painting, and the natural inference would be, that the closer the
imitation the better the picture. But, on the other hand, a picture
which is not an exact counterpart of the object portrayed, but leaves
something to be imagined, is generally more interesting than a more
perfect copy would be.
This fact is particularly noticeable in pictures of flowers, fruit, and
still life generally.
A picture which at a little distance gives thoroughly the character of
the fish, game, or flowers it is intended to represent, will be much
more masterly and artistic if the scales of the fish, the feathers of
the birds, and the petals of the flowers are not individually studied
with microscopic care, but treated in a broad, suggestive manner.
In a painting so handled the loss of a few minute details is more than
compensated by greater freshness of color, and the charm inseparable
from a rapid and dexterous execution.
If it were possible to combine the two qualities, if we could get
breadth and brilliancy united with minute finish, it would even then be
doubtful whether the picture would be any the better for the additional
pains bestowed upon it.
In looking at pictures, we require to be deceived only up to a certain
point, and the whole question depends on where to fix that point. In
to-night’s lecture I intend to investigate this subject, and to extend
my remarks to other kindred questions connected with the finish of a
work of art.
All writers and lecturers on art are pretty well agreed that excessive
finish is undesirable. I mean such finish as one sees in Bellini’s
portrait of the Doge, where each individual hair is painted, and where
every wrinkle or pimple is studied as though it were of the utmost
importance.
There is, however, a kind of finish of an infinitely more objectionable
kind than Bellini’s.
If Bellini elaborated small details to an extensive extent, they were at
any rate thoroughly and honestly studied. His minute, delicate work
always had a laudable object, whether it were the exact rendering of a
stray hair or the microscopic modelling of the wrinkles about the eyes.
But the finisher of whom I am now speaking has no object, beyond
smoothness.
Bad proportions and gross errors in drawing are nothing to him provided
he gets a smooth, uniform surface. Like the old-fashioned provincial
drawing-master, who taught oils, water-colors, and Poonah painting,
smoothness and finish are with him synonymous terms.
Probably most of you are happily ignorant of the lost art of Poonah
painting and drawing, and I certainly do not mean to waste our time in
describing it. It will be sufficient to say that the process was almost
entirely mechanical, and that the results exhibited the maximum of
smoothness combined with the minimum of art.
It used often to be taught in young ladies’ schools. It would be both
invidious and unjust to compare the work of any Academy student with
these inane productions, but I wish to warn you, as I have often warned
you before, against confounding “finishing” with mere polishing.
Intelligent finishing consists in correcting small faults of detail, in
revising the relative values of the shades and half-tones, in giving
definite form to the fingers and toes, or any portion of the figure
which may have been neglected. Unintelligent finishing, or what I call
polishing, consists in getting a nice even grain for all the modelling
of the figure. This polishing process may not in itself be
objectionable, but it becomes objectionable when it interferes (as it
too often does) with necessary alterations and modifications.
You probably all know that I am no advocate of _sketching_ in the
schools. However much I may admire the nude studies of the great
masters, I do not wish to see the same kind of work attempted by the
students. I am decidedly for “finish,” “_high finish_” even, but by the
term _I_ mean accuracy of drawing and modelling, and not neatness or
evenness of execution.
To return to the Doge’s portrait, which I have taken as a thoroughly
good specimen of minute finishing.
It is perfectly true that if you go close up to it and examine it with a
lens, you will find it much more like nature than would be a head by
Titian or Rembrandt if subjected to the same microscopic investigation;
but pictures are not meant to be microscopic objects any more than human
beings.
If in some foreign town I meet unexpectedly my old friend Smith, I
should probably recognize him some fifty yards off. I should say: “That
must be Smith, it is so like his figure and general appearance.” As I
approach him I begin to distinguish his features, and I become more and
more certain, until finally I grasp his hand and all doubt as to his
identity vanishes. It is Smith all over, and, as I remark, not a bit
changed since I last saw him.
I do not pull out a pocket lens and count the number of gray hairs in
his whiskers, or the small warts about his eyes.
