The domains of natural information 73
question but, being imperfect copies of their type, display some divergent characteristics and that, on the other hand, one may encounter individuals that display the seeming characteristics of the species in question but do not belong to
it genealogically. Instances of the latter possibility would be the equivocal traces
of p- and q-birds and the moving shadows of the ‘wrong’ species of Condor,
instances of the former would include any organism with dysfunctional organs
or other phenotypical aberrations. Prima facie, the same conditions will apply to
types of artefacts: an individual Ford Model T will be a Ford Model T because it
was designed by Ford Motor Co., and moved off the assembly lines at Ford Motor
Co. factories between 1908 and 1927. A bootlegged car or an artful replica of the
historical model, meticulously reverse engineered from an original Ford Model
T, would not count, and R would be fooled if having it sold to him as a classic
car from a fraudulent vendor. Conversely, a somewhat wayward Ford Model T
specimen, for having been produced at Ford Motor Co. factories according to the
Ford Model T design, albeit with some obvious aberrations, will still belong to
the same type.
5
However, just as pirated copies of digital media, every bit indistinguishable
from the original and produced in the same way as legitimate members of the type
but only illicitly so, partly undermine the seemingly clear-cut relations of descent
for types of artefacts, treating species as individuals is exposed to some ambiguities contained in their very definition. For example it will be difficult to precisely
determine a speciation event, both in terms of pinning down the moment when a
variety becomes a new species (a problem that has haunted evolutionary theory
ever since Darwin) and in terms of demarcating species from mere varieties in the
first place (a problem that has been haunting natural history even before Darwin).
In these and many other kinds of cases that concern the question of membership of some s in a type s, if this membership shall hold by virtue of certain typedefining properties rather than definitional stipulation on R’s side, the rules
governing the mapping between the domain of signals of an individual and the target set might turn out to be underdetermined. There may be boundary cases where
the mapping of signals pertaining to a putative member of the type in question
onto the target set, that is its being a proper member of the type, may be irreducibly equivocal. If this is so, some of the ambiguity that makes natural information C
probabilistic may well reside at the source of that information. In such cases, the
local character of this kind of natural information does not fall into one with its
probabilistic character. The species would be unequivocally defined at the centre
but less so at the spatio-temporal margins of its extension. Whereas the adherent
of a population view will leave the disambiguation of conditions at the margin to
R and her explanatory purposes, and find this unproblematic, the species-realist
will be more likely to ascribe to R the responsibility for that very residual ambiguity and to remain (implicitly) committed to a fundamentally non-probabilistic
view of natural information. He will also find it difficult to acknowledge that
ambiguity as a matter of fact.
In most relevant situations, however, the practical purposes of the perceiving
organism will dictate the rules of disambiguation so that he remains able to carve
question but, being imperfect copies of their type, display some divergent characteristics and that, on the other hand, one may encounter individuals that display the seeming characteristics of the species in question but do not belong to
it genealogically. Instances of the latter possibility would be the equivocal traces
of p- and q-birds and the moving shadows of the ‘wrong’ species of Condor,
instances of the former would include any organism with dysfunctional organs
or other phenotypical aberrations. Prima facie, the same conditions will apply to
types of artefacts: an individual Ford Model T will be a Ford Model T because it
was designed by Ford Motor Co., and moved off the assembly lines at Ford Motor
Co. factories between 1908 and 1927. A bootlegged car or an artful replica of the
historical model, meticulously reverse engineered from an original Ford Model
T, would not count, and R would be fooled if having it sold to him as a classic
car from a fraudulent vendor. Conversely, a somewhat wayward Ford Model T
specimen, for having been produced at Ford Motor Co. factories according to the
Ford Model T design, albeit with some obvious aberrations, will still belong to
the same type.
5
However, just as pirated copies of digital media, every bit indistinguishable
from the original and produced in the same way as legitimate members of the type
but only illicitly so, partly undermine the seemingly clear-cut relations of descent
for types of artefacts, treating species as individuals is exposed to some ambiguities contained in their very definition. For example it will be difficult to precisely
determine a speciation event, both in terms of pinning down the moment when a
variety becomes a new species (a problem that has haunted evolutionary theory
ever since Darwin) and in terms of demarcating species from mere varieties in the
first place (a problem that has been haunting natural history even before Darwin).
In these and many other kinds of cases that concern the question of membership of some s in a type s, if this membership shall hold by virtue of certain typedefining properties rather than definitional stipulation on R’s side, the rules
governing the mapping between the domain of signals of an individual and the target set might turn out to be underdetermined. There may be boundary cases where
the mapping of signals pertaining to a putative member of the type in question
onto the target set, that is its being a proper member of the type, may be irreducibly equivocal. If this is so, some of the ambiguity that makes natural information C
probabilistic may well reside at the source of that information. In such cases, the
local character of this kind of natural information does not fall into one with its
probabilistic character. The species would be unequivocally defined at the centre
but less so at the spatio-temporal margins of its extension. Whereas the adherent
of a population view will leave the disambiguation of conditions at the margin to
R and her explanatory purposes, and find this unproblematic, the species-realist
will be more likely to ascribe to R the responsibility for that very residual ambiguity and to remain (implicitly) committed to a fundamentally non-probabilistic
view of natural information. He will also find it difficult to acknowledge that
ambiguity as a matter of fact.
In most relevant situations, however, the practical purposes of the perceiving
organism will dictate the rules of disambiguation so that he remains able to carve