🚀 go-pugleaf

Marc A. Belanger (3914MABE@vm1.ulaval.ca) wrote:
: >>
: >>Have you read Kuhn? In his famous book he agrues that falsifiability
: >>does not work as much as we think it does.

Are you talking about Karl Popper?

I think Feynman proposed a simple definition of the scientific method:
using experiment. Popper's falsifiability criterion works well when
an experiment needs to be designed.

Proof by contradiction is hoary, but since it assumes a two-value
logic, is still controversial especially outside of formal systems like
math.

Since linguistics, like psychology, economics, sociology, deals with real
people our formal fictions of reasoning don't work as well as we'd like.
So regardless of how scientific we want to be, we still have to argue
anecdotally, amass a weight of evidence, assess probable doubt and do the
other fuzzy things that are required in reasoning about people.

I like to use the principle of reconciliation of antinomies to invent new
theories, but that is another topic.

Perhaps as we extend the idea of logical positivism to vague reasoning,
storytelling, and other extra-logical means we can really make the 
whole of linguistics scientific, for some new definition of
"scientific". Science Fiction as pedagogy, that's the ticket.


-- 
Grady Ward       |  For information and free samples on | "Look!" 
grady@netcom.com |  royalty-free Moby natural language  |  -- Madame Sosostris
+1 707 826 7715  |  lexicons (largest in the world),    |     A91F2740531E6801
(voice/24hr FAX) |  run:        finger grady@netcom.com |     5B117D084B916B27