168 Environments of intelligence
In line with Azuma’s approach, AR researchers frequently refer to Paul Milgram’s continuum between real and virtual environments, in which AR is located
closer to the “real” than the “virtual” end of the continuum, within the broad realm
of “mixed reality” that stretches between these two poles (Milgram and Kishino
1994). AR appears to be the conceptual counterpart to Virtual Reality in many
ways: instead of cognitive and, possibly, behavioural immersion in a simulated
environment, information that is not or not directly accessible to perception is
added to the perceiving subject’s environment (LFE-a-2015), predominantly on
the level of visual perception. Such additional information may be displayed in
the form of text, audio, images, graphics or video on mobile phones or other
devices. In its most advanced versions, the information and interactive opportunities provided by AR applications are not confined to delimited displays but
projected onto objects and surfaces, so as to become integrated with their user’s
perception of his or her surroundings (Feiner 2002) – which would bring all three
of the aforementioned characteristics of AR to full realisation.
With respect to the intended functions of AR, one should clearly not presume a
contrast between reality and virtuality in which what is virtual appears as unreal,
and hence incurs connotations of being fake, made up and devoid of relevance.
In fact, AR applications are paradigmatically supposed to help to provide information about the real world to their users by means of virtual elements, some of
which may well refer to real objects in their user’s environments, and none of
which are supposed to tamper with correct and effective perception of the nonvirtual components of the environment – although they might end up doing so
under some circumstances. Only in gaming-oriented AR, as discussed earlier
(MMMOG), might the first of these two conditions be waived.
The reality/virtuality distinction in AR is supposed to highlight differences in
the mode of presenting information. If it is displayed on a screen or other delimited interface, the locality- and context-sensitive information provided by an AR
application remains clearly perceivable as having been added to the environment,
and it is still unequivocally localised on an interface that is itself an object in that
environment, equipped with real or virtual buttons, keys or bars for interaction.
In the most advanced cases of AR, however, the optic array itself becomes the
interface in a certain (although probably not properly Gibsonian) sense. Virtual
objects, although not being present in space, are projected onto surfaces in such
a way as to make them appear as integral parts of the environment. Ideally, they
should also allow for some degree of manipulation by their users, which poses
additional challenges for their integration into perception (Zhou et al. 2008).
If there is a mix of realities involved in AR, the reality/virtuality distinction will capture only one dimension of that mix, as it remains focused on the
mode of presentation of whatever kind of information it is concerned with rather
than the nature of that information. In order to account for the relation of what
is afforded by AR systems to natural information for perception, the distinction between informational convergence and isomorphism may be added as a
second dimension. An AR application may provide information about concrete
objects and activities in the environment that are of shared concern, and it might
In line with Azuma’s approach, AR researchers frequently refer to Paul Milgram’s continuum between real and virtual environments, in which AR is located
closer to the “real” than the “virtual” end of the continuum, within the broad realm
of “mixed reality” that stretches between these two poles (Milgram and Kishino
1994). AR appears to be the conceptual counterpart to Virtual Reality in many
ways: instead of cognitive and, possibly, behavioural immersion in a simulated
environment, information that is not or not directly accessible to perception is
added to the perceiving subject’s environment (LFE-a-2015), predominantly on
the level of visual perception. Such additional information may be displayed in
the form of text, audio, images, graphics or video on mobile phones or other
devices. In its most advanced versions, the information and interactive opportunities provided by AR applications are not confined to delimited displays but
projected onto objects and surfaces, so as to become integrated with their user’s
perception of his or her surroundings (Feiner 2002) – which would bring all three
of the aforementioned characteristics of AR to full realisation.
With respect to the intended functions of AR, one should clearly not presume a
contrast between reality and virtuality in which what is virtual appears as unreal,
and hence incurs connotations of being fake, made up and devoid of relevance.
In fact, AR applications are paradigmatically supposed to help to provide information about the real world to their users by means of virtual elements, some of
which may well refer to real objects in their user’s environments, and none of
which are supposed to tamper with correct and effective perception of the nonvirtual components of the environment – although they might end up doing so
under some circumstances. Only in gaming-oriented AR, as discussed earlier
(MMMOG), might the first of these two conditions be waived.
The reality/virtuality distinction in AR is supposed to highlight differences in
the mode of presenting information. If it is displayed on a screen or other delimited interface, the locality- and context-sensitive information provided by an AR
application remains clearly perceivable as having been added to the environment,
and it is still unequivocally localised on an interface that is itself an object in that
environment, equipped with real or virtual buttons, keys or bars for interaction.
In the most advanced cases of AR, however, the optic array itself becomes the
interface in a certain (although probably not properly Gibsonian) sense. Virtual
objects, although not being present in space, are projected onto surfaces in such
a way as to make them appear as integral parts of the environment. Ideally, they
should also allow for some degree of manipulation by their users, which poses
additional challenges for their integration into perception (Zhou et al. 2008).
If there is a mix of realities involved in AR, the reality/virtuality distinction will capture only one dimension of that mix, as it remains focused on the
mode of presentation of whatever kind of information it is concerned with rather
than the nature of that information. In order to account for the relation of what
is afforded by AR systems to natural information for perception, the distinction between informational convergence and isomorphism may be added as a
second dimension. An AR application may provide information about concrete
objects and activities in the environment that are of shared concern, and it might