History of Art in the Digital Age: Problems and
This paper 
aims to provide a broad overview on the impact of computers on the
study of the history of art. It begins by considering the nature
of the information technology revolution, exploring the often-made
analogy between it and the 'Gutenberg' revolution brought about
by the development of the printing press. Like Gutenberg, the IT
development is technologically driven. However it is driven to a
different end, one that emphasizes flexibility as well as dissemination.
This flexibility can be a two edged sword. While it enables many
new possibilities, it also seems to encourage a more fragmentary
and iterative approach to study; to the preference of information
over knowledge. It remains, however, something of an open question
whether this new approach is a necessary consequence of the structure
of the new technology being made available or whether it is more
a product of that wider intellectual change that has grown with
the emergence of Post-modernist discourses. I would argue that the
latter is the case, and that the fragmentary tendencies that can
be accommodated by the new technology can also be countered by those
who wish to do so. The computer has developed in the way it has
as a result of consumer demand. It is up to those who wish to make
different demands to feed these into the technological processes
as they are expanded and modified.
The paper also looks more specifically at issues that particularly affect the
study of images, considering both the potential provided by the digital image
for new forms of exploration and analysis, and the new opportunities that are
emerging via the World Wide Web.
There can be little doubt in anyone's mind now that we are in
the midst of one of the most dramatic technological transformations in the
history of man. Since the establishment of the World Wide Web in the early
1990s, this revolution has affected - both positively and negatively - every
society in the world. It has opened up a rich and exciting range of
opportunities in the visual arts, as elsewhere - ones that seem to be infinite
in their permutations. For many of us the World Wide Web and what it provides
are still simply too good to be true. Hardly a day passes without me staring in
wonder and disbelief with what I have just brought up on my screen, as I hear
outside my window that now all too familiar sound; the beating wings of pigs as
they fly by.
In times of dramatic change it is normal enough - after the initial
shock - to try and stop and take stock of what is going on. Such
surveys cannot, of course, be in any sense definitive, but they
can perhaps help us to collect our thoughts and reach firmer decisions
about what steps to take next. Having been involved in IT and the
Arts in one way or another for more than twenty years I am probably
better qualified now for looking backwards than forward. However,
I hope that my current paper will end up by being more than a relation
of what has happened. I have been involved in a number of projects
myself, involving both visual and textual analysis, teaching initiatives
and museum and archive projects. But it is not my intention to give
an account here of these. Rather I wish to look more broadly at
current practices, and to make some observations, as a user of the
rich resources that are now on offer, of the effects that they are
having on my own expertise - the study of the history of art.
Before going on to consider the ways in which IT is affecting the
study, preservation and promotion of art - I would like to step
back a little further to take in the nature of the IT revolution
itself. In describing this, one previous upheaval is frequently
invoked by commentators. This is the 'Gutenberg Revolution', the
establishment in the fifteenth century of the printing press as
a means for the mass reproduction of texts and images .
This technological advance enabled a new capacity in communication
that proved critical for widespread material and intellectual change.
We can see well enough that the IT revolution has brought about
an unprecedented access to an interpretation of information. But
does this change go so far as to constitute a new mode of thought?
Occurring as it has at a time of rapid intellectual change - the
change summed up in cultural studies by the term 'post-modernism'
- it seems to involve in its own nature that challenge to existing
hierarchies that has been at the basis of revolutions in thought
- such as that caused, for example, by the 'Copernican Revolution'
of the sixteenth century when it was first definitively established
that the earth revolved around the Sun. This view of the IT explosion
as symptomatic of radical intellectual change is certainly supported
by the French Cultural analyst, Jean-François Lyotard. In
The Post-Modern Condition (1979) 
Lyotard famously sees the IT revolution as an aspect of the change
in 'narrative knowledge' that has emerged in the new technological
This challenge is certainly evident in changes in our
perception of both history and of art. It was as long ago as 1979 that the
French artist and philosopher Hervé Fischer proclaimed in a performance in the
Pompidou centre in Paris that the history of art was dead. Fischer claimed that
the 'linear' concept of historical progression was now over, a change that
affected both our understanding of time and of activities like art that were
dependent on it. Now, he claimed, art like history was dead and we were in the
age of 'meta-art'.
The proclamation of the death of art has been a familiar
avant-garde strategy since at least the early twentieth century. To link this
with the death of history, however, was something novel, and reflects the doubts
about linear progress that were soon to grow into a crescendo. Ultimately these
seemed to be justified by the dramatic political changes around 1990 that
brought about the collapse of the communist bloc and the replacement of the
dialectical interchange of the cold war with more mediated forms of discourse.
