All 4 books by Edward Tufte now in
paperback editions, $100 for all 4
Visual Display of Quantitative Information
Paper/printing = original clothbound books.
Only available through ET's Graphics Press:
catalog + shopping cart
Edward Tufte e-books
Immediate download to any computer:
Visual and Statistical Thinking $2
The Cognitive Style of Powerpoint $2
Seeing Around + Feynman Diagrams $2
Data Analysis for Politics and Policy $2
catalog + shopping cart
Edward Tufte one-day course,
Presenting Data and Information
Denver, July 22
Seattle, July 25, 26
Portland, July 28
Minneapolis, August 22
Milwaukee, August 24
Chicago, August 25, 26
San Francisco, December 12, 13, 14
David Hockney: Secret Knowledge
David Hockney: Secret Knowledge
A new expanded edition of David Hockney's Secret Knowledge: Rediscovering the Lost Techniques of the Old Masters was recently published, with a quote for the back cover from my review of the original edition at this website .
-- Edward Tufte
Hockney gave a talk at our departmental or college-wide seminar a few years back. It was a great story and well-told to boot. If the book is half as good as his 45 minute talk, it's worth getting.
-- Matt B (email)
It is notable that there is no mention of the many thoughtful and responsible challenges to
Hockney's optical theory in this new edition of Secret Knowledge. Indeed the book would
greatly benefit from the inclusion of proper comparisons with alternative explanations.
Consider Richard Feynman's advice: " If you make a theory...you must also put down all
the facts that disagree with it....the idea is to give all of the information to help others to
judge the value of your contribution; not just the information that leads to judgment in
one particular direction or another." Hockney need not give all the facts that disagree. It
would be enough to include the best ones.
But as critics point out, much of what Hockney sees as evidence of optics have more
straightforward explanations. To site two examples, Hans Holbein most likely used a
distorted grid method to paint his stretched skull, and Jan van Eyk almost certainly used
proportional dividers to enlarge his drawing of Cardinal Nicolo Albergati (a recent
examination of the original drawing has revealed prick marks that points to evidence of
just such a method, and distortions in the enlarged painting are also consistent with that
At least, in this new edition, Hockney admits: "You can't trace a living head. The smallest
movement of the head seems magnified amazingly in the projection, so only the key
points need be noted ....The rest is eyeballing". It should be noted here that though the
movement is magnified, the detail is not. Projections from concave mirrors are extremely
soft (see page 104 of Secret Knowledge and compare the resolution of the projection with
it's corresponding photograph). When one considers the astonishing detail of a Jan van
Eyck, an artist who painted every individual eye lash, seemingly every individual hair,
delicate veins in the eyeballs of his subjects -details which could not have been resolved in
a primitive projection- one wonders to what degree a low resolution projection would have
I've often wondered what Richard Feynman would have thought of Secret Knowledge.
Needless to say, he had a pretty good grasp of optics. He also learned to draw the human
figure from life. He was no Jan van Eyk, but he could do a little better than noting "key
points" of a soft projection.
-- wwick (email)
I have just read Secret Knowledge (2nd Edition) cover to cover and agree that it does not really describe the main possible problems with the optical hypothesis only what is right with it. Perhaps Hockney should adopt the Popperian criteria that the hypothesis should be falsifiable. I think that this would mean in this case that other methods are shown to be (experimentally side by side with Hockneys expts) equal to his optical efforts. I also miss any examples of his own oil paintings made using the optical techniques.
Nevertheless a great (visual) thought provoking book.
-- Matt Reed (email)
I've read some of the follow-up work, both favorable (Steadman and Kemp are especially
persuasive) and unfavorable (persuasive in a few particular examples) to Hockney. The
Hockney critics regularly send me reprints and preprints of their work, alas now and then
accompanied by edgy letters with vaguely ominous threats that my support of Hockney
might be thought to
my entire book, Beautiful Evidence, particularly when the Hockney critics get a
review my book. I find it insulting
that anyone would believe that such threats could possibly matter to me. I wrote what I believe is a precisely nuanced account of
Hockey's work in Beautiful Evidence.
My blurb on Hockney's book was
probably borrowed from my amazon review. My longer review posted at this site is as
This amazing, powerful, delightful, and beautiful book makes a strong case for the use of
optical projection methods by artists from about 1430 on. Hockney finds all sorts of
telltale evidence of lenses and mirrors (rather than only exquisite eye-brain-hand
coordination) in painting and drawing highly realistic flatland images of 3-space scenes.
