attention? (Or Pixie Dust and Moonbeams)

[Salinas17 at aol.com]

In a message dated 4/7/09 1:22:29 PM, keithjohnson at berkeley.edu writes:
<<The data have to be evaluated on their merits - otherwise we're 
evaluating research on 
the basis of the outcome not on the validity of the process.
So, I look forward to reading the Cosmides, Tooby & New paper.>>

Keith - You SHOULD look forward to reading the paper.   It is one of the 
funniest parodies of a scientific paper you can find outside of the Journal of 
Irreproducible Results.   Either that or it is a stark reminder of how 
close "evolutionary psychology" has come to being the study of pixies, 
moondbeams and extrasensory perception.   I prefer to see the paper as just an 
elaborate inside joke.

There sure are lots of guffaws throughout, but the laughably twisted logic 
that Tom seems to be addressing has to do with the use of the term 
"semantic" to describe the different kind of objects that appeared in photographs 
that were showed to subjects in the experiment.

Why the authors choose to use the word "semantic" isn't clear.   But the 
"five semantic categories" they give are "two animate (people and animals)" 
and "three inanimate... [plants; moveable/
manipulable artifacts designed for interaction with human hands/body (e.g., 
stapler, wheelbarrow); fixed artifacts construable as topographical 
landmarks (e.g., windmill, house)."

Tom's objection I suspect, is to the reason given in the paper for this 
Sesame Street-style scheme: "These categories were chosen because converging 
evidence from neuropsychology and cognitive development suggests each is 
associated with a functionally distinct neural substrate..." citing two papers by 
Alphonso Caramazza from 1998 and 2000.

Of course, the article's title tells us how the authors are using these 
"semantic" categories: "Category-specific attention for animals reflect 
ancestral priorities, not expertise."   But it doesn't tell us why they used the 
word "semantic" as opposed to something like "object-type" or "objects shaped 
like animals or humans." 

It is one thing to suggest that the brain stores language according to 
lexical categories, as Caramazza did. But it's quite another to say that we're 
paying extra attention to something we look at not because of what it looks 
like, but because it falls into a word group.   

Remember that this attention is supposed to be "not goal directed" - so 
you're not looking for Elmo.   In this experiment, you're supposedly not 
looking for anything.   But what you're paying attention to is determined by word 
categories like "animate" or "human", not by the fact you've just been 
shown, say, a blonde in a bikini.

As I think Tom pointed out, it would really take some straight-faced 
explaining to account for how "language-less" animals are so adroit at the 
fundamentally same recognition task.

But, because the authors of this paper were obviously being funny, they 
were looking for some absurd and humorous "neurological" way the human shape 
would grab more immediate attention than a stapler or a wheelbarrow.   Since 
they couldn't find research for a neurological pre-set for the human shape, 
they dug up some old research on the storage of words in the brain according 
to meaning categories. And then they say that's also why we innately pay 
attention to certain things, because of word categories like "animate" or 
"inanimate".   Now, that's comedy.

But that's not the really funny thing about this paper.   What's really 
funny is some great convoluted logic.   For example, we're told that the reason 
evolution favored us this pre-wired attention to "animals" is because:

"Not only were animals (human and non-human) vital features of the visual 
environment, but they change their status far more frequently than plants, 
artifacts, or features of the terrain. Animals can change their minds, 
behavior, trajectory, or location in a fraction of a second, making their frequent 
reinspection as indispensable as their initial detection."

Well, of course, one might suggest that an efficient way to spot something 
that moves is by paying attention to movement.   (A cat's incredible 
response time in reaction to motion is exactly that -- a triggering in the optic 
organ that by-passes normal neural processing.)

But the authors are not bothered by their own reasoning.   True, "animate" 
objects tend to be animated, but motion is not what's catching our attention 
here.

And, in the spirit of Mel Brooks, the authors have made sure that this has 
to be right by using an experimental technique called Change Detection, 
where "viewers are asked to spot the difference between two rapidly alternating 
[STILL PHOTOGRAPHS] that are identical except for a change to one object."

Hilarously, they discount motion by using STILL photographs of cars in this 
phase of the experiment -- because cars move right?   Now, that's funny.

Why not use video instead of stills?   Because still photographs show 
motion just fine, we are told.   

I'll give the authors the benefit of the doubt.   They don't think we are 
idiots.   They are just being comedians.

The best part are the photographs themselves.   We see a small figure of a 
human in a forest or by a resort harbor. This change is circled.   We are 
told the objects are in "natural situations."   Overwhelmingly the photos are 
filled with "inanimate" objects.   Inanimate objects added to the inanimate 
objects don't do as well as human images added to the inanimate objects. 
Wow, what a surprise!

Of course, we see lots of vague human shapes added to a forest scene or 
river scenes or travel scenes. But, with one exception, we don't see is a 
series of photos of something like someone holding a Christmas tree on a crowded 
subway platform.    So the human form always wins out over the wheelbarrow, 
the stapler and the silo.

Of course, it's common knowledge that good artists, graphic designers, 
photographers, film editors know how to drive attention where they want it.   
Each of these photos could have been re-designed to produce eaxctly the 
opposite result.   Luckily, the choice of photos is so obviously biased that we 
might think that the researchers were clumsily trying to fix the result - if 
we didn't already know they were trying to be funny.

And here's another punch line.   How do you know that the attention you are 
seeing isn't goal-directed?   How do you know that the subjects aren't 
already looking to spot fellow humans in the photograhs?   The answer is simple. 
  The subjects have no ulterior interests because they weren't given any.   
To qoute the article: "they are not given any task-specific goal that would 
direct their attention to some kinds of objects over others. Thus, the CD 
paradigm can be used to investigate how attention is deployed in the absence 
of a voluntary goal-directed search."

Thus is a human who has spent most of his waking life interacting with 
other humans is wiped clean of any "goals" when looking at a photo that might 
include other humans, and his "ancestral priorities" are stripped bare.   Just 
because we showed him a photo without telling him what to look for.   

Evolutionary psychology has generated much incredibly funny research, with 
the help of the NAS and numerous institutions listed in the article and 
elsewhere.   Together this effort has done much to satisfy our national need for 
humor.   In fact, sometimes when I read this stuff, I laugh so hard, I 
could cry.


Regards,
Steve Long




















**************
Feeling the pinch at the grocery store?  Make dinner for $10 
or less. (http://food.aol.com/frugal-feasts?ncid=emlcntusfood00000001)