|
|
Jan J Koenderink: Colour, old age, and accepted truth |
|
|||
|
|
|||||
|
Now and then I become so much
interested in a person's work that I really want to know more. I then read all
I can dig up - the science, the reactions of contemporaries, and the bits and
pieces that might give me glimpses of personal life. The last occasion this
happened to me concerned Ostwald. It turned out to be an often surprising,
thus very entertaining, and also rewarding exercise. Wilhelm Ostwald was born in 1853.
His career in chemistry was most impressive on all counts. He received the
Nobel prize in chemistry for his work on catalysis, work he considered a
sideline, his real interest then being electrochemistry, a field largely
founded, developed, and for a long time dominated by him. Ostwald was a true
academic with very wide-ranging scientific, cultural, historical, and social
interests. He was a very outspoken character and at times antagonised people
by his stubbornness. Moreover, although clearly very adept at the social side
of being a professor, he sometimes misjudged the forces that control the
dynamics of social processes. On the whole, the number of his successes far
exceeds that of his failures, though. For instance, he was instrumental in
having the European standards for rational paper sizes accepted (the
`A4-sheet' on which I'm writing right now, for instance). A rather
spectacular - given the energy he put into it - example of a failure was his
attempt to have Esperanto accepted as the language of science - a modern (and
rational) replacement for that lingua franca of science, Latin. Ostwald published his
autobiography in 1927. The book makes fascinating reading since he describes
the laboratories and the academic circles of his time from the inside. He
knew many well-known scientists in several disciplines and one can obtain
much interesting anecdotal material from the book. For perception scientists
his descriptions of Fechner are of special interest. Fechner apparently made
quite an impression on him as he later used Fechner's law extensively in his
work on the colour atlas. I wouldn't be writing on Ostwald
if perception didn't enter the story. In Ostwald's case this happened only
after his retirement, when he put most of his (incredible) energy in
the study of colour. By the time of his death (in 1932) he had published
hundreds of papers and dozens of books on the topic. Moreover, his colour
system had been accepted as the industry norm and his `system' was taught
from the kindergarten to the university level. Boxes of crayons (rationally
ordered) were sold to kids while the monumental Ostwald Colour Atlas was sold
for industrial, academic, and artistic use. Textbooks were available at all
levels. From the 1920s to WWII the “Ostwald System'' was the de facto
standard in Germany, and in fact Continental Europe. When I now take a
standard handbook on the topic (Wyszecki and Stiles' Colour Science - I own
the 1st edition of 1967) from the shelves, I find Ostwald scarcely mentioned
at all (his system gets less than a sentence), so something apparently went
wrong after Ostwald's death. Happily, Ostwald himself knew only success. It is not that there was no
opposition to the Ostwald system in Ostwald's own times. On the contrary.
Whereas the system was very well received by the artists, others - designers,
perception scientists, and especially physicists, but also many engineers -
fulminated against it. For instance, the well-known physicist Kohlrausch
(1921) published a paper ``On the so-called Ostwald colour science''
expressly to finish Ostwald off (sciencewise, that is) and trample on his
grave. One reason why the physicists
didn't take Ostwald's theories seriously was that they didn't understand them
because the ideas deviated too much from canonical accepted truth. He was
simply too original. Ostwald relates in his entertaining book Er und Ich (He
and I), written as a dialogue between a generic scientist of the times and
Ostwald himself, that a physicist when confronted with the question ``what is
the reflection spectrum of a bright yellow paint?'' answered: ``no reflection
throughout the spectrum except for full remission in a narrow band at about
580 nm'', an answer in the true Newtonian tradition: The paint should
remit a `homogeneous light' as a matter of course. But a little reflection
reveals that such a paint would be black (or at least dark brown), because
hardly any radiation would be reflected at all! Ostwald simply put a number
of bright yellow materials in front of his spectroscope (for a chemist the
colour is in the materials, not in the eye) and found that they all
showed the same spectrum: About half of the spectrum is fully remitted, only
the short wavelengths (below about 490 nm) are absorbed. That is why
yellow paint is almost as bright as white paper. Indeed, Ostwald found that
all the best paints (bright and strongly coloured) remit about half of the
spectrum. This theory of `semichromes' became the core of his theory. This
was almost blasphemy to people who understood colours as Newton's
`homogeneous lights'; it is nowadays hard to imagine how strongly people of
Ostwald's times felt about it. The formal reason became clear only after
Schrödinger's (1920) theoretical work. The theory of semichromes is a simple
extension of standard colorimetry. Ostwald is one of my heroes,
because he proceeded intuitively with an uncanny anticipation of what might
be ahead even when most problems still lay in the dark. Perhaps his main
weakness was that he sometimes failed to finish the job properly in the more
formal sense: This is what irritated the physicists most. In retrospect
(after we have discarded some fluff of clearly misconceived notions)
Ostwald's colour system is the only fully rational (firmly founded in
colorimetry and optics) systemisation of surface colours available today.
