On 26/05/2017 9:11 am, David Nyman wrote:
On 25 May 2017 23:18, "Brent Meeker" <[email protected]
<mailto:[email protected]>> wrote:
I have told you my theory of virtuous circular explanations.
"Invoke" is a pejorative attribution. The physical universe is an
*/inference /*to explain appearances (and a very successful one at
that).
Vocabulary. The point is, assuming mechanism (and please do tell me if
you're reasoning in a different theory), that the inference is to a
particular *selection* of computations from the computational
plenitude. And why is that? Because they 'explain' the appearances.
But do they really? Are those computations - in and of themselves -
really capable of 'explaining' why or how they, and no others, come to
be uniquely selected for our delectation? Are they really capable of
'explaining' why or how those selfsame appearances come to be present
to us?
I think you and Brent are using different notions of "explanation". As I
understand your (David's) position, it is a notion of "explanation"
originating with Plato: Plato's theory of Forms offered at the same time
both a systematic explanation of things and also a connected
epistemology of explanation. (Summaries from Jonathan Cohen in the
Oxford Companion to Philosophy.) In other words, the Platonic ideal is
that "Ontology precedes epistemology", to vary Brent's slogan. In the
case of mechanism, the ontology is the natural numbers (plus arithmetic)
and for an explanation to be acceptable, everything has to follow with
the force of logical necessity from this ontology.
As I understand Brent's position (and that is essentially the same as my
position), his concept of "explanation" follows the tradition of British
empirical philosophy, stemming from Bacon, through Hume, to Russell and
others. In this tradition, to explain an observed characteristic is to
show its relationship to a law in accordance with which the
characteristic occurs or can be made to occur, and there is a hierarchy
of such laws -- the more comprehensive laws are deemed more probable.
This leads to the dominant model for explanation in the natural
sciences, which requires the citation of one or more laws which, when
conjoined with the statement of relevant facts, entail the occurrence of
the phenomenon or uniformity that is to be explained. This does not rely
on any assumed ontology; hence, "Epistemology precedes ontology".
Wherever we want to derive a technology from scientific knowledge, we
shall need to know what causes a desired effect. So we need to
distinguish between different levels of explanation, in that while, for
example, the disappearance of a patient's infection may be causally
explained by his antibiotic injection, the operation of that causal
process is in its turn to be explained by correlational laws of
biochemistry. Hence, the understanding of consciousness in any effective
way will be linked to the creation of effective AI.
This is the paradigm of current scientific practice. Sure, as Bruno
says, this stems ultimately from an Aristotelian approach to science
rather than the Platonic approach. But the history of Western thought
has shown the scientific, or Aristotelian, approach to have been
overwhelmingly more successful, both in developing technology and in
reaching understanding of the nature of reality.
Bruce
Or is this really just the latest case of "If I can't see it, it
doesn't exist?" Perhaps you will be content to say that whatever those
computations are capable of explaining sets the absolute limit of what
we can ask in explanation. But if that's the way it is, I can't help
being put in mind of that old huntsman John Peel, who would assert,
with a remarkable satisfaction in the virtue of invincible ignorance,
"What I don't know ain't knowledge."
David
--
You received this message because you are subscribed to the Google Groups
"Everything List" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To post to this group, send email to [email protected].
Visit this group at https://groups.google.com/group/everything-list.
For more options, visit https://groups.google.com/d/optout.
