Nice one, but, a diversion into a side-issue mentioned in the paper:
... "the assignation of approx. 37% of elements to the negative pole of
any construct could make negative evaluations stand out maximally as
"figure" against a "background" of positive evaluations, thereby
highlighting the contrast"...
addresses itself to me as an explanation,
because it recalls the 62%-38% ish proportions of the Golden Section
"diagram". And it succeeds in convincing me in a way that an appeal to
Frank's observation that 0.37 is the maximum value taken by any one event
in contributing to Shannon's H does not: given that the H function is
rather flat at its peak or, as Jack puts it, "relatively small deviations
of psubi from 1/e involve little loss of strikingness".
Then when I accept that there is some brain-functional explanation a la
Lefebvre that makes this 62%-38% proportion "natural", I ask myself
whether a maximal "figure-ground" contrast would be precisely any
proportion which _broke_ this proportion, rather than _adhered to_ it.
For example, putting a single line of text exactly half-way down a page,
instead of 62% of the way up from the bottom, is surely _more_ striking
because it breaks the Golden Rule as applied to page layout which makes
62% of the way up the page "look just right".
I suspect the problem is in the words: "striking", etc. (I was never very
happy with Frank's use of "surprisal" in any case.)
Has anyone any thoughts or should I wait until I've had my first G&T
of the evening and seen the error of my reasoning?
(Jack, are you there?)
Devi Jankowicz
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%