and labeled, and then the labeling becomes the performance. Dan is a serious teacher and this is a serious pedagogy, and the naming has a real purpose. But it acquires a momentum. And there were students — not all of them, and I won't pretend I know their reasons — who had clearly worked out that the fastest route to visible engagement was to produce more terms, faster, and to appear moved by them.

Tab. XIII

I spent a certain amount of the second day thinking about Better Off Dead, which is not where I expected to arrive.

There is a scene in it, one of my favorite scenes in a movie I've loved since I was a kid, in which Lane Meyer is sitting in high school geometry. The teacher is at the board delivering a single unbroken sentence of jargon — connecting lettered points, quintuplicated triangles, a repeated vector sum, and somewhere in the middle of it a phrase about geometric configurations "which have no properties but with location" — before asking the class, mildly, to take out the homework. It runs about eighty words, it's one sentence, its grammar is immaculate, every term in it is a real term used correctly as a part of speech, and the whole of it means nothing at all. That's difficult to write on purpose.

The room is rapt. Enthralled. Gasping. Every hand in it goes up.

Every hand but one. Lane’s best friend, Charles De Mar spends the lecture in the front row with a fetal pig in a jar, nudging it, shaking it, and — if you watch closely — mouthing wake up at it. And when the homework is called for, even Charles surfaces from the pig and puts his hand up.

Which leaves exactly one person in that room who has nothing, and he knows it, and his defense is to raise a sheet of paper in front of his face and hold very still behind it.

He gets called anyway. Along with several others, up to the chalkboard, to present.

I'd always taken the joke to be about the students. Standing at a stainless table on York Avenue with a flood of Latin coming at me, I understood for the first time that it's about Lane — the specific loneliness of being in a room where everyone else is visibly getting something out of the words and you are not. Whether my head had simply hit its capacity, or whether I was in there looking for a different kind of thing entirely, I lost the thread repeatedly and completely. I could not make myself care about the knee bone being connected to the hip bone.

And the scene is more precise than I'd given it credit for. The failure isn't that Lane checked out. Everyone in that room has an answer except him, including the one who spent the entire lecture trying to resuscitate a pig. Even inattention is no protection. Lane's inattention isn't refusal or protest, it's an absence with nothing behind it, and the paper in front of his face is an attempt to make an absence take up less space.

There were parts of those four days where I'd have taken the sheet of paper. Not because I disputed anything being said. I couldn't have disputed it, that's the point. But because I had nothing to raise and could feel it being visible. And the thing at the front of the room was neither the teacher nor the terminology. It was a dead thing that I kept waiting on.

This is where Han landed for me, harder than the material on beauty did.

His distinction is between information and knowledge, and it isn't a distinction about quantity. Information is additive. It accumulates without ever composing into anything; you can always have more of it, and more of it never becomes understanding, because understanding requires the negative — selection, exclusion, the willingness to leave things out. A pile of information has nothing left out and therefore has no shape.

A complete nomenclature is a completed inventory. When every part of a thing has a name that tells you what it is and what it does, the thing has been fully accounted for, and nothing is left over, and nothing is withheld. The naming system is magnificent, and it's also the formal linguistic version of the same absence I'd been feeling all week: the body converted, exhaustively and correctly, into information.

Though the inventory is complete and the bodies aren't. Palmaris longus, one of the flexors in the forearm, is missing entirely in something like one person in ten, in one arm or both, with no consequence whatsoever — which is why it gets harvested for tendon grafts. There's a name in the system for a thing a substantial number of people simply don't have, and nobody who has it knows, and nobody who lacks it misses it.

The geometry teacher's line is better than the movie needs it to be. Everything in that room had a location. Nothing in it had properties.

The other drawings

Tab. XIV

A forearm and hand, anterior side up, with the skin taken off the arm and off the palm, so what's in front of me is the flexor side: the muscle bellies coming off the medial epicondyle, then the long tendons, then the hand.

