LINGUIST List 13.1334

Mon May 13 2002

Disc: Falsifiability vs. Usefulness

Editor for this issue: Karen Milligan <karenlinguistlist.org>


Directory

  1. Dan Everett, RE: 13.1330, Disc: Falsifiability vs. Usefulness
  2. Dan Everett, RE: 13.1330, Disc: Falsifiability vs. Usefulness

Message 1: RE: 13.1330, Disc: Falsifiability vs. Usefulness

Date: Mon, 13 May 2002 06:41:50 -0300
From: Dan Everett <dan_everettsil.org>
Subject: RE: 13.1330, Disc: Falsifiability vs. Usefulness

Of the three sets of remarks in 13.1330 re falsifiability vs.
usefulness, George Gale's posting seems to me to raise the only new
perspective on the issues.

Gale says: "Sometimes falsifiability works exactly as Popper described
it, and for a straightforward reason: the theorist designs the theory
to be a candidate for Popperian falsifiability." The example Gale
gives of a theory so designed is Bondi's research on Steady-state
cosmology. Bondi said what kind of counterexample would lead him to
abandon the theory and, when faced with an exemplar of such, did
abandon the theory.

In fact what I think this illustrates is that Bondi simply refused to
allow himself the equally Popperian breathing room of 'shelving' the
anomaly. More generally, if I tell you that 'If you do 'x' (e.g. find
an example of shape 'z'), I will do 'y' (e.g. abandon my theory)'
then, although this is an interesting bit of scientific history,
nevertheless, it need not be taken as support for falsifiability,
merely for my penchant for gambling. Or, in Bondi's case, how he chose
to behave in a given circumstance. There is nothing inherent in the
notion of falsifiability that would explain Bondi's behavior, except
that Bondi so interpreted it. Elsewhere, people (e.g., any major
linguistic theory) do not behave this way, i.e., they do not feel so
constrained by falsifiability because they know that one can avoid
falsifiability by *not* designing their theories so as to be deprived
of the right to 'shelve' problematic data. What Bondi did seems much
stronger than falsifiability, pretty much just a choice he made. I
look forward to reading Gale's discussion of the matter, however.

I think that Gale gets to the heart of the problem in his final
paragraph: "If someone wanted to deny that there were *genuine* cases
of falsifiability, I suppose that they could defend their hypothesis
and attack the Bondi case. As Lakatos always said, if you don't want
up a hypothesis in the face of contrary evidence, then don't. Nobody
can make you do it if you don't want to."

Just so - but this is what renders falsifiability less than
useful. This is exactly why falsifiability doesn't work, i.e. one is
never forced to admit that his/her theory or statement has been
falsified (so long as one says 'we'll just set this problem aside for
now and come back to it later'). Do I think it is unreasonable to set
aside problems, i.e. to shelve them? Not at all. I think this is
useful and that it makes good sense. But it renders falsiability a
mantra, rather than an advance in epistemology.

- Dan Everett
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue

Message 2: RE: 13.1330, Disc: Falsifiability vs. Usefulness

Date: Mon, 13 May 2002 07:40:33 -0300
From: Dan Everett <dan_everettsil.org>
Subject: RE: 13.1330, Disc: Falsifiability vs. Usefulness

Although I think these issues have already been dealt with, I will go
ahead and respond to Lofti and Sharifian, because I suspect that their
questions are shared by others on the list.

Lofti begins by citing my statement that "falsifiability has no
application ever."
He then goes on to say that "The concept "usefulness" is also vague and
empty unless one first decides to whom a theoretical statement is going
to be useful." Lofti seems right here. But I wouldn't have claimed
otherwise.

However, the examples which Lofti goes on to give about the earth and
sun turning miss the point. The point, again, is that either *all
statements can be made falsifiable (by conjunction)*, as per Hempel,
or *no statement is* unless the one who made the statement wants it to
be (Gale, Lakatos, and Hull). And in the history of science, as Gale
observes, cases where anyone ever claimed to have had a theory
falsified are so rare as to be aberrations (subject to alternative
explanations, as I pointed out relative to the Bondi case).
Therefore, we cannot merely invent hypothetical examples to support
falsifiability. We must address first the prior points of 'shelving'
and 'conjunction' (well, too, we would need to discuss the strange
idea that anyone would think that science is out to discover the
'Truth' or 'objective reality', which have merely substituted
religious concepts for scientific ones, but we should resist the
temptation to take on these subjects on this list).

Sharifian thinks that the problem I am having with falsifiability is
because I have failed to see that Popper is concerned about 'empirical
sciences' and the 'logic of scientific discovery'. In fact, I have
already explained in previous postings why I think that dividing off
the 'empirical sciences' from nonempirical work is not possible in
Popper's framework (as Lakatos, inter alia, has argued at length). And
as for the logic of scientific discovery, one simply finds no (or few,
if Gale is correct) examples of falsifiability playing a role in
acknowledged advances in science. So either science progresses
illogically, or falsifiability is not part of its logic.

- Dan Everett
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue