Tuesday, September 14, 2010

Caricature and Likeness

If you want a portrait likeness, don’t paint exactly what you see.

John Singer Sargent certainly didn’t copy appearances, as this photo/painting comparison attests. He exaggerated the character of his subject, Coventry Patmore. Sargent made the neck longer, the hair fuller, the face bonier, and the mustache more “twirly.”

To be clear, the painting was done from life, not from the photo, but both date from the mid 1890s.

Patmore was so impressed with Sargent’s interpretation of him that he wrote: “He seems to me to be the greatest, not only of living English portrait painters, but of all English portrait painters; and to be thus invited to sit to him for my picture is among the most signal honours I have ever received.”
From the book John Singer Sargent by Richard Ormond, 1970, figure 21 and 22. Thanks, Barry!

ADDENDUM: Don't miss the fascinating string of comments on the subject of sight-size practice.


Daroo said...

This is one of my favorite portraits -- it has so much more animation than the photo.

My memory is that although Patmore liked it, some of his family didn't. I wonder if they were influenced by the assumed "truth" of the photo. (This side by side comparison that you present on your blog may have first occurred on the Patmore fireplace mantel)

Kessie said...

Huh! That is really interesting! I've noticed that really good modern portrait painters, like Chris Saper, tend to exaggerate like that, too. But I don't think even she does quite that much. The exaggeration makes for a much more interesting picture than just a static photo.

Rebecca.Yanovskaya. said...

Mightn't Patmore's enthusiasm be attributed to Sargent's making him quite a bit thinner...I'd prefer the painting too, heh.

अर्जुन said...

You just did the sight-size copyist schmucks super dirty… AWESOME!

Anonymous said...

Food for thought indeed ... I guess that's one of the reasons that a painting can be much more life-like than a photo. I think that we as humans (unlike cameras) don't see what we see, so to speak - we have a brain that underlines and highlights what is peculiar and characteristic about a person's appearance, and a picture that can mimic that process will be more life-like than a mere reproduction.

My Pen Name said...

i would love to see an extensive comparison of Sargent's portraits and photos of his subjects -

Madame Gaterau for example - was not very photogenic and other portraits are competent but no madame x.

another interesting comparison is photos of Carmecita (the famous dancer) film of her, (which exists) and the Chase and Sargent portraits.

we all know people who 'look great' in person but just don't photograph well- I don't know the reason for this but maybe the portrait artist captures what the photo doesn't

Mark said...

I wonder if Patmore would have been as impressed with the likeness if it had been painted by an unknown fresh from the atellier rather than the great Sargent.

Max West said...

People usually don't seem to understand that drawing is NOT photography. A drawing or a painting is a representation, not a precise duplicate of an actual subject.

That concept was made very clear by Rene Magritte in his famous painting, the Treachery of Images.

Vicki said...

Interesting also that, although the portrait looks quite different from the photo, you know in an instant that it is the same person. It means to me that what makes a person recognizable is not esactly the combination of their features, plus the spaces between them, plus the shape of their face. There is, as Cegebe also pointed to, some indefinable something about a person's appearance that gives them their "look". I have seen simple line drawings with dots for eyes that were recognizable as a particular person.
I have done portraits, and it is always easier to get a likeness when you know the person well.

slinberg said...

I saw this painting this summer in the National Portrait Gallery in London. Sargent just knocks you to your knees every time, doesn't he?

My Pen Name said...

though i hope people don't see this as an excuse for NOT learning to draw accurately- because i think you have to have the ability to draw accurately first before you change things - it would never look like CP if he didn't get the proportions right - which is even HARDER when you are actually changing what you see.

Begnaud said...

Really good find and comparison. There is a substantial difference between photography and perception. Most everyone looks thinner in real life (the camera adds ten pounds.) I think this is due to the eyes ability to focus on the nose, then the eyes, with the cheeks and ears being most distant. Most photographs place the ears on the same plane as the nose, flattening the face. A portraitist knows how to exaggerate for volumetric effect. To counter, I certainly agree that everyone likes to look thinner, and Sargent was a known flatterer. The physiognomy also looks older in the painting, and the subject may have simply gotten older and hairier.