What you draw is the narrowing. The flexor mass at the top of the forearm is broad and meaty and undifferentiated, and as it runs down it converges — the bellies give out and become tendon and the whole thing draws into a bundle you could close a hand around at the wrist, and then immediately past the wrist it opens out again and fans into the fingers. A wide thing, cinched, and then a wide thing. Once I'd seen that cinch I couldn't un-see it in a living arm, and it explains the wrist, which as a piece of drawing has always given me trouble and which I'd been treating as a joint when it's really a waist.

Tab. XV

A head.

Tab. XVI

A head and torso.

All of these were easy to draw. Alarmingly easy. They hold still — perfectly, indefinitely still, in a way no model on a stand can — and they ask nothing of you and they won't be there tomorrow to see what you made of them. That ease is a poverty. The three days on 57th Street were hard because a living body won't stop doing things at you, and won't consent to be measured, and has a shelf life. Nothing was being done at me on York Avenue. I was drawing an object, and objects are easy.

Which produces the arithmetic of the trip, and I didn't notice it until I was packing. Two days at the League produced one grey sheet and a ruined eraser. The third day produced a few thin, fast, unfinished drawings of a living woman. Four days on York Avenue produced a sketchbook — orderly, rendered, permanent, entirely correct — of things that were already dead.

Two stacks, opposite in every respect. The recordings of life are few and elusive and barely hold together. The renderings of death are numerous and are the best-made objects I brought back. I've gone through both a number of times since and I know precisely which ones I'd run into a burning building for, and it isn't the good ones.

IV. The Last Day

One day left, spent entirely at the Metropolitan, which is the wrong way to use that building and the only way I was able to manage it.

Tab. XVII

Vermeer, A Maid Asleep. Velázquez, Juan de Pareja. Lawrence's Elizabeth Farren. Bastien-Lepage's enormous Joan of Arc. Dagnan-Bouveret, The Pardon in Brittany. Repin's portrait of Garshin.

Tab. XVIII

Carolus-Duran's Fantin-Latour. Homer, Dressing for the Carnival. Hopper's The Lighthouse at Two Lights. Kollwitz's ploughmen, bent double and hauling. Dalí's hypercube crucifixion, which I didn't expect to stop at and stopped at.

Written out, it's almost entirely faces, and almost all of them are looking back at me.

Juan de Pareja looks at you with a composure that's difficult to describe and impossible to mistake. Velázquez painted it in Rome in 1650 as a demonstration piece and hung it in the Pantheon, where it made his reputation in the city before he went on to paint the Pope. Pareja was in his household as an enslaved assistant at the time. He was freed, and he painted, and The Calling of Saint Matthew is in the Prado with his signature on a piece of paper on the floor. Two centuries after Velázquez sat him down and painted him like that, a publisher in Philadelphia was quietly taking Black men out of an anatomy book. Both things are true. The painting was there first.

Fantin-Latour looks at you. Garshin looks at you with an unbearable gentleness. Elizabeth Farren is caught mid-turn, mid-stride, wind in the fur, smiling back over her shoulder — the impulse, in oil, at full length, by a man of about twenty-one.

And two that refuse to look, which is why they're in the list. Bastien-Lepage's Joan looks through you, focused on nothing in the room, painted in the actual instant of being called. And Vermeer's maid has her eyes shut, and is I think the most withholding painting in the building: Vermeer had put a man in the doorway and a dog in the foreground and then painted both of them out, so what's left is a woman who won't tell you what she's thinking, in a room that will no longer tell you what happened.

Tab. XIX

The self-portrait

At the end of the day I went back to the Rembrandt and sat down with the sketchbook.

The 1660 self-portrait, signed and dated, painted the summer he turned fifty-four, which was the worst year he had: bankrupt, the collection he'd spent a lifetime assembling sold off at auction, the house gone, the studio gone, the students gone.

There are two entirely different pictures on that canvas.

The shadow side is done the way the late work is always done: thin, transparent, glazed, the transitions so soft you can't locate the place where one value becomes the next. Nothing in the shadow declares itself. You could go over it with a magnifier and not find a single stroke you could put a name to.

