[FRIAM] Spandrel

[jon zingale]

EricC, What again is the connection between goal orientation and function
in the evolutionary theory literature? All I can remember, from this
summer when we were discussing it, is that it was a way to distinguish
those things selected-for from the spandrels.

Not to muddy the waters, but nose *decoration* today is selected for, at
least to some extent, secondary adaptation? The same can clearly be said
of consciousness. Some of us select our mates out of an appreciation for
their perspectives, insights into their subjective experience. The "nose"
then becomes a *thing*. I mention this relative to Nick's question:

"""
...or has the fact of consciousness been seized upon by subsequent
selection to give it causal properties?
"""

I never did grok Glen's idea of *algorithmic depth*, and now I am
wondering if selecting at the level of goal versus selecting at the level of
function is an example? One may try to argue that my preference for a
sense of humor or compassion is a short-hand or substitution for
*deeper* underlying causes, but it seems a stretch. I somehow doubt the
computational ability of my person to do such a thing.

"""
Is the shape of the armpit been selected for dispersing hormones?
"""

To add to the confusion here, I do wonder if the shape is selected for,
but not in the sense that pheromones cause armpits. Instead, assuming
the inevitability of glands, I can imagine mutation driving their
migration and armpits becoming a basin of attraction.

As far as I can imagine, the locality question is possibly related. Should
it be the case that one type of selection giving rise to the shape of my
nose[d] also gives rise to a not so easily detectable change to my ass,
while another gives rise to the "same" nose[d] together with a not so
easily detectable change to my leg?

For whatever reason, I really enjoyed thinking about this problem from
the perspective of sorting algorithms and tuples. Extensionally, the
colors become sorted regardless of the final configuration of the balls
within any given level. That is, the goal-directed "algorithm" is free
two vary while holding the function invariant. Another variation I found
interesting was to imagine, instead of color, that the smallest balls
are labeled with numbers divisible by 2, the next smallest have labels
divisible by 3, and so forth. Now when I sort, some balls divisible by
2 will be at the top, but the majority of them will sort to the bottom.
In general, we see striation up to some probability. IDK, maybe there is
some nice way to think about the connection this schema may have to the
locality problem and the additivity of variance problem?

Anyway, just entering the conversation, so please pardon any pedantry
and staggering ignorance.

[d] Up to "a quality easily detected by humans".




--
Sent from: http://friam.471366.n2.nabble.com/