matter S in a marked surface M, a theory of pictorial representation will also
have to provide a standard of correctness for this seeing. For paintings and
other hand-marked surfaces, this standard of correctness will generally be
set by the maker’s intentions; for photographs it will generally be deter-
mined by the object in front of the camera. Given Photoshopping and other
techniques of photographic image manipulation, author’s intentions may,
however, also play an important role in setting the standard of correctness
for seeing a subject matter in photographic images.) Explanatory theories of
seeing-in can be divided into roughly three kinds: phenomenological-
subjective; objective; and information-theoretic. In phenomenological or
subjective theories, the subject matter S that is seen in a marked surface is
determined by theexperienceof the viewer. Thus Robert Hopkins argues that a
viewer sees a subject matter S in a marked surface M when theexperienceof
looking at the marked surface resembles theexperienceof looking at the
actually existing object that a painting has taken as its subject. Specifically,
both in looking at the marked surface and in looking at the actually existing
object that is depicted, one must see (have the visual experience of) roughly
the sameoutline shapein order to see that actually existing objectinM as its
subject matter, where an outline shape of an existing object is the contour
that would be placed by tracing its outline on a pane of glass at a certain
distance from the object.^44 Adjustments to the theory will then have to be
made for depictions of nonexistent objects, since we are able to see a unicorn,
say, in a unicorn painting, and for deliberately distorted depictions such as
caricatures. Phenomenological theories have the advantage of making evi-
dent that a pictorial representation is generated from a specific point of view
and of emphasizing that a particular kind of visual experience does occur
when we see S depicted in M.
Objective theories are motivated in contrast by the thought that paintings
and photographs are primarilyofobjects andtheirproperties and by suspicion
of the thought that painters and photographers are in any way trying to
produce marked surfaces that are to engender psychological experiences.
Instead, painters and photographers are rendering or capturing howthings
actually look or appear. Thus John Hyman argues that resemblances that
enable us to see S in M are viewer-independent, direct resemblances between
(a) the occlusion shape or outline shape and the aperture color (roughly, the
(^44) Robert Hopkins,Picture, Image and Experience(Cambridge University Press, 1998), pp. 53–63.
Representation, imitation, and resemblance 39