Anonymous said...

Wow! This takes some of the pressure off. I still paint from photos quite a bit and get frustrated when they don't look like the digital copy.

Johan said...

I would say it is wise to be careful when exaggerating the facial features of women.

My girlfriend refuses to sit for me because she thinks I always make her nose too big.

Anonymous said...

I think that another important difference between a photo and a painted portrait is the influence of time. Try to find a video recording of some talking head - a TV anchor or someone like that. Put it on pause every now and then - chances are that in many cases, the person will be caught with a really awkward or downright silly expression. If we just see the recording, we don't notice all those silly faces, even though they are there for a substantial amount of the time.

The same thing applies to people we encounter in real life. When we see a face, our brain presents us with an image made up from information gathered in time - and also sorted. A photo presents a snapshot, one moment in time, and we hope that it is not one where we look too silly (a professional photographer would of course take a number of shots to choose from to get the one that really catches the model's personality), but a painting can accumulate all those little changes that will occur during a sitting, no matter how good the model is at keeping a pose, thus presenting an image that has never existed in one brief moment, but shows how we actually perceive that person when the dimension of time is allowed to have a say too.

Brian said...

Seems like Sargent brought the eyes closer together which elongated the prominent nose. The structure of the nose is also exaggerated. As a caricature artist I have learned to study how one face is different from another. Those small differences is what brings out someone's likeness.

James Gurney said...

अर्जुन: I wasn’t disrespecting sight-size methods in any way. Sargent himself was schooled in it, and as MyPenName suggests, it's very important to develop the skills of painting exactly what you see.

One of the benefits of doing that is the understanding of how accurate observational drawing differs from photography, as Max pointed out.

I think CeGeBe has expressed it well: our understanding of what someone looks like involves a synthesis of impressions in time and space, perhaps comparing the person to a "norm" or to other individuals we've met. Regarding the movement point, Sargent insisted on painting models who were talking or moving, not holding still.

Sargent's form of "caricature" is also a special kind of idealization--he obviously didn't just make noses bigger, but rather subtly enhance the individual features in their most noble form.

अर्जुन said...

""Sargent himself was schooled in it""

No he wasn't. Do you honestly believe Carolus-Duran taught such a thing? Read the excellent books written by Sargents friend and fellow classmate Frank Fowler. Have you seen a single painting/drawing/photo of a class/atelier/academy in which "sight-sizing" is being done by any member (pre-1940). Or is it a lost secret of the masters which they purposefully hid?

"A vast majority of private art schools today are teaching drawing and painting by the sight-size method. This is a copying method, resulting in complete dependency on “tracing” the subject in the same size as it intersects the picture plane. This method has been used by some artists, previously classically trained, specifically for portraiture. However, it never has been taught in any educational institution, academy or atelier until it was invented in the 1970’s by Richard Lack* as an educational method.

Any claims that the sight-size method has been used for education in academies or ateliers are without merit. The Ashland Academy of Art finds the sight-size method harmful and stands in opposition to it." ~Semyon Bilmes

*1940's by RH Ives Gammell, would be accurate, yet amounts to the same thing.

My Pen Name said...

Do you honestly believe Carolus-Duran taught such a thing?
maybe, maybe not, but École des Beaux-Arts, where sargent attended, certainly did.

To claim that no atelier used it before 1940 is just silly - by 1950 art schools were doing away with classical methods

Lastly, remember sight sizing is a way of TEACHING not 'the only' way to paint but it is the easiest way for students to easily see the difference between what they are drawing and their drawing. I would think it would be 'common sense' that having the object the same size would aid beginning students.

अर्जुन said...

""but École des Beaux-Arts, where Sargent attended, certainly did.""

No they didn't. What proof do you have (aside from that which the linear source is Gammell/Lack)?

""To claim that no atelier used it before 1940 is just silly - by 1950 art schools were doing away with classical methods""

If you can say "just silly", I can say stop drinking the kool-aid. The Art Students League of New York founded in 1875, emulated the French model of the time. The physical set up of ASL classroom has never changed. Look at any picture of the École, Art Students League, or Académie Julian, 25-50 students 3-4 rows deep, all working on canvases of uniform size depicting figures of the same size regardless of the students position.

2-Académie Julian
3-Académie Julian
These guys don't even have the canvas in their cone of vision with the model! Intérieur de l'atelier de David
Académie Julian circa 1950 The set-up hasn't changed!

अर्जुन said...

part 2

~Drawing Class taught by Allyn Cox, son of Kenyon (1896-1982), Art Students League, New York City, August 1940

~Classical Copenhagen

~current Repin Academy

~Art Students League, 1936


~Drawing after the antique.

More drawing from the antique.

~Proper technique for copying the flat.

~La leçon de dessin.

~École des Beaux-Arts


The loss wasn't method/procedure, it was culture and market.

Alas though, I know I know, the sight-sizin' was done on the sly at the private super secret backroom ateliers. (insert smiley face of choice)

Just one more.

My Pen Name said...

Your pictures do not prove the absence of a technique, anymore than the absence of a plumb line indicates that it is some 'invention' created in 1940

William Rothenstein... memoirs he wrote, "Sargent, when he painted the size of life, placed his canvas on a level with the model, walked back until canvas and sitter were equal before his eye, and was thus able to estimate the construction and values of his representation ..." -William Rothenstein, Men and Memories, Coward McCann, 1931 *

it was essential to portraiture
in fact I don't know ANYONE who doesn't use some form of sight sizing while painting - perhaps not exclusively -but as a tool in the paintbox.

to claim such a method was NOT taught in schools in a profession where portraiture was an essential part of the business is beyond credulity.

this is a silly argument and I am not going to continue to waste my time. Very well DON"T learn how to do it. call everyone a fool for using it. I don't care.

James Gurney said...

अर्जुन and My Pen: This is obviously a rich topic, perhaps meat for several future posts. Thanks for sharing those photos and first-and quotes.

My own feeling is that it's a good idea to experience a variety of different methods depending on the painting or drawing challenge you're facing.

अर्जुन said...

I agree Mr. Gurney.

My Pen Name~ I'm not arguing, just presenting facts. Placing the canvas side by side with a 6 1/2 head tall woman, and then painting her 8 or 9 heads tall is sight-size?!?! You do Sargent, and any other great portraitist, an artistic injustice.

>>some 'invention' created in 1940<<

I know it wasn't invented then, that is circa when the obsession with it began. Millais is a known ardent practitioner, and to a lesser extent John Collier. They were Englishmen. Why doesn't Speed or Solomon J. Solomon tout the practice. Speeds clear procedure for drawing from life is the same as Loomis' and Willy Poganys.

>>to claim such a method was NOT taught in schools in a profession where portraiture was an essential part of the business is beyond credulity.<<

Not to be a snob, but the fact of the day was: They taught high art, not the low art of portraiture.

19thc photo~Ecole cast hall. Looks like the cast hall at the Met.!

~Cast drawing at the ecole.

Warning for true believers, do not click this link~ Ecole drawing class.

For those that have made it this far~ Attributed to Hippolyte-Jean Flandrin (French, 1809–1864)

अर्जुन said...

Whilst traveling across internet I found this excellent article~ The Sight-size Method and its Disadvantages, by Hans-Peter Szameit of ATELIER STOCKHOLM.

I should have linked a Sargent in the last post, so: The Wyndham Sisters: Lady Elcho, Mrs. Adeane, and Mrs. Tennant.

Daroo said...

I wonder if part of the differing viewpoints here stem from matters of degree.

?????? -- You've obviously had run-ins with the sight-size fundamentalists and feel oppressed by their dogmatic ways (do they knock on doors giving away copies of "The Sight-size Tower"?)-- but you also know a lot more about their methods. I only know GENERALLY about sight size so I've always considered that anytime you have your canvas next to the model and back up so you can see them both as the same size -- you are measuring sight-size (as opposed to when you are parked next to your canvas you have to pull comparative measurements from your model then scale them to the size that is on your canvas.)

But I suspect, as the link suggests, that you don't (and the sight-size fundys don't) consider that the sight size "method".

From the Hans-Peter Szameit article:

"However, merely setting-up the canvas next to the model, even side-by-side, does not constitute working sight-size unless it also involves taking accurate measurements from a fixed viewing position using tools other than the naked eye, and carefully transferring those measurements to the canvas in order to create an image that is exactly the same size as the subject – and it is precisely these criteria that are consistently missing from any historical references."

So maybe we need a different term?
Reformed sight-size? Sorta-sizing? Wearing out the carpet?

अर्जुन said...

Daroo >>I've always considered that anytime you have your canvas next to the model and back up so you can see them both as the same size<<

Once upon a time that definition would have been fine. Only now "they" control the conversation, define the term, and are rewriting history. It has come to imply that Sargent, and every other artist they wish to name, practiced sight-size in their manner.

Daroo >>(as opposed to when you are parked next to your canvas you have to pull comparative measurements from your model then scale them to the size that is on your canvas.)<<

That drawing can also be done same-size sight-size. Your drawing will be smaller than life (the size you see it).

For Further reading: On the Art of Drawing by Robert Fawcett, Chapter 4: On Drawing Naturally.
Drawing naturally is what he calls his science-philosophy (sight-size). Observations, not mechanical measurements, are made, the drawing being done freely. Note that he does not attest to historical precedent, on the contrary, he presents it as replacement for the standard procedure (comparative mechanical measuring).

A revealing quote from Fawcett, "for mathematics is replacing observation". For you see "sight size" is the product of an anti-academic, anti classicism concept known at the "naïve eye", an aspect of pure vision impressionism. Follow the line Monet-Bunker-Paxton-Gammell. Gammell, longing for what he believed to be the lost academic standard, found himself surrounded by American Impressionist. His education and study led him to combine approachs~ Modern Sight-Size*.

Now can someone provide proof that the École des Beaux-Arts ever taught such a thing.

*my previous post does suggest Millais or Collier practiced this new version

अर्जुन said...

For scholars and aesthetes, via google books,

On Drawing and Painting by Denman Waldo Ross


The pertinent section, Drawing from Nature , pg 125~142, contains the following;

• natural way of drawing (sight size)
• Mr. John S. Sargent
• drawing method for non life-size portrait
• the method of the Schools (art)
• the fallacy of measuring
• visual feeling
• method for copying a master drawing
• tracing
• Claude Monet

2 choice quotes~

"No one who wants to learn to draw should go near any of the Art Schools. They are blind alleys that lead nowhere."

"Why in the world should we shut ourselves up for years in a School just to draw from those stupid casts and models with idea of doing something else presently? The School of Drawing and Painting and those casts and models should be given up. They have had more than a hundred years in which to prove their value and they have not proved it."

António Araújo said...

Great post, James.

Now, for the comments: the idea that this somehow shows that one doesn't need to know how to draw accurately is nonsense. One should learn it precisely in order to go beyond it. One should know how to draw realistically not just because one cares for drawing, but because one cares for reality. The camera can make copies for you, but it cannot see for you. Being able to see accurately all by yourself opens up great new pleasures in life.

On techniques, sight size is just another one. I think we should learn them all. Personally, I've tried a lot of techniques of measurement, try to use them all once in a while, and try to do without them once in a while also, and draw spontaneously, or make blind drawings, or mass drawings, or whatever. Specialization is for insects.

In the last few years I came up with a personal method involving traingulation. I spent a lot of time coming up with special grips for measuring angles, and even a method for using muscle memory. I love these angular methods and can go on and on about their advantages. But do I think for a moment that I should drop the other methods? Not on your life. One should know them all. Each has its advantages. Sight-size is actually the method I find less agreeable, but I'm happy it is available - another tool in the belt.

Less is not more, people. Knowing stuff is good. All the time we spend inventing excuses (realistic drawing is obsolete because of photos! Mental calculation is obsolete because of computers!) is time we could spend learning new stuff.

My Pen Name said...

. It has come to imply that Sargent, and every other artist they wish to name, practiced sight-size in their manner.
you hae a direct quote above. which you ignored. but...

wait! its not just site sizing that's evil.. now.. cast drawing too!