It appears to me that a life-size portrait should be treated in the same
way. Viewed from the end of the gallery, it should resemble the person
it represents, and the likeness should become more and more striking the
nearer we approach, until we get within a very short distance.
If we approach still nearer, the brush-work of the artist begins to
appear, and finally, if we examine the features with a lens, we can
discover but very little left of the resemblance which was so striking
at a reasonable distance.
The whole question is, What _is_ a reasonable distance? In portraits by
the minute finishers, the point at which the work looks its best is
evidently too near, and I think that in a good deal of modern painting
it is too far off.
A life-size head should look its best at from about six to ten feet
distance. Nearer than six feet the impasto and brush touches of the
painter would be too apparent, and beyond ten feet the delicate
modelling of details would begin to be lost.
Artists and the art-loving portion of the public delight no doubt in
going close up to a fine Titian, Rembrandt, or Vandyck, but this is to
see how the marvellously life-like effect has been produced (to learn a
lesson in short), but not to view the work from the most favorable
standpoint. I think it will be found that, generally speaking, the old
masterpieces of portraiture are best seen within the distances I have
mentioned.
There are, no doubt, exceptions. Thus the portraits of Holbein gain by
being studied closely, and those of Velasquez are best appreciated at a
considerable distance, whilst the figures of Van der Helst are so
admirably painted that they will bear a very close scrutiny as well as a
distant view.
If an artist has the precision of a Holbein or the consummate execution
of a Van der Helst, there is no harm in his following finish in
portraiture almost to its extreme limit; but if not, he had better rest
and be satisfied with less literal work.
In spite of a few honorable exceptions, the tendency of modern artists
is, however, not toward the finish of Holbein, but rather in the
opposite direction.
No one can walk through a Paris exhibition without being struck by the
enormous amount of sketchy, imperfect work; the best specimens of which
have, at a great distance, a look, a reminiscence of nature, but when
viewed nearer, resolve themselves into smears of paint, generally
plastered on with the knife.
Now it is this kind of work which is so attractive to the modern
connoisseur. The peasant, the workman, the soldier pass it by with a
laugh, or sometimes with an expression of bewilderment. The cultured
artist shrugs his shoulders, but tries to view it leniently, as he would
the work of a savage; but the _dilettanti_ and those who have a
smattering of art-knowledge delight in it. It flatters their vanity to
supplement out of their inner consciousness the artist’s short-comings.
These pictures get talked about in the salons and praised in the
newspapers, whilst good, honest, sober work is comparatively ignored.
Public taste having thus declared itself, it is not surprising that an
ever-increasing crop of these young “impressionists” should be
forthcoming to minister unto it.
There is another kind of departure from truth in connection with finish,
which is, I think, almost as much to be deprecated. I mean where the
heads are painted in a different style to the rest of the picture.
If we go back to the old masters, we shall never find this fault.
Examine any of their works. Recall to mind the Raffaelles, the Titians,
the Correggios, or the Poussins of the National Gallery, and observe
that the draperies, accessories, and backgrounds are all in keeping with
the heads. If, as in Perugino’s and Raffaelle’s early works, the
painting of the flesh is delicate and smooth, though dry and hard, you
will find the same qualities and defects in the whole picture.
If, on the other hand, as in Titian and Paul Veronese, the
flesh-painting is rich and free, the draperies will be equally so. Take
Rubens, again; how homogeneous is _his_ work! Let us suppose that a
picture by this master were unexpectedly discovered, and that by some
accident all the flesh-painting in it had been destroyed, would any one
hesitate, on inspection of what remained, in attributing it to Rubens?
Would not the good and bad qualities of the master be apparent in every
part?
As the opposite extreme to the slapdash Rubens, take the careful Gerard
Dow, and observe how the delicate and minute finish of the heads is
carried out into every detail of his pictures. If we examine any genuine
work of Rembrandt or of David Teniers, we shall always find the same
homogeneous qualities. The heads may (as is often the case in Rembrandt)
be more carefully painted than the unimportant parts of the picture; or
contrariwise, as in David Teniers, we may sometimes find a stoneware
flagon more elaborated than the hand of the boor who is holding it, but
we recognize everywhere the touch of the master.
