> It absolutely amazes me that Popperians like Deutch have no problems at
all
> talking about the many-worlds interpretation of quantum mechanics and
other
> things which appear on the face of them to be utterly unfalsifiable, but
> have a scunner against astrology, which provides me with twelve
falsifiable
> predictions free every day with my newspaper.
>
Miles Jackson (in reply to DD above):
>Yep. But just like any practical scientist, Popperians ignore or
>explain away examples that contradict their favorite theory. And >this is
not meant to diss scientists: holding onto a theory in the >face of
"unruly" data is in fact a crucial characteristic of science
>(PKF's stuff on Galileo is great on this). Anybody who thinks that
>scientists digilently throw out an elegant theory just because some
>data contradicts it needs to put down the Popper and actually talk >to some
practicing scientists.
And in a related thread, Jim Farmelant (Re: falsifiability, apparently both this thread and the Deutch and Popper one evolved from the Is Marxism a Science thread still going):
>I think that we got to be Duhemian-Quinean here and
>recognize that when a hypothesis seems to be contradicted
>by observation or experiment, we almost always have
>a choice of saving our hypothesis by invoking ad hoc
>hypotheses (which may be nothing more than that
>of "experimental error" or even "experimenter incompetence"!),
>or we can choose to reject our hyothesis, in which case
>we say that it was "falsified." The point here being we
>are not forced a priori into following one course or
>the other. Very often, if we have a theory that is able
>to explain most things in a very comprehensive but
>parsimonous manner, then we will most likely
>make great efforts at saving our theory if it is apparently
>contradicted by a few observations. In the case of less
>parsomonous, less elegant theories, we may be more
>likely to "allow" them to be falsified, if we run into
>anomalous observations.
I quote Jim here because he answers the Deutch and Popper sub-thread very well. The Duhem thesis is well worth invoking here. Falsification and verification are not simple matters and occur holistically and in a highly contextualized manner, if at all. I'm at a university where a lot of 'science' and 'technology' (Fukui University's College of Science and Engineering dominates the university) get done and have to remark just how much experimental activity is total failure. One thing that gives an illusion of science's efficacy is that only the 'successful' results get written up in highly edited and selective articles. I firmly believer that what 'progress' there is is more by accident than any design. And since so much science is done really for technology, creativity and intuition (yes, however, inscrutable, 'genius') are key. Most 'scientists' are just a bunch of plodders like the rest of us.
Also, in my 'profession' (foreign language teaching in an academic setting) I have to deal with pervasive naive positivism. At the level of helping seniors write their 4th year research projects I see them gamely setting out to 'prove' this or that isolated hypothesis about language learning or language teaching, etc. with nearl-silly 'experiments'. I see very little more sophistication at the PhD level where they set out to 'disprove' or 'falsify' their isolated hypotheses with just as silly 'experiments' about things like 'tasks', 'language processing', 'recall', 'motivation', etc. (You would think linguistics was the strongest influence, but it's not, it's largely a neobehaviorist experimental framework on language learning).
Most big steps forward in language teaching and learning have been more a result of certain, well-placed academic polemicists convincing others about the emptiness of a previous approach (such as the shift from structuralist-behaviorist audio-lingual approaches to communicative approaches). The problem is the current paradigm (as a teaching practice) is just as empty. It's also largely cut off from all that research and theorizing going on at universities.
Charles Jannuzi