Although Ostwald conceived of it intuitively, all major components of the
theory have found respectable formalisations, starting with Schrödinger's
monumental work in the early twenties. The systems currently in common use
(mainly the Munsell system and the CIE Lab system) are feeble, ad hoc
constructions based upon uneasy combinations of colorimetry (the objective
ingredient) and `psychophysical data', that is to say arbitrary selections
and functional fits decreed by a committee. I know many people will hate me
for pointing this out, but I'm firmly convinced. An interesting issue is whether
Ostwald can be said to have developed the theory of surface colours. Opinions
are bound to differ. From my perspective, Ostwald almost singlehandedly
developed the theory from scratch. However, his thinking was intuitive, not
formal, and often sloppy. He did make mistakes (some spectacular ones; he committed
his errors in a grand style) without noticing. People who hold that Ostwald
was a clown and other people did all the real work have a point, although I
hate to admit it. I tend to value the creation of the novel concept, out of
the blue, and less so the formally correct but essentially - in my view -
derivative work, even though I highly value mathematical elegance. Of course,
both are needed. In this respect Ostwald reminds me of James Gibson, another
interesting case good for considerable controversy. Although it would be interesting
to go into the technical details, this is clearly not the place to do so. I
have only unearthed Ostwald's case because it illustrates so many facets of
the sociology of science (and society in general) in interesting ways. I'll
simply pick out a few aspects that interest me. Although on most days I feel ready
to beat the world, the newspapers tell me that, out of all people of my age
and over, only 30% are currently part of the labour force of my country (The
Netherlands), and this despite the fact that officially people are fired on
their 65th birthday (it is called retirement, and one has to prepare oneself
for congratulations). Instead of making people appalled, it is generally
considered to be the desirable state of affairs, since it is commonly
accepted that the `elderly' (mostly people over 55, increasingly over 50
even) are unproductive (there is talk of lowering wages for the same job as
age progresses) and most certainly no longer creative. In the sciences people
are hardly expected to add much of value after the age of 30 even, especially
in fields of `hard science' such as physics. The more my own age progresses,
the more I cherish examples such as Wilhelm Ostwald, who after his retirement
entered a new field, developed it on his own steam, and published more on it
than many scientists will ever manage in their chosen field in a lifetime.
Not that such examples impress our national science foundation or ministry of
education. Indeed, far from it. Why was Ostwald's Farbenlehre
(Colour Science) so soon forgotten after WWII, after several decades of a
most spectacular flowering? No doubt the war itself has been a major factor
here. Continental Europe (and particularly Germany) looked on the USA as the
paradigm for scientific endeavour after the war. Consider experimental
psychology of perception: It is mainly Gestalt psychology injected by leading
(mainly German) European psychologists into the USA that made a lasting
impression on the rebound. The ones that stayed in Europe are not nearly as
well known as the ones who left. Another factor that is probably even more
important (it might causally explain many of the phenomena for which I
mentioned WWII as a possible influence) is the issue of language. The two
factors are not at all independent though, since before the war German (like
French) was a highly respected scientific language in continental Europe. For
instance, in my country many theses were written in German or French (with a
few Latin ones thrown in), whereas nowadays the overwhelming majority is
written in English. This might explain why Ostwald remained unknown in the
USA. Even an eminent physicist like Erwin Schrödinger, who originated the
pertinent formalism of surface colours very elegantly, was largely ignored.