I know of no example amongst the old masters where the kind of disparity
in style which I am deprecating is observable.
In certain modern pictures, however, this homogeneous quality in
painting is sadly wanting. In the so-called Spanish school (by which I
mean the school of Fortuny), the background, draperies, and accessories
are painted with a crisp dexterity which is quite marvellous, whilst the
heads are labored like colored photographs.
The contrast is sometimes so great that it is difficult to believe that
the picture is not the work of two artists. This fault has become
apparent in certain pictures of the Austrian school; but the contagion
does not appear to have extended to us, at least not to our
oil-painters.
I have, however, noticed a tendency amongst a few of our water-color
figure-painters toward this singular modern peculiarity.
The difficulties of giving color, form, and expression to a head, and at
the same time preserving a free style of painting, are no doubt much
greater in water-color than in oil, but I think it so desirable that a
work should be homogeneous that I would sacrifice a good deal in the way
of finish and even of expression in the faces, to obtain that quality.
If a man has great versatility with his brush and wishes to display it,
let him paint one picture in the style of Holbein or Memling, and
another in the style of Velasquez, but he should not in the same picture
(and _à fortiori_ in the same figure) attempt to unite two dissimilar
styles of painting.
One of the principal difficulties young artists have to encounter in
finishing a figure-picture is the management of their drapery.
If they are painstaking and make an intelligent use of their models they
will succeed with their heads, hands, and all their flesh-painting; or
if they do not succeed, the way to success is so obvious that I need say
but little about it.
I assume that our young artist has gone through a course of study, and
is able to paint a nude figure or a head from nature tolerably
correctly. His difficulty will be, not in copying his models, but in
making use of them without copying them.
He should form an ideal in his mind of the personage he means to
represent, and take care to select either from professional models or
from his friends those who approach nearest to this ideal. He will
probably have to make use of casts. The small heads of the warriors on
the Trajan Column are admirable in character and very suggestive. Casts
from the mediæval heads of Pisano, Donatello, and Luca della Robbia are
also very useful.
All these and other means toward his end will suggest themselves to him,
but his course is not so clear when he comes to tackle his draperies.
Every student must be aware that draperies adjusted on the lay figure
and carefully copied, have always an unnatural and trivial look about
them. The form underneath, if expressed at all, is the form of the lay
figure, and not that of nature. It will not do therefore in a picture to
adjust the drapery on a lay figure and copy the result.
This _may_ be done with advantage for those folds which hang altogether
independent of the figure, but for all those which are in the slightest
degree connected with the form underneath, some other method must be
adopted.
No doubt the best method of all (were it possible) would be to dress up
the living model and paint direct on to the picture, but this is seldom
practicable.
Long before the artist has had time to study the folds, the model moves,
and all has to be done over again.
If an artist has great experience with drapery, and the attitude is a
very easy one, he may make a charcoal study which will serve him for the
picture without having subsequently to readjust his drapery on the lay
figure, but no young hand would be able to do this in a satisfactory
way. He must go more systematically to work. He must first get a
characteristic study of the nude figure. I mean such a study as the old
masters used to make, giving the exact attitude and the form of the
salient parts. He must then make a replica in charcoal of this study and
adjust the drapery on his living model. On this replica he will now, as
far as he can, reproduce the arrangement of the folds he has before him.
There are plenty of studies by Raffaelle and the old masters which
explain better than words can, the process I am trying to describe.
He has now two working drawings to guide him, viz., his original nude
study, and the study from the draped model. Having thus as it were laid
his foundations, he may drape his lay figure and paint direct from it on
to his picture, taking care (as he proceeds) to correct the form from
his preliminary studies. He will thus be doing sound, honest work, and,
even if dissatisfied with his finished drapery, he has always his
studies to fall back upon.