The 'death of history' has now become a commonplace statement amongst cultural
analysts, suggesting we are now in a world in which events no longer unfold in a
monumental and predictable fashion, and in which none of the old values can be
taken for granted.
We can see the impact of this in historical studies generally.
'Classic' studies, in which pride of place was given to 'objective' evidence and
to 'leading' areas such as politics and economics, have ceded territory to all
manner of investigation and to sometimes bewildering degrees of subjectivity.
The history of art has been one of the many branches to be affected. The concept
of history as a progression of styles orchestrated by Great Masters - has given
way to a questioning of aesthetic canons and to the very notion of artistic
development. It is significant from this point of view that the schools and
departments that teach the subject in the U.K. are now increasingly changing
their titles from 'History of Art' to 'Visual Culture' - a term that
simultaneously obviates both history and art, replacing these with a temporally
unspecific and aesthetically non-discriminatory exploration of the pictorial.
This change in academic practice is evident enough. But are
we in fact dealing with a phenomenon that has any application beyond that
rarefied world? Are we talking here of no more than an 'Ivory Tower' revolution?
We are told that history is 'dead', a victim of the new perception of time as
multi-layered and multi-dimensional. Yet events still seem to unfold in this 'post-historical'
world in a sequential manner as they did before, and to be susceptible to very
much the same kinds of description and analysis. We are told that art is 'dead'
and that now all forms of visual manifestation should be of equal interest. But
this doesn't seem to stop the public flocking to the old guardians of outmoded
aesthetic values such as the Uffizi, the Louvre and the National Gallery. In
fact they come in increasing numbers. Nor does visual culture's exposure of the
myth of the 'masterpiece' seem to have put a dent in the auction houses' habit
of selling these discredited items for countless millions.
It may turn out in time that the 'revolutions' in thought
that we have experienced are less Copernican than they might at first seem. But
perhaps we are simply too close to what is happening to understand it. What we
can do, however, is to describe the visible symptoms of change, hoping thereby
to build up in time a fuller picture of what is happening.
In the following sections of this talk I shall look at some
of the symptoms that seem to me to be most telling, particularly in relation to
the study of works of art.
I will start with a general issue - that of the nature of
the experience we gain via IT. I would argue that the new IT process foregrounds
information over knowledge. The latter is a long-term process, conceptualized
within the mind. Information is a form of short statement that can be delivered
easily by automated processes. The gathering of information becomes much easier
by these means. It remains an open question about whether this change is
actually driven by the new technical processes, or whether those processes are
themselves a symptom of a deeper cultural transformation.
We are constantly
being made aware of the increasing shortness of our attention span, and the ways
in which this seems to be related to the diversions of a consumerist society. It
would appear that we prefer, nowadays, the short reports offered in journals and
newspapers to the long distance reading required by novels and scholarly
investigations. Similarly the process of spending hours in the company of a
single image is replaced by a practice that expects the stimulus of continual
visual transformation. Reading of books on screen is becoming more common - but
is still not easy for most of us. The ability of IT to fragment large works -
for example the potential offered by DVD to subdivide film narratives into
sections - offers a quite different way of approaching texts - both visual and
aural. This could lead in time to them being reduced to a mass of information,
explorable through all kinds of analytical processes but never appreciated, as
they once were, as a totality.
This practice of fragmentation extends to the digital image.
The very process involved in its construction re-presents pictorial continuity
as a series of distinct units, even when these are perceived by the spectator as
an integrated whole. It should be remembered in this context that a digital
image is not a ‘reproduction’ in the way that an analogue image is. Rather
it is a transformation of an image, a translation from a continuum to a set of
discreet units. When displayed on a screen the image is re-performed according
to a set of encoded instructions.
The physical means of display encourages a fragmentary
approach. The limited definition offered by most screens restricts quality
encouraging this process of fragmentation in the way we look at them. The 'whole'
reproduction of a work offered on the screen is usually a schematic mnemonic,
put up as a guide for the spectator. It is only the individual details that can
be provided in anything approaching their actual quality. Such processes can
perform brilliantly for certain types of technical analysis - for example those
required in conservation processes. But they do raise real questions when it
comes to the issue of offering a surrogate for the experience of a traditional
work of art.
As screen sizes increase, making possible larger and more
detailed visual representations, it may be that this problem will diminish.
Nevertheless, I suspect that the temptation to explore the fragment rather than
absorb the whole will persist.
Quality and the Aura
Digital imagery opens up a new point of entry into the
debate surrounding the issue of the 'Aura' of the unique work of art - that
quality famously identified by Walter Benjamin in his essay "The Work of
Art in the Age of Mechanical Reproduction" . Much attention has been
focussed on the notion that the 'aura' of a work of art is related to its 'uniqueness'.