Why in the world should we shut ourselves up for years in a School just to draw from those stupid casts and models with idea of doing something else presently? The School of Drawing and Painting and those casts and models should be given up.They have had more than a hundred years in which to prove their value and they have not proved it."
probably the most ignorant thing I have seen posted in the comments section of this blog.

maybe we should smash them again? Eh, whatever you name is?

take a look here.
when you can draw with 1/10 to the ability of the average GCA student, then perhaps you lecture everyone about the evils of cast drawing.

अर्जुन said...

António Araújo said…" the idea that this somehow shows that one doesn't need to know how to draw accurately is nonsense"

Who but you said that? By all means do what you wish! Though it should be understood that a subject drawn sight-size need not necessarily be photo-copy accurate, only that it be same as the size seen.

I see that you are on the way to recreating *Charles "Shorty" Lasar's pocket angle machine… awesome!

*Charles A. Lasar, Gerome student, friend of Monet, taught plein air landscape painting mainly to women, notably Cecilia Beaux and Violet Oakley.

अर्जुन said...

My Pen Name,

The operative phrase being in their manner.

I thought it was clear, the quotes are from the text by Denman Waldo Ross, an ardent proponent of the natural way of drawing (sight size). I politely suggest that you read the 17 pages for your own edification. It offers far more than the Rothenstein quote.

"…when he painted size of life… …to estimate the construction and values of his representation ..." -William Rothenstein

"to estimate the construction and values of his representation", yes, to see the full breadth of the effect, and to be sure that it reads properly at what will be its eventual viewing distance

"size of life" does not mean sight-size. An accurate same as sight painting in no way addresses his gross distortions of form: elongated arms, legs, torsos. Sargent painted people with fashion model proportions. Some greater than 9 heads tall. While some no doubt were*, a far greater amount of 19th c. men and women of European heritage would fall between 6 ~ 7 & 1/2 heads tall, standing 5'1" ~ 6'3".

If you wish to continue slighting him with the idea that he calipered and copied the model at sight-size, by all means do so, for what else can I say?

*8 1/2 anyway

RE: casts
Draw'em if you got'em! Though I prefer them in my foyer and garden.

António Araújo said...

अर्जुन, I wasn't answering anything you said. The question of the usefulness of accurate drawing was mentioned by other people above.

>Though it should be understood >that a subject drawn sight-size >need not necessarily be photo-copy >accurate

I think that is understood. The complicated stuff with levels and whatnot is optional. Sightsizing can just mean raising up your pad in the street and making your cathedral as big as the one you see in front of you. I use that when it doesn't mean moving too much from where I'd like to stand. Why waste a good thing?

As for your satements on the history of sight-size I have no opinion, I haven't studied it. I just know that Fawcett used the simple sort of sightsizing because he said so himself in a book I have, and as for Sargent I heard it from secondary sources so who knows. I'll be sure to follow your links.

>I see that you are on the way to >recreating *Charles "Shorty" >Lasar's pocket angle machine…

Nothing new under the sun, so I won't be surprised. I'd appreciate immensely a link if you have it, or a description of the contraption.

But I doubt I am recreating the machine itself, because I am not interested in using a machine at all. I am interested in using triangulation itself, without instruments. And I have no pretense of having "invented" something the greeks knew, that would be funny coming from a geometer :).

I'll have to tell you though, that a few years ago I did think that people didn't use triangulation *for* life drawing, for some reason, because it wasn't mentioned in the dozens of books I had, but even that I put to rest sometime ago (as I do mention in my site, I found in a 1971 book and I am sure you'll have some 19th century reference at hand though I don't). Still it remains a fact that the process is unknown by almost everyone taking art classes, or even teaching (none of my teachers knew it, maybe I was just unlucky...). If you doubt it go on the internet and count how many sites teach you proportional measurements or sightsize vs triangulation, and in how much relative detail. And I can't see why. Personally, nobody taught me and I have to invent the whole thing all over, including the grips to use, how to test it, work out the geometry to see if there are distortions, etc. I learned later that some people do teach it, but nobody around my area, and I don't know how much my way of doing it differs in the details. You seem to be pretty much informed so any links would be appreciated.

