[lit-ideas] Re: Patrick and the Snakes: the logic of falsification
- From: Donal McEvoy <donalmcevoyuk@xxxxxxxxxxx>
- To: "lit-ideas@xxxxxxxxxxxxx" <lit-ideas@xxxxxxxxxxxxx>
- Date: Tue, 19 Mar 2013 10:50:12 +0000 (GMT)
Deep breath...
________________________________
From: Omar Kusturica <omarkusto@xxxxxxxxx>
>Just so that this wouldn't be too simple, would the statement: "There are
>snakes in Ireland" be scientific ? It is difficult to see how this claim could
>be falsified, although it could conceivably be verified if we found a snake.>
As with statements that prima facie are scientific, like "Here is a snake (or
'At time t, at location l, a snake was observable')" or "All snakes have
scales", the scientific status of "There are snakes in Ireland" depends on the
methods adopted to defend it from possible falsification by observation.
"There are snakes in Ireland" falls, logically, somewhere between an
unrestricted or purely existential statement like "There exists a snake
(somewhere, unspecified)" and a restricted existential statement of the sort
that constitutes a 'test statement' like "Here (specified) is a snake (or 'At
time t, at location l, a snake was observable')". It falls logically between
these because it is not so unrestricted as the purely existential statement and
it is not so restricted as a test statement.
An unrestricted or purely existential statement is not scientific per se
because it cannot be falsified - as the absence of snakes wherever we have
searched would not logically falsify the claim that somewhere else (say in a
gallaxy far away) there is a snake. The test statement is scientific because
the claim "Here (specified) is a snake" would be falsified if we looked and
observed no snake "here".
However, statements do not give more than a prima facie indication of their
scientific status - the crux is always what methods are used to defend the
statement from possible falsification. So while prima facie scientific, the
claim "Here is a snake" could be rendered unscientific if defended against
'falsification-by-observation' by maintaining it is an unobservable snake: for
clearly "Here is an unobservable snake" is not testable by observation.
Likewise, while prima facie scientific, the claim "All snakes have scales"
could be defended in a way that renders it unscientific - say, if an unscaly
counter-example were dismissed on the basis that because it is unscaly it
cannot be a snake (then it would be clear the claim is being defended as a
definitional truism and not as one testable by observation).
From "Here (specified) is a snake" we can deduce "There exists a snake
(somewhere)": so the latter statement, though unscientific per se, may be
accepted as 'scientific' where it is deducible from a scientific statement. But
this deduction is not a matter of 'verification by observation'. Strictly
speaking, the latter statement is not per se 'verified' or 'verifiable' (for
that 'there are snakes' is too broad a statement to be 'verified', or
falsified, by any observation) - rather its truth may be deduced or deducible
from the truth of a test statement like "Here is a snake". It is only the test
statement that may be 'verified' or 'falsified' by observation, not the purely
existential statement (even though the truth of the purely existential
statement is deducible from the truth of the test statement). [Though a
digression is here avoided, it is important to note that talk of 'verifying' a
test statement does not amount here to any inductive theory of
'verification' - in truth, a test statement is only ever 'verified' to the
extent that its truth is not falsified by observation, and so when we speak of
'verifying' a test statement by observation we are really speaking of whether
it survives attempts to falsify it by observation.]
As above, from "Here (specified), in Ireland, is a snake" we can deduce "There
exists a snake in Ireland (somewhere)": so the latter statement, though perhaps
unscientific per se, may be accepted as scientific where it is deducible from a
scientific
statement. Strictly speaking, the latter statement may not per se be verified
or verifiable (for that 'there are snakes in Ireland' may be too broad a
statement to be 'verified', or falsified, by any observation) - rather its
truth may be deduced or deducible from the truth of a test statement
like "Here, in Ireland, is a snake".
But if the truth of "There exists a snake in Ireland" may be deduced from the
truth of a test statement
like "Here (specified), in Ireland, is a snake", so its falsity may be deduced
from a set of 'test statements' that exhaustively cover Ireland if those 'test
statements' are all false (i.e. if none of the 'test statements' in the
exhaustive set is true, we may deduce that it is not true that "There exists a
snake in Ireland").
There is an obvious problem here: namely, that the set of such 'test
statements' can never be given exhaustively in a strictly logical sense: for it
is always logically possible that some other 'test statement' is missing from
whatever set we assemble. Nevertheless, we can make the set as exhaustive as
practically possible, and we can decide to accept that sufficiently exhaustive
set as determining the matter. So that if a sufficiently exhaustive search of
Ireland turns up no example where it is true, by observation, that "Here is a
swan", then we may conclude that we have tested and falsified the claim that
"There are swans in Ireland".
What matters in the case of "There are snakes in Ireland" is, then, what always
matters - the methodology that underpins its defence from possible
falsification. If that methodology accepts a search of a high degree of
exhaustiveness as determining the matter (so if no snakes were found, the claim
is falsified), then the claim might be deemed scientific (because falsifiable).
But if, no matter the exhaustiveness of fruitless searches, the claim is
defended on the basis that searches are inadequate to find the snakes, then it
becomes an unfalsifiable, untestable, unscientific claim - for the claim is, in
effect, being defended as beyond our powers of observation to test it by
falsifying it.
Hope this helps clarify the "logic" of what here would constitute 'science' -
and what would not. As may be seen, this clarification involves no induction
but only a clear grasp of the logic and methodology of 'possible falsification'.
Donal
London
Other related posts:
- » [lit-ideas] Re: Patrick and the Snakes: the logic of falsification - Donal McEvoy