The digital image can present a challenge to such claims in two ways. First, it
is by its very nature infinitely reproducible. Indeed it is nothing but
reproduction. There is, literally, no original of a digital image, since every
version has equal status by virtue of being absolutely identical. Variation does
occur in practice, but only at the point where the image is performed, as the
performance is dependent upon the character of the apparatus displaying it. Even
here, however, there is no sequential hierarchy. Each performance has an equal
relationship to the code on which it is based.
The ‘quality’ of the performance is entirely dependent upon the
apparatus used for display - just as the ‘quality’ of a piece of
music performed is dependent upon the skill of the musicians performing
it. The second challenge is also dependent on this performative
nature. The digital image is not a 'passive' reproduction in the
way that photographic copying is. It can therefore be used for interpretation,
fragmentation and analysis as well as for reproduction. Yet there
are questions about how much these implications can as yet be fully
accepted. It seems significant that while contemporary artists incorporate
the digital into their work, the production of pure digital art
remains a minority activity. I suspect that it is the very lack
of uniqueness that hampers development here, in an art world geared
to reward individuality above all other criteria.
When looking at the opportunities offered, then one turns almost
inevitably back to the question of information. How much easier
is it now to access information! A single keyword typed in to a
search engine like Google (but let’s face it, there is actually
NO search engine like google when it comes to quality of performance)
can deliver a cornucopia within milliseconds. Yet we all know too
that such information can be highly different in quality. Ironically
it is the knowledgeable person who gains most here, since scholarly
practice familiarized him or her with the process of sifting and
critically evaluating large bodies of information. Even such seasoned
explorers, however, give up thanks for the increasing number of
sites compiling accurate well-researched material. In the visual
arts there are textual indicators as there are for other historical
studies - such as the Inventory of Artists Papers in the UK .
As yet there are far too few actual art texts available online -
something that contrasts strongly with historical and literary studies.
I have myself been involved in recent years in putting up Hogarth’s
Analysis of Beauty - which is viewable on the
Birkbeck website .
I do hope that art historians will join in making more and more
classic texts available online - particularly those that are not
readily available these days in modern editions.
But the key area for the visual arts is of course the visual archive.
Here we have seen great strides in recent years, both in collections
making their images available online - such as the
Tate Gallery in London which has a full text listing of their
holdings and the vast majority of its images .
Equally impressive are those collections put together by consortia
of museums, such as the American group Amico
While the Tate site is free, Amico makes a charge - though one that
seems to me to be a highly reasonable one. Yet this does lead to
other benefits. While the quality of reproductions on the Tate is
limited to that which is useful only as a screen display, Amico
gives you images that can be useful for more thorough exploration.
The quality of information provided, too, is far more scholarly
than that given by the Tate, which is aimed more at a general public.
There are also equally important virtual collections, such as
that of the Corpus Vitrearum
. This aims to give a comprehensive and
international inventory of stained glass windows (an art form, incidentally,
uniquely well suited for screen display because of its transparency). While the
information provided by the Tate is limited, this shows the highest scholarly
standards. Such a work will surely in time render the printed catalogue
raisonné obsolete - the more so since the online catalogue can be instantly
All this is heartening. But there does, perhaps, remain one
unsettling question in connection with such projects. This is the question of
durability. In theory the digital image has an indefinite life. The code that
creates it does not decay. However, such code is dependent for its survival and
communicability on the electronic processes that store it and perform it. How
reliable are such processes? An Egyptian hieroglyph, carved on a wall or even
inscribed on a parchment scroll remains readable to this day, thousands of years
after it was made. How long will a digital record last? When our civilization
follows the course of all previous ones and meets its end (either by catastrophe
or decay), how will it be possible for future beings to gain any kind of access
to the information that we have been storing in our idiosyncratic and highly
vulnerable machines? There may be an answer to this question, but it is unlikely
that we will be around to find out what it is.
To the opportunities for documentation can be added those of
new forms of presentation. ‘Virtual’ exhibitions mushroom. Some time these
are surrogates for the real thing. One example is the National Museum of
Uruguay, which exists as yet only as a website . In this website we are given a
virtual tour of the building that Uruguayans hope will one day be built. In the
meantime they can still make us aware of the work of their leading artists via
the website. The simulation of the museum visit on this site is perhaps
important because it helps confirm a ‘museum’ status on the works that are
looked at digitally. Elsewhere, however, the simulation of the museum visit can
be dispensed with and other issues can be stressed.