As for machines: I think one just needs a couple of pencils to train the brain and gain confidence at first, and then you quickly learn to calibrate your mind. Nowadays whenever I draw 10 minute or 5 minute poses I just spot the angles of the limbs and spine and triangulate the joints and get mostly everything located in seconds and then I can happily blind-draw the rest if I want. I can mostly do the triangulation with my eyes now, just raising the pad to check when in doubt with the simplest of grips and a single pencil. It's very fast, much faster and more accurate than taking measurements. You can sketch in the air "over the model" and muscle memory will tell you what angle it is and you can even transport the angle to a pad on your lap. Much more practical than any machine and no imaginary gridding to slow you down. Even better is to just eyeball the whole thing and use triangulation to check and correct, never to measure then draw. It's the very opposite of the old adage "measure twice draw once"; if you want to build a measuring circuit in your brain you'd rather guess once and correct twice.

>*Charles A. Lasar, Gerome stude, >bu I nt, friend of Monet, taught >plein air landscape painting >mainly to women, notably Cecilia >Beaux and Violet Oakley.

Well, there's a job I'd like to have. :)

Jeremy said...

Erm, going back to the original point of this post. I believe Sargent did indeed try to slightly caracature his subjects as did many famous portrait painters (Ingres included). However, I believe the mismatch between photo and portrait is mainly down to the camera position and lens, plus the heavier coat etc compared to sargents 'further out' position. This is a distortion that profesional photographers usually avoid.


R. A. Davies said...

If painting from life, the subject is always in front of you. If you cannot see the subject, then you are pulling something from your nether regions. Using any form of measurement is essentially sight size even if you do not hold up your brush to compare lengths. We measure. We have to. Otherwise you would have a jumble of marks randomly placed. I find this "argument" pointless, tedious, and bordering on trolling. I suppose it is related to realism vs. all else. There are many "artists" out there who cannot draw and who rely upon artist statements to explain their messes. I suppose they might be jealous. What I have found is that a good painting is one which is composed well, uses color effectively, and allows the surface textures to speak as necessary. I've often noted that most corporations prefer abstract art because they are completely inoffensive. They have zero cultural content and can't offend those who wander through their offices, plus they have the imprimatur of the art critic/theorist. So the owner, or the followers/synchophants, have that certain cachet of being "in the know" and thus conveying a sense of superiority. I'm not saying that realism doesn't have its own tedious works; I'm saying that at least they automatically have some cultural content.

Unknown said...

There is a wonderful research paper on Sight-Size method: "The Sight-size Method, a Critical Overview"
By Semyon Bilmes


James Gurney said...

Johnny, thanks for that link to the essay about the sight size method. You have brought us a lot of food for thought.

Unknown said...

A very secretive man who embraced photography wholely then gave it a SECRET SHIFT

Unknown said...

A very secretive man who embraced photography wholely it's so obvious now , when you see it. Good luck to the man.

Keith Patton said...

I came to this from the link in the Sargent and Caricature post. Wow, quite the discussion about sight size, including a long article. I don't practice sight-size as taught by Lack and the like. But the detractors seem to neverify understand what sight-size is, and usually attack straw-men. The article kept referring to it as "sight size tracing," when in reality, nobody traces in atelits that teach this method. After setting up the straw-man, they then pick and choose which historical references to "refute," bit never refute the ACTUALLY position held by these practicioners. The reality is that sight-size simply means that you're painting or drawing what you see at the size you see it. That's it. It's about being able to compare at a 1:1 ratio. The historical evidence that artists of the past drew or painted the size they were seeing is easy to establish. Frederic Leighton explicitly talked about it, and explicitly said that one should draw the same size they see their subject. He then gave historical references to old masters doing the same thing. And that's well before 1940 and Lack, and coming from the old academies. Frederic Leighton's writings settle the debate.

Keith Patton said...

Whoops, lots of typo's there! I was typing on my phone just before getting off the train, sorry.