The lights are the opposite. Loaded, physical, standing up off the surface, put down in strokes so large and unblended that from eighteen inches they stop being a face and become a small landscape of ridges with actual shadows falling across them. Every one is legible as an individual act. You can read the direction the brush travelled, roughly how much paint was on it, and in places the order.

And they are organized. The strokes in the lights aren't distributed evenly over the lit side of the head. The great majority wheel around the eye nearest the viewer — turning up over the brow, down along the side of the nose, out across the cheek and back — so the entire lit half of the face is one slow spiral with that eye at the center.

I stood there a long time working out whether I was inventing it. I don't think I was. The Met's own curators noticed something adjacent from the other end: Walter Liedtke's commentary points out a slight disparity between the two eyes, one softer and less in focus, which he reads as giving the face a distracted or anxious quality, and observes that there's more paint invested in the face than anywhere else on the canvas. He was describing the effect. I was standing there trying to work out the cause. Between us I think the account is that the near eye is the one the paint is organized around, the far eye is the one that got left alone, and the disparity people feel as anxiety is partly a disparity of handling.

I've looked at reproductions of this painting my whole adult life and it's in none of them. It can't be. The impasto is a three-dimensional fact, and what you're looking at is raking gallery light catching the raised edges of the paint — the stroke is visible because of a shadow it's casting now, in that room, on that afternoon. Photograph it and the relief flattens into value and the organization vanishes. Same as the plate mark in the 1543 folio: the information is in the physical relief, there's no way to transmit it, and you have to go and stand there.

What the spiral does is pull. It works the way spirals work — it takes your eye and walks it inward and hands it to the one place the painter wanted it, and the place is a pupil, and the pupil is looking back at you.

And in the middle of it, the flesh. I've been trying to paint flesh for years without understanding it and I understood it slightly better standing there. The lights are opaque and dense and sit on top. The half-tones are thin enough that light goes down into them and comes back out warmed, the way it does in a real hand held up against a window. Subsurface scattering, which is the term the computer graphics people use because they had to solve the problem numerically, and which is very close to the single physical property separating living tissue from meat.

I'd just spent three days with tissue that had none of it. The cadaver was opaque all the way through. Light didn't enter those specimens, it stopped on them, and that's a large part of why they read as objects: nothing was happening underneath the surface, so the surface was all there was.

Rembrandt put it back. He put back the light going into a face and coming out again.

And then the eye. It isn't a beautiful eye and he wasn't a beautiful man and he was fifty-four and ruined. What looks out of the middle of that spiral isn't defiance and isn't self-pity. It looked at me with something I can only call concern, and with understanding, and the understanding seemed to be about me rather than about him.

That is beauty.

I'd been circling it for eight days and four hundred years of books and I didn't have a definition at that moment and I'm not sure I have one now, but I know exactly where it was. Not in the anatomy. Not in the bodies. In a spiral of grey and ochre paint about the width of my palm, laid down by a broke fifty-four-year-old in Amsterdam and arranged so that a stranger four centuries downstream would be delivered to an eye that appeared to know something about him.

I made a study of it in the sketchbook. It's a poor drawing — you can't draw impasto — and what I was really doing by the end was mapping stroke direction with arrows, like a weather chart. It's the most useful page in the book.

Tab. XX

V. On Beauty, Provisionally

Han's argument, which I resisted while reading it and now can't get out from under, is that beauty requires the veil — that it lives in concealment, in resistance, in the not-yet-given. Total exposure is not the maximum of beauty but its cancellation. This is why he thinks pornography is the enemy of eros rather than its extreme: eros needs the withheld, and pornography withholds nothing.

The cadaver lab is anatomical pornography. I mean that descriptively, with no wit at all. It's the condition of total exposure. Nothing veiled, nothing withheld, everything named and fully described, and the result is not intensified beauty but the flat unmysterious presence of an object. I went in expecting the sublime and got a spec sheet. Han predicts this exactly. I just hadn't believed him until I had my hands in it.

So the veil on the body is the skin. But not only the skin — the veil is also the motion, the warmth, the drift, the shelf life, and the fact that the model is going to stand up in eleven minutes and put on a robe and look at her phone and have a life I know nothing about. The living figure conceals its interior and conceals its intention, and that double concealment is precisely what I was chasing on days one through three under the name impulse.