A good example of a thematic virtual exhibition is the one mounted
last year by the National Gallery London and sponsored by the BBC.
This explored representations of the weather in art by means of
showing images of pictures from collections throughout the British
Isles. This might simply be seen as an exhibition on the cheap.
But there was a further point being made. Not only did it engage
the spectator in a particular theme. It was also a means of raising
consciousness about the works of art on display in provincial museums
that are all too little visited. Visitors to the site were made
aware of the actual location of each work reproduced and encouraged
to look at it there.
The increasing access to imagery provided by the web has also
led to the growth of the teaching of history of art via the web. A splendid
example of this has been provided by Britt Kroepelien in her account of the
courses for History of Art that she has developed at the University of Bergen
for online teaching throughout Norway .
Kropelien’s success in mounting her course can lead to
envious eyes from colleagues in many countries. Her project received strong
government funding, which enabled her to deal with one of the most enduring
problems facing those in Britain wishing to teach using digital imagery. I refer,
of course, to the problem of copyright. Perhaps this problem is now being
overcome in most countries at the teaching level. Yet in Britain there is still
no security offered, and the current copyright law is vicious in its
implications. This means in practice that institutions have no affordable means
of dealing with copyright and do not on the whole want to run the risk of
infringing a law whose implications have as yet not properly been tested. The
absurdity of the situation is that many private individuals make personal use of
the huge wealth of imagery available to them via the web - or simply by the
process of photographing or scanning reproductions, while not being able to use
such material in a teaching situation.
Recently I tested this absurdity by
deciding to take two routes to gain a reproduction of a famous Scottish Painting,
the one of the Rev Walker Skating (familiarly known as the ‘Skating Minister’)
by Raeburn. Knowing it was housed in the National Gallery of Scotland, I went to
that institution’s website. Only the tiniest of thumbnails of the picture was
available, despite the fact that it has been adopted as an advertising logo by
the institution as a whole. Knowing, too, that the National Gallery of Scotland
was a member of SCRAN, the consortium of Scottish Museums  that make information
available at a charge for teachers, I visited that site where, after searching
through their database structure - a process that took some minutes, I finally
came across the picture I wanted.
That was the ‘correct’ thing to do - but
even that was only possible to me because my institution happened to have signed
up to SCRAN and was paying the consortium an annual subscription. However while
pursuing this virtuous path, I was all too aware of the temptation to fall into
vice and use the alternative - namely an image search on the search engine
Google. And in fact, an image search on Google, using only the keywords ‘Raeburn
skating’ instantaneously provided me with examples of not just one, by over
forty, reproductions of the skating vicar, many of which were equal in quality
to the image being offered me via SCRAN. In view of such circumstances, it is no
wonder that image bootlegging is the order of the day. But however satisfactory
such illicit image usage is for the private individual, it still begs the
question of when a fair system for the public use of images in teaching and
research will come into being.
So far I have been looking at the acquisition of information
via IT. But what about the possibilities of analysis and interpretation? This
still remains one of the most disputed areas in IT. Even nowadays there are
differing voices about the possibilities offered by artificial intelligence,
expert systems and similar forms of analysis. In the visual arts this issue
focusses upon the problem of how far an image can be analyzed by computational
means. We all know that there are specific forms that can be described and
identified, and that such search possibilities have been widely used in
scientific analysis, for example to identify different kinds of cells in
clinical analyses which can be of great value in medical diagnoses or in the
codification of the forms that will allow the automated processing of
fingerprints or DNA samples. When we come to forms as complex as the visual
image, however, it seems as though the complexity multiplies beyond the possible.
There are, it is true, some areas of design where specific formal
characteristics can be identified. Computers have been used, for example, to
identify certain types of furniture or spoons or drinking vessels. In all cases
where the results have been effective there have been forms of sufficient
rigidity and regularity to enable codified matching to take place. There are
indeed forms within pictures that have such regularity. Face recognition - which has been used and which functions on the fact that the human face has
sufficient regularity and predictability in its forms to enable identification -
could be applied in pictures. Yet the problems of setting this up would be
too complex probably to justify the returns.
Another approach is to take existing forms of image analysis
and attempt to apply them to computers. As elsewhere this process is a testing
one that leads often to the clear demonstration of differences of forms.
Attempts have been made, for example, to codify the iconological system that was
constructed by Panofsky for classifying the differing levels of meaning in an
image. Yet as far as I know, this imaginative structure had proved too complex
and elusive to ever lead to a systematic application. Panofsky’s imaginative
construct was, like so many structural systems created for the purposes of
cultural analysis, more of a conceptual than actual model, and as such it is not
really susceptible to mechanical application.