Some artists, especially French and Italians, make a great use of
photography, and, if kept within bounds, I see no objection whatever to
the practice. It would hardly be legitimate art to dress up and pose a
number of models and have them photographed with the intention of
transferring the group to canvas, but it is perfectly allowable to call
in the aid of photography for draperies or costumes, where, from the
action of the figures, it would be impracticable to draw the folds from
nature.
All portrait-painters know that it is not easy to get ladies and
gentlemen to sit for their clothes, and it is far better to get help
from a good photograph than from a model or a lay figure whom the
clothes do not fit. I have no doubt that if photography had been known
in the time of Raffaelle he would have largely availed himself of it. He
often copied whole figures from his predecessors, and this is certainly
more reprehensible.
I now approach the vexed question as to how far the draperies,
background, and accessory parts of a picture should be finished without
detracting from the heads.
Most of you will doubtless recollect the passage in Sir Joshua’s
discourses where, speaking of drapery, he says: “It is the inferior
style that marks the variety of stuffs. With the historical painter the
clothing is neither woollen, nor linen, nor silk, nor satin nor velvet:
it is drapery, nothing more.”
I would fain believe that Sir Joshua meant to say that it was beneath
the dignity of high art to trouble itself with surface texture, in which
case I should certainly agree with him; but I am afraid this is hardly
what he _did_ mean. However that may be, it is certainly not the
inferior style which by intelligent arrangement and careful study of the
folds expresses the nature and quality of the various stuffs. How is it
possible (using Sir Joshua’s own words) to “dispose the drapery with the
nicest judgment, and to copy it carefully,” without clearly expressing
the material out of which it is made?
Satin and velvet are very seldom wanted in pictures of subjects taken
from the Bible, ancient history, or mythology. The figures should be
clothed in woollen or linen stuffs, and without descending to minute
imitation of texture, the nature of these garments should be clearly
expressed.
If in Raffaelle’s frescoes, and in the works of the Roman school
generally, we are in doubt as to whether the draperies are meant for
wool, linen, or silk, it is because their folds were _not_ “studied
with the greatest care,” and often _not_ “disposed with the nicest
judgment.” In many of Raffaelle’s works, and particularly in those of
Giulio Romano, we feel that the draperies are wholly imaginary, and
hence the vague uncertainty as to the material.
This uncertainty, instead of being a quality to be imitated, appears to
me as a blemish to be avoided.
In the highest style of landscape-painting, again, although it would
doubtless be absurd for the artist to elaborate his foliage leaf by
leaf, yet there would be nothing beneath the dignity of his art in
faithfully giving the general characteristics of the oak, the beech, the
ash, the bay, and the olive, so that each species should be distinctly
recognized in the picture.
I am quite aware that in many classical landscapes by Poussin and the
old masters, it is difficult to specify the kind of trees they contain,
but the botanical uncertainty in which we are left, instead of enhancing
the merit of the work, rather lessens it.
I remember going through an Italian gallery with a mixed company, and
coming upon a magnificent Titianesque landscape.
This arrested the attention of all the party, and was greatly admired,
until some botanical Philistine asked what kind of trees the artist had
meant to represent. We none of us could tell; I thought they were
evergreen oaks, another said they were elms, a third apple-trees, and so
on; but we were all in doubt.
“Well,” says the questioner, “it cannot be much of a picture if the
trees are done so badly that no one can tell what they are.”
Our Philistine was no doubt wrong, but, at the same time, the work would
have been all the better, and would have lost none of its imposing
grandeur if the specific characters of the trees had been given with
greater care.
I am glad to note that almost all modern landscape-painters are fully
alive to the fact that a tree is not merely a tree, but a particular
species of tree, and that the species can be thoroughly indicated
without in any way lessening the grand character of the work.
To return to draperies and costumes.
The artists of the Byzantine and Romanesque periods used to paint their
heads of a conventional and very ugly type, without any attempt at
individuality, and bestow all their care on the draperies, nimbi, and
accessory parts, often enriching their work with real jewels.