His work on colour is conventionally attributed to MacAdam in the literature
written in English (a fact which can't be blamed on MacAdam, by the way).
Many other examples could be quoted here. No doubt most often these - in my
view painful enough - developments took place unwittingly through people with
the best of intentions. However, sometimes these things are more worrisome
and seem to border on the type of `rewriting' of history we remember from the
late Soviet Union (remember Ronald Reagan's `Empire of Evil'?) and the fate
of Ostwald's colour system appears (to me) a case in point. Ostwald visited
the USA in 1905 and on that occasion met Munsell (1858 - 1918, an artist) and
the two compared notes (much later - in 1913 and 1914 - Munsell also lectured
in Europe). Given the time needed to produce a colour atlas the Munsell and
Ostwald atlases appeared almost simultaneously (Munsell in 1915, Ostwald in
1917). I found rather overt suggestions in several books on colour science
that Ostwald (after the 1905 meeting) emulated Munsell's system and imported
it into Europe under his own name. To one who understands both systems
this sounds truly ludicrous! But then, the (mistaken) way the Ostwald system
is typically described in the English literature (its mention alone is a rare
enough event) is as a parody of the Munsell system, so most readers will be
fully prepared to accept such allegations. The fact of the matter is that
Ostwald's colour system is squarely based on colorimetry (that is to say it
is fully objective), whereas the (original) Munsell system is based on eye
measures (which goes to say that it is subjective). The difference is indeed
a categorical one. These are two possibilities between which one has to
make a choice. One actually needs both, but taken one at a time. The modern
`fusing' of these complementary systems is in bad taste and to my mind bad
science (or, rather, not science at all; perhaps it can be said to be an
`engineering solution' if you want). Because so many scientists are not able
to read the original sources, such myths will slowly but certainly evolve
into truth as people know it. Of course `truth' as a causal factor is what
people believe. If there is an absolute kind - but unknown - it must be
irrelevant. In retrospect it is perhaps to be deplored that Ostwald's efforts
to promote Esperanto as the language of science didn't work out the way he
imagined. Reading up on Ostwald and delving
into his work proved to be rewarding. I like to remember Ostwald as his
children describe him, after his retirement and working like mad on his
monumental colour atlas: his beard unkempt and shining in all the colours of
the rainbow because of the numerous jars with pigments all over his lab. |
|||||
|
Jan J Koenderink (Jan Koenderink was born in 1943. He studied physics, mathematics, and
astronomy at Utrecht University. His thesis (1972, summa cum laude) is on
neural networks, the structure of color space and psychophysics of spatial
vision. He joined the faculty of psychology of Groningen University where he
worked on the theory of optical flow and shape from shading. In the late 70's
he joined the faculty of physics and astronomy at Utrecht University where he
now holds a chair, "The Physics of Man". In 1987 he received an
honorary doctorate in medicine from Leuven University (Belgium), and in 1990
he was appointed by the Queen a member of the Dutch Royal Society of Arts and
Sciences. His current interests include the mathematics and psychophysics of
space and form in vision and active touch, the structure of perceptual space,
and ecological physics, including applications in art and design. He is
currently on the boards of a dozen scientific journals in various fields. |
|||||
|
References Kohlrausch K. W.F., 1921 „Bemerkungen zur Ostwald'schen sogenannten Farbentheorie'' Physikalische Zeitschrift 22, 402 Ostwald, Wilhelm, 1927 Lebenslinien, eine Selbstbiographie (Berlin: Klasing) Ostwald, Wilhelm, 1936 Er und Ich (Leipzig: Theodor Martins Textilverlag) Schrödinger, E., 1920 „Farbenhalb Theorie der Pigmente von grösster Leuchtkraft'' Annalen der Physik 62 603 Wyszecki,
G., Stiles, W. S., 1967 Color Science (New York: John Wiley) |
|||||
|
|
|
|
|||
|
|
|
|
|||
|
|
|
||||
|
|
|
|
|||
|
|
|
||||
