I was trying desperately to write something
about the Kamp/Raffman/Soames/Graff theory of vagueness, and I noticed that
both Kamp and Raffman note that their theory makes an empirical claim, one that
they are apparently sure is true, but which they have never tested. Well, I
haven’t tested it either, because testing theories costs real money, and my
research fund is lucky to run to a couple of conferences and a few books a
year, let alone real experiments. But I did come up with a way to do an
uncontrolled experiment on the hypothesis in question.

If
you want to take the experiment yourself, open this file and unzip it. Then
open the Word document in it and answer the onscreen questions until you get a
summary sheet of the results. Take note of the two numbers you are given, they
will be important. I won’t yet tell you what Kamp and Raffman’s prediction is
concerning those two numbers, it might be better (well, less appallingly awful
from the pov of experimental design) if you don’t know that yet. Note that the
test works best if you have your computer set to run in True Colour, and
probably doesn’t work at all if you aren’t running Word 98 or later. (If I
could only learn to program the latter two problems could be fixed – my skill
set is still, sadly, set-sized.)

Vagueness
Test

If you have advanced virus protection you
may have to be quite insistent with your computer or it won’t let you run the
attached macros. Trust me, you won’t catch a virus this way. (Not that I
guarantee anything in case you do :))

Have you taken the test yet? Good, keep
reading. If not, go back and take it you slacker!

The result Kamp and Raffman want is that
the first number is higher than the second number. Essentially, they claim that
among the many technical flaws in our perceptual system is a hysteresis in our
colour perception. If you slowly change a colour from red to purple (they both
use orange, but the experiments are easier I find with purple) then the change
in apparent colour will lag the change in actual colour. So the effect will be
that we judge some colours as red if we have previously been looking at reds,
but we will judge the very same colours as non-reds if we have previously been
looking at non-reds. If this is true, then when you run the experiment, the
second of the two numbers you get at the end should be lower.

For
what it’s worth, I do get this result when I run the test. I was rather hoping
I would not, so I could quickly refute this theory and go back to working out
the true ‘truer’ theory. If you run the test, let me know the results, and I’ll
keep a very unscientific running tally of the totals. If the test doesn’t work
also let me know. If running the test causes grave computer malfunctions, call
an expert, I’m going to be of no help whatsoever.