And you can't dissect out an impulse. This isn't a poetic claim, it's an anatomical observation: the impulse is not in there. There's no organ of intention. The thing I'm trying to draw isn't a structure of the body at all; it's a relation between the body and the next second — a leaning-toward — and a preserved specimen has no next second in it, which is exactly why it has no mystery. The mystery of a body is its future.

What you put back

Which is the question those books were arguing about, and the specimen has had something removed that turns out not to be only the skin.

Vesalius put back the tense. He stood the corpse up and gave it a gesture and a horizon and an unfinished afternoon. He falsified the specimen in the one respect that mattered, and by doing so produced the truer picture, because his subject was never the corpse — it was the person the corpse is a ruin of.

Lairesse put back the frame. He didn't pretend the body was alive. He left it on the table, flat, pinned, roped, exactly where it was, and he composed it — and the composition is a veil, because to compose is to select and to select is to withhold. The fold of skin he kept for its beauty is occluding a muscle, and that occlusion isn't a failure of the plate, it's the plate's entire method. This is why his figures are less informative and more real at the same time: information and beauty were trading against each other on the same sheet of paper, and he knew which one he was buying and what it cost. And the still life was the correct genre — not a compromise but the right formal answer. Dead nature was what was on the table, and there exists a four-hundred-year-old tradition specifically for the problem of making dead nature hold a viewer's attention. It works by arrangement and by light and by withholding.

Vanderpoel put back nothing, and didn't need to, because he never removed anything, and his book has outlasted all of them in daily working use by people who actually draw.

And then Rembrandt, who isn't in that argument because he's the answer to it. Rembrandt put back the gaze.

A specimen doesn't look back. That's the entirety of its availability: you can examine it as long as you like, from any angle, in any light, and it will never once examine you. There's no reciprocity anywhere in the transaction. A thing that's completely available is a thing with no interior, and a thing with no interior cannot regard you.

The eye at the center of that spiral regards you. Which means the painting has an inside I can't get at and never will, and that inaccessible inside is the veil, restored at the only level where it can survive being known about.

And the paint is completely unveiled. Nothing whatsoever is hidden about how that face was made. The strokes are enormous and unblended and fully legible; you can read the direction, the load, the sequence, the entire record of the hand that made them. It's the most exposed surface in the museum. And out of the most candid material available, at a range of eighteen inches, he produces the least available person in the building.

So the concealment isn't in the material. I'd assumed for years that the painter's job was to hide the machinery — to blend, to soften, to sink the anatomy under the skin and the brushwork under the illusion. It isn't. The machinery can lie right there on the surface in ridges you could catch a fingernail on. The veil isn't made of paint. It's made of what the paint is organized to do.

Which leaves the anatomist-painter in a strange position, and it's the thing I flew across the country for without knowing it: you have to learn what's under the veil in order to draw the veil convincingly, and then you have to forget it. Not discard it — metabolize it, sink it far enough down that it stops presenting itself as knowledge and starts operating as instinct. The muscles have to become the reason a form turns the way it turns rather than a set of shapes I'm inserting. Anatomical knowledge deployed visibly on a canvas is a failure state. It's the écorché showing through, and it looks like exactly what it is: a man demonstrating what he learned.

The box is wrong and gets me closer. The Pantone chip is wrong and gets me closer. The straight line is wrong — nothing in a body is straight — and gets me closer than the honest curve. The flayed man standing up in the Euganean hills is wrong in every particular and is the only one of them that looks like a human being. The fly in the middle of the Bidloo plate is a compositional accent, put there by a painter who needed a dark note in that passage, and it's also the truest thing in the book. And a slab of ochre standing a millimeter off a canvas is not a cheekbone, has never been a cheekbone, and is the closest thing to a living human being I encountered in nine days in New York — including the days I spent among living human beings.

One sheet of grey paper. A dead eraser. A few thin drawings from the last afternoon of the first workshop. A Pantone deck with fifteen holes in it. A page of arrows. And a painting to make.