Hockney tells an often hilarious visual detective story.
The book has such a wonderful intensity of seeing and thinking. The computer
(presumably using Adobe Photoshop) made it possible for Hockney to write with images
and to construct a beautifully designed and printed book.
This is one of my favorite works in art history, along with Meyer Schapiro's essay on the
semiotics of visual art (in Meyer Schapiro, Theory and Philosophy of Art: Style, Artist
Society), and Martin Kemp's The Science of Art. Hockney, Schapiro, and
their readers into seeing and thinking more deeply.
For a fuller account
of my views, see Beautiful Evidence, pages 28-31.
-- Edward Tufte
I think it is important to note that he is describing possible by which some artists were working, and that the pictoral influence this had on image making as a whole around all of Europe was not such that everyone was using lenses, but that optics based technology had a large influence on how people without optics began to look. This in turn is where Hollywood got its sense of lighting from in a nice ironic loop.
A favourite addition of mine to the new edition is the section on Brunellesci's 'invention' of perspective, and the tests which Hockney took at the very spot.
Sorry for the late comment on this but a friend only recently pointed this out. It is a marvelous book.
-- Luke Hart (email)
I have been thinking about the link between optical resolution and science. One of the key tools of science remains detailed visual observation. Until the late 1500's the ability of experimental scientists to detect differences visually was limited by the resolving power of the human eye. In addition, their ability to make detailed records of what they observed was limited by the artistic ability of their eye-brain-hand system. Although our image recording technology is now mainly digital, even today some specialised fields of science rely on the ability of humans to visually resolve differences and make hand drawn records.
For microscopic specimens or macro specimens with very fine detail a camera lucida attachment can be added to an ordinary light microscope. This allows you to view the field of view and trace and draw what you see. Having used one I can recommend it.
I think that we under-appreciate the power of forcing ourselves to draw, it commands us to really see what is going on and to focus on the essentials of the object under scrutiny. In fact it was precisely this aim that motivated the noted English art critic and social thinker John Ruskin (1819 - 1900) to enthusiastically teach workingmen how to draw. He believed that learning to draw would teach people to really see and understand nature and thereby be happier. He told a Royal Commission on Drawing in 1857 that, "My efforts are directed not to making a carpenter an artist, but to making him happier as a carpenter".
The camera lucida is no longer mainstream but research paleontologists such as Simon Conway-Morris (one of key scientists in the re-interpretation of the Burgess Shale fossils) still rely on camera lucida attachments on their light microscope to record the fine details of their fossils. These are great examples of what ET calls Mapped Pictures in Beautiful Evidence. For a recent example see, "New Malacostracan Crustacea from the Carboniferous (Stephanian) Lagerstatte of Montceau-les-mines, France" by Patrick R. Racheboeuf, Frederick R. Schram, and Muriel Vidal in the Journal of Paleontology 83(4):624-629. 2009. This paper includes excellent hand drawn camera lucida images of their newly discovered fossil Crustaceans.
Palaeocaris secretanaeSchram, 1984, MNHN-SOT 12595, Assise de Montceau, Saint-Louis open cast, bed 0. Camera lucida drawings from latex casts; 1-2, latex cast of the left side of the exoskeleton exhibiting the thoracic and pleomere segments; note the occurrence of possible (?) epibiontic organisms on the thoracic segments (bold arrows); the thin arrows indicate the pereion/pleon limit; 3, latex cast of the right side of the anterior part of the exoskeleton; 4, tentative partial reconstruction of the exoskeleton showing only three pleomeres with posterior spinose margin; compare with Fig. 12 of Perrier et al. (2006). Scale bar = 2 mm
-- Matt R (email)
Further to my last posting on this topic here are some recollections from Prof Harry Whittington, leader of the team who systematically re-evaluated the Burgess Shale fossils.
"The most abundant arthropod (some 15,000 specimens) in Walcott's collection is Marrella, on which I began my work by looking at, even cursorily, all the specimens, and then choosing a particular sample to illustrate my account. In order to compare the specimens, I drew them using the attachment to the microscope known as the camera lucida. This is a salutary procedure that forces one to look at a specimen carefully, and reveals the different levels at which parts of the body lie in the shale."
