The nature of cognitive artefacts 139
dysfunction. To recall, the change in question has to be, first, persistent in order to
qualify as a change in an informational environment (see p. 104 in Chapter 6). The
introduction of new affordances or modifications of existing affordances by whatever means needs to be well-structured, which means that a limited and clearly
circumscribed set of the variables within R’s informational environment is subject
to a non-transient change. Second, the change in question requires a coherent
pattern of modification, and hence has to be consistent in order to be accommodating by some means for R. Whether the change opens up new informational or
practical opportunities or increases informational uncertainty, it must allow him
to methodically alter his responses towards the signals or towards what is being
signalled.
To illustrate my threefold point in (e 1, 2, 3) with a contemporary example: if
I have information on my precise location available at any time, using a mobile
device that displays to me my GPS (Global Positioning System) coordinates projected into a map-like visualisation, I will be provided with information that is
factually identical with the information that I would pick up from a map that
I have learned to read, from a verbal description or from an environment with
which I am familiar. In the latter two cases, however, I will not have all that
information available at all times, or some of it may not be available at all. For
example I have to fill in the information on my current location myself when using
the map, which I might sometimes not be able to do. Or I must have at least partial
acquaintance with some place when trying to get my directions only from looking
at my surroundings.
This example might look a bit mundane, and it was deliberately chosen to do
so, because even in such a comparably simple case, which involves a technology
that is not intelligent on any AI-related account and whose use is presumably not
ecologically or selectively relevant in a life-or-death manner but may become
practically relevant in many respects and in many situations, all three conditions
mentioned previously apply: the mechanisms involved will have changed, my
purposes will have been affected and the availability and quality of the information on which I rely will have been modified. Think of a world in which almost
everyone is almost always guided by GPS devices, a world where maps and signposts have been consigned to obsolescence as archaic, primitive and insufficient
means of orientation, and where that device-based guidance is designed to feel
so natural that the use of the technology is hardly noticed by its own users, as in
Mark Weiser’s vision of “ubiquitous computing” (1991) quoted in Chapter 1.
Concerning the issue of mechanisms of information uptake (e 1), different cues
are used, and different skills are required for different modes of accomplishing
the same task, namely orientation. I do not need language to find my way by
purely environmentally bound visual orientation. Recognising a landmark might
involve the use of language but could, in principle, be accomplished without it.
I do need language to learn reading a map, but once familiar with it, guidance
will be more direct than from a description in natural language, as my orientation
remains operating on the level of visual cues. Asking someone for directions will
be helpful if I get lost and cannot match what I see on the map with what I see
dysfunction. To recall, the change in question has to be, first, persistent in order to
qualify as a change in an informational environment (see p. 104 in Chapter 6). The
introduction of new affordances or modifications of existing affordances by whatever means needs to be well-structured, which means that a limited and clearly
circumscribed set of the variables within R’s informational environment is subject
to a non-transient change. Second, the change in question requires a coherent
pattern of modification, and hence has to be consistent in order to be accommodating by some means for R. Whether the change opens up new informational or
practical opportunities or increases informational uncertainty, it must allow him
to methodically alter his responses towards the signals or towards what is being
signalled.
To illustrate my threefold point in (e 1, 2, 3) with a contemporary example: if
I have information on my precise location available at any time, using a mobile
device that displays to me my GPS (Global Positioning System) coordinates projected into a map-like visualisation, I will be provided with information that is
factually identical with the information that I would pick up from a map that
I have learned to read, from a verbal description or from an environment with
which I am familiar. In the latter two cases, however, I will not have all that
information available at all times, or some of it may not be available at all. For
example I have to fill in the information on my current location myself when using
the map, which I might sometimes not be able to do. Or I must have at least partial
acquaintance with some place when trying to get my directions only from looking
at my surroundings.
This example might look a bit mundane, and it was deliberately chosen to do
so, because even in such a comparably simple case, which involves a technology
that is not intelligent on any AI-related account and whose use is presumably not
ecologically or selectively relevant in a life-or-death manner but may become
practically relevant in many respects and in many situations, all three conditions
mentioned previously apply: the mechanisms involved will have changed, my
purposes will have been affected and the availability and quality of the information on which I rely will have been modified. Think of a world in which almost
everyone is almost always guided by GPS devices, a world where maps and signposts have been consigned to obsolescence as archaic, primitive and insufficient
means of orientation, and where that device-based guidance is designed to feel
so natural that the use of the technology is hardly noticed by its own users, as in
Mark Weiser’s vision of “ubiquitous computing” (1991) quoted in Chapter 1.
Concerning the issue of mechanisms of information uptake (e 1), different cues
are used, and different skills are required for different modes of accomplishing
the same task, namely orientation. I do not need language to find my way by
purely environmentally bound visual orientation. Recognising a landmark might
involve the use of language but could, in principle, be accomplished without it.
I do need language to learn reading a map, but once familiar with it, guidance
will be more direct than from a description in natural language, as my orientation
remains operating on the level of visual cues. Asking someone for directions will
be helpful if I get lost and cannot match what I see on the map with what I see