Transcript A dilemma for the extended mind

Scheduling error
Our officially scheduled final exam period is
Saturday at 10:30 a.m. We put Monday at
7:30 a.m. on the syllabus.
We still plan to hold the Monday session.
But if you cannot make it on Monday
morning and you want to give an extracredit presentation, please let me know.
We can have an additional session on
Saturday.
Distinction from Thursday’s class
Two general approaches to Iriki’s studies:
1. The self is wherever the subject
represents it as being.
2. The self is what does the representing (if
the representing is somewhere other than
what is represented, the representation is
inaccurate).
Consider two of your mental representations: one
of a hill, the other of yourself.
You can represent yourself as being on the hill, for
example, by the co-activation of these two
representations (properly bound), whether or not
you are actually on the hill.
So, the representing is happening in one place,
what is represented is represented as
happening somewhere else. Where is the self?
Applied to Cognition
It’s one thing for you to represent an external
thing as a literal part of your cognitive
process. It’s another for this representation
to be correct.
It’s possible that cognition is only where the
representing goes on, not at the place that
is represented as having cognition
occurring at it.
Assume for the sake of argument that your
cognitive system is in the brain.
There’s nothing to stop a cognitive system
located there from forming a
representation of itself as being
somewhere else.
Standing in a field, you can represent
yourself as being on a mountain, thinking.
Is your thinking taking place on the
mountain?
Back to Iriki’s monkeys
Possibility: All of the cognitive processing in
Iriki’s monkeys occurs in the brain (the
bimodal neurons in IPS are part of this
process), whether the monkeys are
retrieving food with their own hands or with
a rake.
The response patterns of the neurons in IPS
in fact shows exactly where the relevant
part of the cognitive process takes place.
Alternative View
The movement of the monkey’s hand is part
of the cognitive process to begin with.
So, anything that plays the same role as the
hand counts as part of the cognitive
process.
The response profiles of the bimodal
neurons in IPS shows that the rake and
the image of the hand etc., do play the
same role.
Markman and Brendl: Background
Pencil experiments (Strack, Martin, and Stepper,
1988:)
Cartoons rated funnier when holding a pen in hand
than when holding it with lips; even funnier if
holding it with teeth.
(But note that the effect goes away when it
pertains specifically to “objective” funniness, as
opposed to the subjective feeling of
amusement.)
Chen and Bargh (1999): Push-Pull
Two experiments:
First is a “conscious” condition:
Congruent: Push lever if word on screen is
bad, pull if good.
Incongruent: Push if good, pull if bad.
Responses are significantly faster in the
congruent condition.
Second experiment
Uses immediate reaction; no instructions
given regarding evaluation of words.
Always push or always pull, as soon as
stimulus is recognized.
Congruent responses (e.g., pull subjects
responding to good words) significantly
faster than incongruent response (e.g.,
pull subjects responding to bad words)
Does it matter where one
represents oneself as being?
Simon effect: same-sided response to stimulus is
facilitated (faster).
But it’s faster even if response is made with
opposite hand, so long as response button is on
the same side as the stimulus.
Suggests that there’s a distinction between a more
abstract representation of the location of the self
and something more like spatially guided brute
reflex (or low-level space-based associations).
Markman and Brendl’s Design
Subject’s name displayed in corridor. Instructions
given only in relation to location of the name.
Word far condition: word is on the far side of the
name
Instructions are either push lever if negative, pull if
positive or vice versa.
This is consistent with the arrangement in Chen
and Bargh. The relation of the word to the self is
the same as the word’s relation to the
representation of the self.
Word near condition: the word is in the
corridor between the viewer and the
viewer’s name. (same variation in
instructions)
In this case, pushing the lever away from
(where one represents) one’s actual body
also pushes the lever toward where one
represents (a representation of) one’s self.
Supposed to distinguish the effect of actual
location of the body from effect of
representation of the self.
Results
Significantly faster to move toward the name
(whether or not that involved pulling
toward oneself) in response to positive
words.
Significantly faster to move away from the
name (whether or not that involved
pushing away from oneself) in response to
negative words.
Does this implicate the use of abstract
representations in cognition? Only in some
kinds of cognitive processes?
How artificial or natural is the experimental
design? Does it take extra (or an unusual
kind of) effort to project oneself onto the
screen via the use of a linguistic
representation?