MUSEUMS, art galleries and their advisers often appeal to the authority of
science when their picture restoration policies come under attack. 鈥淭he
techniques used by conservators today,鈥� claimed Peter Fellgett, an adviser to
the National Gallery, 鈥渁re as microsurgery is to the methods of the old
barber-surgeons鈥� (The Independent, 15 April 1993). Such claims are
naive and simplistic, and must be resisted. The idea that restorations are
rigorous and reliable because they are 鈥渟cientific鈥� is, itself, a thoroughly
unscientific conceit. Applied science is only as good as the uses to which it is
put. A satisfactory application of science to conservation work must rest on a
full understanding of the rules of artistic creation by which the old masters
operated.
This understanding can only come from first-hand familiarity with the
practices used in the creation of the works. And because artistic practice is a
unique compound of technical method and aesthetic intentions, its secrets cannot
be unlocked by mere physical analysis, however advanced the apparatus: no
scientific process exists that can identify an aesthetic value. Analysis
uninformed by practical insight is not only futile; it is potentially disastrous
when applied to picture restoration.
I can cite many examples of such uninformed analysis, but I shall limit
myself to one, from my research, covering some thirty years, into Leonardo da
Vinci鈥檚 painting methods. By successfully attempting to recreate the distinctive
effects that Leonardo achieved in the Mona Lisa, I have come to realise
that when depicting flesh areas, he departed from the specific rules then laid
down for producing resistant oil paint films. In these zones, his method was
more akin to the techniques of the watercolourist than to those of a painter in
oils. It consisted of a very gradual, deliberate build-up of successive layers
of ultra-thin, overdiluted glazes. (I call this technique
鈥渕颈肠谤辞-诲颈惫颈蝉颈辞苍颈蝉尘鈥�.)
Advertisement
For this reason, the glazes in areas of paint where Leonardo was depicting
flesh are extremely thin and, in all probability, porous, friable and powdery.
Therefore, the varnish lying on these areas acts as a useful reinforcement. Any
attempt to remove and replace this integrated varnish by solvents or scalpels
could seriously endanger Leonardo鈥檚 own work. Since the last stages of artistic
鈥渇inessing鈥� form the upper layer, the most crucial effects are also the most
vulnerable ones.
In 1994, the Louvre in Paris considered cleaning Leonardo鈥檚 The Virgin
and Child with St Anne. As a consultant to the museum, I was able to point
out in advance these physical and aesthetic vulnerabilities. My testimony was
listened to and accepted by the Louvre鈥檚 committee for restoration. Happily, the
project was dropped, but other experts challenged my views鈥攊n particular
David Bull, head of painting conservation at the National Gallery of Art, in
Washington.
Although Bull agreed that the surfaces produced by Leonardo are 鈥減aper-thin鈥�
and finely manipulated, he doubted that they would be fragile structurally. He
recommended that the varnish be removed. Bull contends that Leonardo achieved
his impalpably vaporous effects not by delicate brush-work, but much more
robustly by spreading paint with his fingertips and, even, with the palm of his
hand. I reject this hypothesis. It might, at first glance, seem to rest on hard
physical evidence鈥攏amely, the traces of fingerprints found on certain of
Leonardo鈥檚 paintings. In reality, it reveals a misreading of evidence and a
misconstruing of technical practice.
I am familiar with the method of work that Bull suggests, and in my attempts
to 鈥渞econstruct鈥� Leonardo鈥檚 art I have experimented with it over several years.
It is a process which I abandoned when I realised that no matter how skilled or
practised the hand, the master鈥檚 famous sfumato effects, characterised by an
extremely gradual transition between areas of different colour or shade, could
not be so obtained. It might be argued that the failure was my own. But, having
tried to create the effects with both brush and hand, and comparing them at
length, I can justifiably claim sufficient expertise on the issue.
Those who continue to champion 鈥渇ingerwork鈥� have yet to demonstrate the
practice. My 鈥減ractical鈥� position is widely supported by historical evidence and
scientific pointers, but nowhere in Leonardo鈥檚 extensive theoretical writings on
painting does he mention anything like the method which Bull adduced.
In fact, the fingerprints are only encountered in unfinished
sections鈥攚hich is to say those that have not received the delicate
multilayered treatment described above. The 鈥渇ingerwork鈥�, self-evidently, served
as a quick method of laying down an intermediate stage, and not as a final
effect. The more fragile of these two distinct types of work is the more complex
and subtle one (that is to say, 鈥渕icro-divisionism鈥�), which means that
restoration must be equally delicate and cautious.
As chief conservator at Washington鈥檚 National Gallery, Bull has at his
disposal a huge array of scientific apparatus. My contention is that such
material can only be of value if placed at the service of understanding through
practice. No important progress is likely until this basic notion is integrated
into working procedures which today, more often than not, are in contradiction
with the high ethical claims of true scientific conservation.