The nature of cognitive artefacts 133
mode of reference. Both the indexical and the symbolic mode of reference may
become part of an iconic picture, too (think of a photograph of a crowd at a rock
concert, furnished with an arrow pointing to an individual and a caption “That’s
me at Sleater-Kinney’s 1997 show in Cologne” – which the author of these lines
happened to miss).
However, the question of modes of pictorial reference is in many respects independent of the Gibsonian point that the primary function of pictures will not be
that of representation but that of providing a partial, delimited analogue of direct
perception of an environment and the guidance of activities in that environment.
In the latter respect, an abstract painting or a visualisation of a simulation might
operate in a manner quite distinct from a photograph of someone or something
I am acquainted with. However, the “mode of reference” variable only partly
covaries with the “perceptual guidance” variable. Concerning the latter, the central question will be what guidance something as limited and de-contextualised as
a picture could provide, and how.
A first approach to an answer will be to say that the possibility of pictorial
guidance depends on a selection of invariants to be preserved by the picture that
is suitable to the observer’s needs. A detailed re-creation of what has been seen in
the original perceptual situation cannot be the aim of a picture, but only the preservation of some of the information available in that situation. The observer may
also be supposed to identify the mapping relations of pictures of quite abstract
and formal kinds, and act upon them. The empirical rendering of a simulation is
often designed with quite pragmatic criteria in mind: directing different observers’
attention to what is deemed most relevant in the results. Indexicality may hence be
combined with iconicity, but may also override it if what the observer is supposed
to pick up so requires.
A second approach to an answer to the question of pictorial guidance will be to
say that the information which is inevitably missing in the picture will be filled
in by the observer: he or she may be related to the picture, or learn how to look at
the picture, in ways that allow him or her to identify it as a picture of some concrete individual entity or event that he or she knows, or as exemplifying types of
things or properties that he or she is familiar with. If the filling-in of the missing
information is done properly, a picture may well guide the observer’s behaviours
in a way partly comparable to the presence of the original world affair. For more
abstract picturing relations, as in simulations, the filling in will be of an equally
more abstract kind. In order to make the unobservable variables observable, a
preliminary concept of the structure and behaviour of the target has to be in place
when first-hand experience, via perceptual encounters with the target, is either
unavailable or insufficient.
A third approach to pictorial guidance will specify that one key element of
filling in the information that is missing in the picture will be to know how to
apply the relevant transformations of the mapping relations between the original
view of a scene or an object and those involved in seeing the picture. On these
grounds, Ruth Millikan (2004, Chapter 9) argues that, if we know what transformation rules are to be applied, we actually see, through a different medium but
mode of reference. Both the indexical and the symbolic mode of reference may
become part of an iconic picture, too (think of a photograph of a crowd at a rock
concert, furnished with an arrow pointing to an individual and a caption “That’s
me at Sleater-Kinney’s 1997 show in Cologne” – which the author of these lines
happened to miss).
However, the question of modes of pictorial reference is in many respects independent of the Gibsonian point that the primary function of pictures will not be
that of representation but that of providing a partial, delimited analogue of direct
perception of an environment and the guidance of activities in that environment.
In the latter respect, an abstract painting or a visualisation of a simulation might
operate in a manner quite distinct from a photograph of someone or something
I am acquainted with. However, the “mode of reference” variable only partly
covaries with the “perceptual guidance” variable. Concerning the latter, the central question will be what guidance something as limited and de-contextualised as
a picture could provide, and how.
A first approach to an answer will be to say that the possibility of pictorial
guidance depends on a selection of invariants to be preserved by the picture that
is suitable to the observer’s needs. A detailed re-creation of what has been seen in
the original perceptual situation cannot be the aim of a picture, but only the preservation of some of the information available in that situation. The observer may
also be supposed to identify the mapping relations of pictures of quite abstract
and formal kinds, and act upon them. The empirical rendering of a simulation is
often designed with quite pragmatic criteria in mind: directing different observers’
attention to what is deemed most relevant in the results. Indexicality may hence be
combined with iconicity, but may also override it if what the observer is supposed
to pick up so requires.
A second approach to an answer to the question of pictorial guidance will be to
say that the information which is inevitably missing in the picture will be filled
in by the observer: he or she may be related to the picture, or learn how to look at
the picture, in ways that allow him or her to identify it as a picture of some concrete individual entity or event that he or she knows, or as exemplifying types of
things or properties that he or she is familiar with. If the filling-in of the missing
information is done properly, a picture may well guide the observer’s behaviours
in a way partly comparable to the presence of the original world affair. For more
abstract picturing relations, as in simulations, the filling in will be of an equally
more abstract kind. In order to make the unobservable variables observable, a
preliminary concept of the structure and behaviour of the target has to be in place
when first-hand experience, via perceptual encounters with the target, is either
unavailable or insufficient.
A third approach to pictorial guidance will specify that one key element of
filling in the information that is missing in the picture will be to know how to
apply the relevant transformations of the mapping relations between the original
view of a scene or an object and those involved in seeing the picture. On these
grounds, Ruth Millikan (2004, Chapter 9) argues that, if we know what transformation rules are to be applied, we actually see, through a different medium but