I see your point about the logical difficulty of implications in construct
research. Using your example it is clear that having a name like "Mary"
correlates with that a person is a female, but, as you say, being a female
does not imply that one has the name "Mary" (or any other name). If we
assume that people have traditional names, perhaps we can assume that
people named Mary are female, but this too seems suspect. The
relationship is conventional, and conventions are made to be broken.
Even if conventional names are presumed, it isn't clear, as you explain,
if Mary should be a superordinate concept implying "Maryness" and all
female traits or not. This would, as you say, make "the individual more
complex than any of the abstractions that we may use to describe her."
Your comparison with the issues in Jung's theory of archetypes is
interesting, too. You note that for Jung, individuation consisted in
fostering the individuality of the person around the archetype, not
reducing the individual to a simple archetype. The analogy, presumably,
is that we don't want to reduce the individual "Mary" to imply an
archetype of femaleness. We want to sustain her individuality, her
ability to break the mold.
Then you switched to talking about the meaningfulness of whole figure
constructs versus traits. This has me confused. I suppose you have the
model of Mary in mind here. If Mary implied her archetype (to give this
a Jung spin), then would this be a "whole figure construct"? You say
that you have argued that such whole figure constructs may be more
meaningful than traits. But, on the the surface, that seems counter to
everything you explained in your last note. Unless I'm missing
something, you're torn about that. And, if you're torn, perhaps you can
share with us your pondering of this puzzle. You say "the stats held
up". Do you men the regression weavings? And how is this significant
in your mind for both the regression studies and the coordinate grid studies?
Maybe it would help if you used the example above to elaborate the
following concept:
In the coordinate grid method, I assume that
perfect integrative complexity occurs when
whole figure constructs like Mary (figures) are not
subsets of other whole figure constructs. In such
a case, the whole figures are coordinates-not
subordinates or superordinates- hence the name
coordinate grid.
Then you gave an example of rating myself along 5 trait dimensions that I
consider essential to my character and do this for 50 situations.
The point, I believe, is that if I rate myself as very Lois-like in these
situations, this presumes I have chosen traits that are very Lois-like,
and this involves me, it seems in a circularity of definition. I may be
defining my "Lois-like traits" by my peculiar manifestation of them. For
example, I might consider myself intelligent, when by intelligent I mean
make good purchasing choices. This, too, seems to be a logical problem
in the study of implication.
Seems like a conceptual hornet's nest. Does your statistical weaving
offer relief?
..Lois Shawver
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%