Conceptual Aspects: How Landmarks Can Be Described in Data Models - Landmarks: GIScience for Intelligent Services - page 114

representations of an a priori given external world. Varela wrote: “Precisely the
greatest ability of all living cognition is, within broad limits, to pose the relevant
issues to be addressed at each moment of our life. They are not pre-given, but
enacted or brought forth from a background, and what counts as relevant is what
our common sense sanctions as such, always in a contextual way [...]. If the
world we live in is brought forth rather than pre-given, the notion of representation
cannot have a central role any longer” ([ 59 ] , p. 250f.). Similarly, Maturana said
earlier [ 38 ] that already the idea of a world out there implies a realm that preexists
its construction by an observer. Thus the reality out there comes into existence for a
living being only through (embodied) interaction.
Just as an example let us return to Sadalla et al. [ 50 ] , who reminded that
landmarks have multiple roles and are used in different spatial tasks: “The term has
been used to denote (a) discriminable objects of a route, which signal navigational
decisions; (b) discriminable objects of a region, which allow a subject to maintain
a general geographic orientation; and (c) salient information in a memory task.
These different referents suggest that landmarks may play a role in a variety of
spatial abilities” (p. 516). Thus, the Eiffel Tower, which had been previously called
generously a global landmark, may have low landmarkness according to (a), because
of its location and its quite extended footprint, high landmarkness according to (b),
because it is highly visible from many locations in Paris, and also high landmarkness
according to (c) because ...well, just everybody remembers or seems to know the
Eiffel Tower.
This brings up a dilemma. There is no (complete) formal model of landmarkness
without context, but context, as an open-ended and arbitrary-dimensional system
cannot be captured completely in a formal model. Even worse, if context cannot be
captured completely then differences between computed or assigned landmarkness
values, especially marginal ones, may be misleading or meaningless. So after all,
what can we model and state about landmarkness?
With hard to defend differences in the quantity of landmarkness, the fundamental
basis of above's landmarkness value remains the distinction between (a) objects
with landmarkness (in some context), (b) objects with no landmarkness (in this
context), and (c) objects with unknown landmarkness (in this context). The rest
is a matter of grading. And grading is, considering the aim of this landmark model,
actually irrelevant. The purpose of the model is two-fold: supporting the machine
interpretation of human place or route descriptions, and supporting the machine
generation of human-like place or route descriptions. If I am asking passers-by in
a street for a route I will get a different description from each one of them. Each
communication context is different because in each communication I speak to a
different person. They may describe the same route (perhaps the only possible or
reasonable one in this environment), but they may refer to different landmarks, and
to different numbers of landmarks. The speakers choose from their mental spatial
representation what they find relevant in the given situation. Furthermore, by taking
into consideration their description in my wayfinding process my attention would be
already primed to search for the objects referred to. I will not look out for alternative,
possibly more outstanding objects and wonder why they have not been used, which