A different situation is the one presented by literal iconographical
systems. The most notable success here has been the use of the codification
Unlike Panofsky’s iconological model, iconclass does not seek to
organize layers of meaning. Rather it seeks more simply to assign
a specific code for each element of meaning within a work. The directness
of this approach - as well as the hierarchical order in which such
elements are ranged - makes the assignment process achievable -
although this has to be done via the input of specialists rather
than by automated means. Nevertheless, the resultant outcome is
a structure of information that can be entered into a database and
that can lead to all the rewards that database structuring allows.
systems are now regularly being used where collections of iconographically
orientated imagery exist. Yet while straightforward in one sense, Iconclass and
similar classification systems also throw up fascinating complexities. It should
be stressed that Iconclass does not provide an unique identifier for a specific
picture. What it does is classify a specific subject visible within a picture.
The same picture can in fact bear an almost inexhaustible number of Iconclass
code identifiers, depending on the interests of the classifier. To quote a
specific example, Reynolds’ portrait of Kitty Fisher as Cleopatra could have
one iconclass identifier as a portrait and another identifier as a
representation of a subject from classical antiquity . It could also be
identified with reference to costume or jewellery. Subject and image have an
independence of each other, something that would make little sense in terms of a
Panofskian exploration of specific meaning. Yet nobody could deny the practical
use of Iconclass as an encodable system of identification - albeit one that
aids that fragmentation of knowledge into information that I mentioned earlier on.
It may also be possible to use a formal codification of
pictures to analyze and relate compositional forms. As yet work in this area
remains unproved, though the evidence provided by the practical application of
processes that encode the structure of the image via digitization are
encouraging. Recently the University of Northumbria have been applying IBM’s
QBIC system to the collection of the Guildhall in London, with encouraging
results . Earlier I myself was involved in a system used in connection with the
Van Eyck project, a project that unfortunately has not reached fruition. Yet I
feel enough was done there, too, to show how effective simple form matching can
be . The mistake
- in my view - with those criticizing form analysis - has
been to expect it to answer highly specific cultural related questions rather
than to see it as the kind of visual equivalent of word searching. Once people
have got over the fact that high cultural searching of image via the computer is
unlikely it may be possible to make the kind of progress with simple form
searching that has been achieved so spectacularly with word searching already.
This thought brings me to the final point that I wish to make
about the computer. This is that it has only developed in the way that it has
through the significant input of the user. The computer was invented by
scientists and in its early days it looked as though it would remain the machine
controlled by men in white coats, controlling all with their arcane knowledge.
But this was never the vision of Turing, the British inventor of the computer.
He always saw it as the ‘universal transformation’ engine. The
transformations that it can achieve depend on what is asked of it. The whole
history of IT has in fact been that of a tug of war between the scientist
specialist and the amateur enthusiast. Or perhaps more importantly, between
science and commerce. The original creators of computer systems were happy
enough to speak to their machines in the highly complex machine code. It was
business demand that turned it into the engine that could be manipulated by all
in the office, and then in the home.
The internet was invented by a British
physicist seeking a way of communicating more complex material to his colleagues
world wide than that which could be sent by text emails . Yet this new graphic
environment rapidly moved from being a domain run by physicists to the universal
communication system that it is today that has become absolutely central to all
forms of commercial transaction. That was never the intention of its inventor.
It was other pressures that caused this development. Those who make no demands
of computer will receive no benefit. It is up to all of us to make demands, to
press for what we think it will be possible for it to do. IT is far to important
a resource to be left in the hands of experts.
But what this history has shown is that both enthusiast and
expert are needed. What we have now is the product of their struggle, and more
so, of the creative solutions that have come out of their engagement. Without 'consumer' demand, the Personal Computer (PC) would
never have been developed - nor would the huge number of periphereral products
that accompany it. The Web arose from the demands of a user group - a consortium
of physicists who wished to send each other material using a graphics
environment. It was its unsuspected commercial potential that then caused it to
spread beyond that group.
The main message is flexibility. This, in the end, is the
difference between Gutenberg and the IT revolution. Gutenberg brought in
reproducibility - but it was inflexible reproducibility. The book, once printed,
can't be changed. It can only be refuted. The website can be updated every
second. And when it is updated the older version disappears - not like the book,
which lies in libraries still proclaiming the same old message, irrespective of
what has happened since. The book is the agent of those inflexible ideologies
that have caused such misery in mankind. It is the servant of those smug,
inflexible, predictive theories of history that hopefully have now come to an
end. Nobody would predict with confidence nowadays what the situation will be
like in even five years time from now. All that we know, is that things - both
great and small - will be different. Art will be different, and so will its