Two things put photographs, among all other
forms of representation, in a class by themselves:
first, once the apparatus is in place, there is a
straightforward causal connection between object
and image, not requiring the intervention of man-
ual dexterity, interpretation, or judgment; second,
the photograph can captureeverythingabout the
object (suitably illuminated and within reach),
even details not noticed at the time of the pic-
ture’s making. This power of fixing the states of
affairs as they are at a given moment accounts
for the immense importance of photography in
contexts where truth practically matters: scientific
data, forensic evidence, personal identification,
and so on. But the proviso ‘‘in the absence of
manipulation’’ is of primary importance in these
contexts. ‘‘The camera cannot lie’’ cliche ́assumes
that the proviso is satisfied, and practically speak-
ing, most of the time, this has been a reasonable
assumption. It has rested on more basic assump-
tions about the accuracy of optics, the adequacy
of lighting, the quality of the light-sensitive emul-
sion and of the chemicals used to develop it, and
the competence of the professionals involved; but
by now the technology has been brought to a
high degree of reliability and these things can
for the most part be safely left below the level
of awareness.
This comfortable state of affairs obviously
requires qualification. ‘‘While photographs may
not lie,’’ said Lewis Hine, ‘‘liars may photograph’’
(Mitchell 1992, 30). Deception with the camera (or
in the darkroom) has always been possible, yet its
relative difficulty in years past has led to the situa-
tion where most viewers of photographs believed in
their veracity. Since the advent of digital photogra-
phy and the wide availability of programs like
Photoshop there has been a shift in popular percep-
tion of the truthfulness of photography, and impli-
cit in looking at late-twentieth century images is the
idea that they may have been totally artificially
created, never mind merely manipulated. This con-
sensus, incidentally, arises more from the popular
audience’s experience with digital filmmaking and
virtual reality games than through photography
per se. This may seem an overnight revolution in
the popular perception of photographic truth. But
granted that the manipulation of images is now
easier and more seamless than before, this percep-
tion of truth still remains a matter of degree. For it
is also true that the suppression of photographic
evidence, by manipulating either the negative or the
final photograph through retouching or overpaint-
ing of the kind practiced, for example, in group
shots of the Soviet leadership during the 1940s
and 1950s, has been familiar for a long time. It
goes back to the retouching that was done in studio
portraiture almost from the beginning of photogra-
phy, and the techniques involved had been perfect-
ed in the multiple-negative works of photographers
like Oscar Gustav Rejlander and Henry Peach
Robinson as early as the mid-nineteenth century.
The question of truth did not pose much of a
problem in the cultural context of that era, given
that the photographic image emulated the painted
image and the representational tropes were easily
decipherable: the difference between the artful and
the truthful is not too difficult to detect in senti-
mental representations of domestic life or tragic
death. Yet as photography developed its own for-
mal language specific to its intrinsic nature, the
apprehension of truth became more complex. An
essential truth about photographs is that they are
almost impossible to be seen except within a con-
text—of other photographs, of texts, of supporting
images and documentation, and of specific display
environments, and the context will affect the expec-
tation of truth. The same portrait of a baby in its
baptismal finery on the mantle of a bourgeois home
says one thing; attached to a gravestone it says
quite another.
The correspondence theory of truth works best
when the object can be directly compared with its
representation. ‘‘Snow is white’ is true,’’ said Alfred
Tarski, formulating his principle of the ‘‘material
adequacy’’ of truth, ‘‘if and only if snow is white.’’
We can validate the truth of photographs of Paris
when we go to Paris. It is normal for the represented
object (or person) to be absent when we have the
photographic representation at hand, but a few
cases of direct comparison to establish an analogy,
and some information about the provenance of the
representation (it is by a known photographer, in a
reputable travel magazine) reassures the viewer that
the image is veridical, within obvious limits of per-
spective, lighting, detail, and so on.
Confidence in the basic reliability of photo-
graphic representation, however, means that photo-
graphy has been implicitly trusted in cases where
explicit comparison has never been possible, even
when the evidence has run counter to entrenched
beliefs. The paradigm for this phenomenon is the
photographs made by nineteenth-century innova-
tor Eadweard Muybridge. An ages-old puzzle
about the galloping of horses—whether and if so
when all four feet are off the ground—was solved
photographically in 1878. In the words of Beau-
mont Newhall:
Though the photographs were hardly more than silhou-
ettes, they clearly showed that the feet of the horse werePHOTOGRAPHIC ‘‘TRUTH’’