This fashion, which was rampant in the Byzantine period, began to wane
in the fourteenth century, but lingered on almost till Leonardo da
Vinci’s time.
During what may be called the golden age of art (that is, from
Leonardo’s time down to Poussin’s) the proper balance of finish between
flesh and drapery seems to have been well observed, but in the last
century (especially in this country) the artists of the time reversed
the practice of the old Byzantine painters; that is, they painted the
heads of the sitters as well as they could, and left the dress and
accessories to be put in by their assistants.
Bad as the flesh-painting was, the treatment of the dress was still more
slovenly and inartistic. The apologists for this style of work say that
the head is everything in a portrait, and that no one cares about the
dress and background, but this was certainly not the opinion of the old
masters. To take a familiar example. Is not the head of Gavartius
greatly improved by the exquisitely-painted frill which surrounds it?
Or, again, is not the life-like flesh in Bordoni’s female portrait
rendered still more life-like by the gorgeous color and masterly
execution of the crimson dress?
Our National Gallery teems with examples of the same kind, where
judicious finish of the accessory parts assists rather than mars the
effect of the flesh-painting.
I do not wish to be understood as insisting that in all cases the dress
and background should be as much finished as the heads, but there is a
great difference between unfinished work and bad work, and it is this
difference which the advocates for neglecting accessories seem unable to
understand. I do not find fault with a certain charming unfinished
portrait group by Rubens in the Louvre, because the accessory parts are
merely indicated, but I _should_ find fault with it if they were
clumsily and inartistically painted. The kind of work I am protesting
against is that which is often noticeable in portraits of the
Gainsborough and Lawrence schools, where the shoulders and hands are
quite shapeless, and the folds of the dress utterly impossible.
It is very refreshing to me to emerge from a gallery containing pictures
of this class, and to enter one devoted to pictures of the Dutch school.
I feel as if I had reached _terra firma_ after floundering about in a
quagmire. We never find a want of intelligent and careful drawing in the
hands and dresses of portraits by Rembrandt, Van der Heist, Franz Hals,
Terburg, and all the other masters of the school.
It may be objected that I am deprecating a fault which no longer exists,
that my expressions of antipathy to a slovenly treatment of the
accessory parts of a picture are out of date, and that the commonplace,
simpering full-lengths of fifty years ago, with their impossible
shoulders and badly-drawn hands, are no longer seen in an Academy
exhibition. I am quite willing to grant this, but it does not follow
that because this pseudo-Lawrence sort of work is no longer seen on the
walls of the Academy, that therefore it is defunct.
There is a large and ever-increasing class of young artists who are
treading in the footsteps of the old masters, who grudge no time and
spare no pains in the study of their hands, costumes, and every thing
which will give finish and completeness to their work; but, on the other
hand, there are still many who, to save themselves trouble, and perhaps
misled by the present extraordinary popularity of Gainsborough, are
satisfied with the most careless and weak treatment of all accessories
in their pictures.
That these pictures are not often seen on the Academy walls is due to
the rejecting power of the Council, and not to the non-existence of
their authors.
* * * * *
Having thus, I trust, given you to understand that by the word “finish”
I mean something quite different from mere smoothness or polish, I will
now give you a few hints as to how the work of finishing a picture is to
be accomplished.
It was the habit of Horace Vernet to make a very rough pen-and-ink
sketch for his elaborate battle-pieces. He would then, without further
preamble, have his models, and paint direct from them on to the blank
canvas, finishing every thing as he went on.
When the whole canvas was covered, the picture was finished.
I remember, on one occasion, he painted a most gorgeous Arab saddle,
holsters, stirrups, and all, and several weeks afterward painted the
horse which bore it.
Cocked-hats, kepis, etc., he would knock off by the dozen, and then,
when he could get his trooper models, he would paint the wearers. He was
always, in the matter of finish, putting the cart before the horse. I
don’t think that he did this intentionally, but he was of an impatient
nature, and could not bear to sit idle, waiting for his sitters.