Whittington echoes the comments above of John Ruskin - the act of making a manual drawing 'forces one to look at a specimen carefully'. This aspect of drawing - that it enables an observer to engage in a more intense manner with the scene (either landscape in the case of Ruskin or microstructure in the case of Whittington) is an under examined part of the parallel that ET talks of - that art and science are both intense ways of seeing.
-- Matt R (email)
I have been reading up on drawing - the complex interplay between visual observation and hand coordination required to make a pencil record of what an artist or scientist sees either with the naked eye or with optical aids (lenses, mirrors, camera obscura, camera lucida).
This is clearly of direct interest in art (e.g. the Hockney-Falco thesis) but it is also of real interest in the history of scientific illustration (from the work of Galileo, through that of Audubon to the camera lucida work of Harry Whittington in the Burgess Shale) - as ET put it in Beautiful Evidence "Science and art have in common intense seeing, the wide-eyed observing that generates empirical information".
As mentioned previously with respect to Ruskin drawing forces us to really see. This is true for science as well - Prof Harry Whittington, one of the lead academics on the re-analysis of the Burgess Shale fauna, reflecting on his methods said the following on the occasion of the award of the Geological Society's Wollaston Medal to him in 2001;
"I soon realised, in my work on Burgess Shale fossils, that explanatory drawings would be needed as well as photographs, to describe these fossils. This is where Dr Wollaston enters the scene - a late 18th to early 19th Century physician, who practised in London for many years, and made valuable contributions to chemistry and optics. He had a cracked shaving mirror, but instead of throwing it away he puzzled over the refractions and reflections of light caused by the cracks.
This led to his realising that by inserting a prism into a microscope tube, the image could be directed laterally, then down on to paper beside the microscope, and provide a way to draw an accurate picture. In much refined form this is his invention, the camera lucida, which I used to make my drawings."
Our ability to make detailed records of what we observe is dictated by our artistic ability and the capacity of the human eye-brain-hand system. Drawing is an extremely complex and interesting process requiring high levels of motor co-ordination and only achieved by long practice.
To illustrate this complexity the figure below will repay careful study. It shows a montage of data that was extracted from a study by John Tchalenko and colleagues at Camberwell College of Arts. http://www.arts.ac.uk/research/drawing_cognition/.
In this study they used an eyetracker, movement sensor and close-up video to study how the painter, Humphrey Ocean, drew portraits. Their analysis concentrated on the painter's eye-hand coordination. They observed that in general his eye closely followed the drawing hand, with fixations on, or very near, the line being drawn. They also found frequent exceptions to this behaviour when the artist's eye moved from the drawing hand to look at other parts of the drawing or he turned to look at the model. The study concludes that they show evidence, `illustrating the process of visual memory fading and refreshing, and the possible action of a motor memory component in the drawing method of this painter'.
The figure to the right shows data about the fixation of the eye above the horizontal line and data about the position of the hand below it as the artist drew the picture of Nick to the left. In both cases vertical distance away from the horizontal line corresponds to physical distance away from the paper; which is the physical fulcrum of the exercise (eye, pencil and paper come together). There is an initial period of practice during which no contact is made between paper and pencil but the eye flicks back and forward from the model to tracking the pencil. Then the eye focuses on the paper and immediately fater the pencil hits the paper to begin the first four strokes H1-H4. These take 2.68 seconds and a total of 10 cm of pencil stroke is created. The eye evaluates what the pencil has done then back onto the model and so on.
The graphic is quite special - the physical paper is represented by a horizontal line and the two times series (eye data and hand data) are running in synchrony.
-- Matt R (email)
What you see whilst drawing, is what you need for drawing.
The figure below is re-drawn from Visuomotor characterization of eye movements in a drawing
task. Ruben Coen-Cagli, Paolo Coraggio, Paolo Napoletano, Odelia Schwartz, Mario Ferraro and
Giuseppe Boccignone. Vision Research, 49, Issue 8, 2009, pp. 810-818.
This is an interesting eye-tracking study that quantifies how much more you focus on an
object when you are drawing it versus just viewing it.
"We find that a peculiar feature of the drawing behavior is that the gaze does not move back
and forth among different objects, but proceeds sequentially, and most fixations on an
object are executed within a time interval in which no fixations occur on other objects."
In addition the study also found that the proportion of inter-object saccades to total
saccades (a saccade is a quick, simultaneous movement of both eyes in the same direction) is
four times higher for free viewing (c) versus drawing (d).
-- Matt R (email)