30 Informational environments
If, however, we follow Daniel Dennett (1987, Chapter 8, “Evolution, Intentionality, and Error”) in his account of the “Great Divide” in the contemporary philosophy of mind, the difference at issue here may ultimately be more fundamental than
between systematically grounded preferences for variant perspectives and variant
methods. Only to the extent that Millikan and other authors view information from
the perspective of its consumer, that is the receiving and interpreting organism,
and take this to be the relevant perspective for a theory of meaning, the problem
of markers or meta-information on the reliability of the signal arises – which is
to be solved by a functional-historical account of organisms dealing with signals
that, with varying degrees of reliability, are related to world affairs. Whether by
phylogenetic or by ontogenetic means, they acquire mechanisms for coping with
those varying degrees of reliability and for tapping into other information channels
that help to assess the affairs on the source side. Not so for Dretske and the authors
Dennett subsumes on the other side of the Great Divide: the firmness and consumerindependence, if I may call it so, of the informational relations that Dretske insists
on shall account for a solid grounding of whatever further operations in the organism’s nervous system will obtain in terms of representing world affairs. There is an
indefinite amount of natural information, and because that information is an entirely
objective affair, all the information there is could be, in principle at least, assessed
with determinacy. Only under this condition of a firm, nomologically governed,
rooting in natural informational relations, symbols can become meaningful, and be
endowed with what John Searle (1983) calls “intrinsic intentionality”, also known
as “original intentionality” – a notion that, on Dennett’s account, Dretske endorses
(Dennett 1987, 288f ) and that, if tenable, would spell bad news for computational
models and simulations of meaning-conveying processes.
Despite its rooting in natural relations, this latter view implies the assumption of
a discontinuity between intrinsic intentionality and the definition and characterisation of normal biological traits. Organic traits allow for the existence of intermediate and proto-forms and thus for some degree of ambiguity in definition and
function. The debates within and about evolutionary theory will never ultimately
resolve, nor can they possibly evade, questions of delimitation between biological
characteristics, such as, for example: at what stage in the course of evolution does
a (dinosaur’s) arm become a (bird’s) wing? Is the demarcation between species
determined by variance of traits, too, or by reproductive isolation only? When
precisely does a speciation event occur? Is evolution always a continuous process?
Even the Cambrian Explosion and Punctuated Equilibria, viewed in relation to the
time-frame of evolution, remain gradual processes, albeit with a steeper gradient.
8
In these processes, there are no unequivocal, predetermined thresholds to tell us
when something really has become something else, and when a fast-paced process
really amounts to a rupture in the course of evolution. In contrast, the phenomenon
of intrinsic intentionality, as conceived of by of Searle and like-minded authors,
makes its appearance as an either-or affair. It will be difficult to imagine it as
being partially realised or as an intentionality-in-the-making. Nor is it unequivocally admitted to being a product of processes of adaptation by natural selection in
the first place. An explicit plea for such a discontinuity that is based on an explicit
Précédent

- 43/231

Suivant