[lit-ideas] Re: My father's wound
- From: Donal McEvoy <donalmcevoyuk@xxxxxxxxxxx>
- To: "lit-ideas@xxxxxxxxxxxxx" <lit-ideas@xxxxxxxxxxxxx>
- Date: Thu, 16 Aug 2012 22:21:17 +0100 (BST)
Just a minor point for now, for Eric, in a part of his comments, has trespassed
into the problem - or problems - of induction.
________________________________
From: Eric Yost <mr.eric.yost@xxxxxxxxx>
>If that were otherwise, there would not be so many different contradictions
in philosophy of science, e.g., Hempel's Raven Paradox>
For "Hempel's Raven Paradox", I take it we have something in mind like the
following:- "All ravens are black" is equivalent to "All non-black entities are
non-ravens"; in which case, what is an example that supports one must be an
example that supports the other; in which case, it would seem that my big toe
[being non-black and not a raven] is just as much supporting evidence for "All
ravens are black" as is a black raven.
According to Popper, and here he is right, Hempel's so-called paradox is
neither a genuine logical paradox nor something that grounds a contradiction:
far from being either a paradox or some contradiction, the "paradox" is in fact
a demonstrable theorem in the calculus of probability.
The fact is that anything anywhere - except a falsifying instance - is an
instance of something that is consistent with "All ravens are black". A black
raven may be a positive instance of the generalisation "All ravens are black",
but it no more supports that generalisation than my big toe or indeed empty
space. The black raven, my big toe and empty space are logically on the same
level - they are each logically consistent, and equally consistent, with "All
ravens are black"; and it is wrong to think one of these [the black raven] is
more consistent with "All ravens are black" than either my big toe or empty
space.
The Hempel Paradox is only a paradox if we make the (logical) mistake of taking
a positive instance to be some kind of supporting (inductive) evidence for a
generalisation: for the Paradox indicates that a positive instance cannot play
this role any better than any other kind of instance such as my big toe or
empty space. Indeed, far from being a "paradox", this is logically the case: if
we are considering "All ravens are black", and look in a series of boxes to
test its truth, only boxes that contain non-black ravens tell us logically that
the generalisation is untrue; boxes containing black ravens no more tell us the
generalisation is true, or probably true, than boxes that contain severed toes,
or that are empty or that contain anything that is not a raven.
Of course, as the Wason Test may indicate, we are inclined - and may even be
taught - to take a positive instance as supporting evidence; or as greater
supporting evidence than any other kind of instance. But a positive instance
is only supporting evidence in the sense that it is the absence of a falsifying
instance; but so, equally, any instance of anything - except a falsifying
instance - is supporting evidence in the sense that it is the absence of a
falsifying instance.
From a logical POV, it is only the absence of a counter-example or falsifying
instance, and not the existence of an example or positive instance, that
provides critical reasons for accepting a generalisation. In this we should
regard the generalisation as being well-tested only if we have searched
rigorously for a counter-example - that is, only if we have tried and failed to
find a counter-example. No scientist would regard a generalisation as
well-tested merely because they could adduce one or more positive instances of
it: indeed, the accumulation of positive instances would be misconceived - for
if a single counter-instance were produced the generalisation would be
falsified. So the scientific method is geared not to the accumulation of
positive instances but to the search for counter-instances [thus the
accumulation of positive instances is only significant because there are no
negative instances in that accumulation; which means the significance of
any such accumulation lies in how it results from a severe but fruitless
search for negative instances].
Hempel's so-called Paradox is only a "paradox" within an inductivist philosophy
of science that tries to ground science in "supporting evidence" that is
constituted by the existence of positive instances of scientific claims. But
this is a mistaken philosophy for many reasons, especially logical. The paradox
does not arise within a non-inductivist and critical methodology that regards
the testability of claims as equivalent to their falsifiability; and which
regards such claims as well-tested not because they have positive instances but
because there is an absence of negative instances despite a critical and severe
or rigorous search for them.
It is because scientists see this clearly enough that they often find Popper's
philosophy of science so enlightening as an account of their method and its
logic (unless their heads have been turned by some philosophical jiggery-pokery
of the Hempel Paradox kind, for it is nothing of the sort [i.e. a "paradox"]
except to philosophers trapped in inductivism).
Donal
Other related posts:
