Transcript romistalk - Marieke Rohde

Space, time and the nervous system
Romi Nijhawan
University of Sussex
Summary of the talk
• Philosophy; Neutral monism; the flip-flop analogy
• Basics of visual processing. Time delays in the
transmission of neural signals
• Neutral monism and the Flash-lag effect
• Neutral monism and the life-span development of
the scientist (from zygote  embryo  fetus 
newborn  adult)
Neutral monism
• Holds that ultimate reality is of one kind
• Intrinsic nature of ultimate reality is neither
mental nor physical; it is neutral between the two
–
–
–
–
Baruch Spinoza (1632-1677)
Ernst Mach (1838-1916)
William James (1842-1910)
Bertrand Russell (1872-1970)
Bodies do not produce sensations, but
complexes of sensations (complexes of
elements) make up bodies. If, to the physicist,
bodies appear the real, abiding existences,
whilst sensations are regarded merely as their
evanescent, transitory show, the physicist
forgets, in the assumption of such a view, that
all bodies are but thought-symbols for
complexes of sensations (complexes of
elements).
- Mach (1897)
Thus the great gulf between physical and
psychological research persists only when we
acquiesce in our habitual stereotyped conceptions. A
color is a physical object as soon as we consider its
dependence, for instance, upon its luminous source,
upon other colors, upon temperatures, upon spaces,
and so forth. When we consider, however, its
dependence upon the retina…it is a psychological
object, a sensation.
Not the subject matter, but the direction of
investigation, is different in the two domains.
- Mach (1886)
The data of psychology do not differ in their intrinsic
character from the data of physics. I have
maintained that sensations are data for psychology
and physics equally…
- Russell (1921)
Common sense imagines that when it sees a table it sees a table. This is a gross
delusion. When common sense sees a table, certain light waves reach its eyes, and
these are of a sort which, in its previous experience, has been associated with
certain sensations of touch, as well as other people’s testimony that they also saw
the table…
The light waves caused occurrences in our eyes, and these caused occurrences in
the optic nerve, and these in turn caused occurrences in the brain…
(Of course, if matter in general is to be interpreted as a group of occurrences, this
must also apply to the eye, the optic nerve, and the brain).
- Russell (1959)
The modern would-be materialist thus finds himself in a curious position, for,
while he may with a certain degree of success reduce the activities of the mind to
those of the body, he cannot explain away the fact that the body itself is merely a
convenient concept invented by the mind. We find ourselves thus going round and
round in a circle: mind is an emanation of body, and body is an invention of mind.
Evidently this cannot be quite right, and we have to look for something that is
neither mind nor body, out which both can spring.
-Russell (1929)
Neutral monism: an analogy
(In the analogy Flip and Flop represent
opposite extremes of a dualistic philosophy)
Flop
Flip
•
Real
•
Studied by physicists
•
Relevant to psychologists and
neuroscientists as a stimulus only
•
Well known proponents: Newton,
Einstein
•
Perceived
•
Studied by psychologists and
neuroscientists
•
Relevant to physicists only as a source
of intuitions about ‘forces’ and other
quantities; or in interpretation of data
(including design of measuring devices)
•
Well known proponent: Berkeley
In the analogy a move toward neutral monism
Flipflop
Neural delays in the retina
LIGHT
Receptor hyperpolarization
Delays due to signal transmission from retina
to primary visual cortex and beyond
Neural delays in response to a discrete visual stimulus
Neural delays in processing of a continuous (moving) visual stimulus
Flop?
Perceived
Flip?
Real
• Assume:
• Object velocity = v
• Neural processing Delay= t
“vt-lead premise”:
• Due to neural delays a moving object’s real
position should lead its perceived position by vt.
• Dilemmas and paradoxes
• Phenomenon: Flash-lag effect
On the ‘real position = perceived position + vt’ premise
• Real and perceived are spatially separated for motion but
not for stationary objects (i.e. when v = 0)
• For stationary objects, are the real and the perceived one
and the same, or are the real and the perceived co-localized
(congruent)?
stationary
object
Flash-lag effect
Dark lab, two light sources
and a bar painted white.
Light source 1
Light source 2
What do observers see at the instant of the flash?
Percept
F
M
F
“Flash-lag effect”
Movement direction
demo
What does the flash-lag effect have to
do with neutral monism?
“I see, therefore, no opposition of the physical and the
psychical, no duality, but simply identity.”
-Ernst Mach, 1890
Mach denied quantities such
as absolute space and absolute time
due to the fact that these quantities
were not observable
“This book exercised a profound
influence upon me… while I was a student.”
--Albert Einstein
•Can the spatial lag between the perceived
and the real position of the moving object
be observed (measured)?
•Is this observation possible even in
principle?
NO!
A
B
midpoint
01:004
Following Mach: If it is unobservable it does not exist
A
B
midpoint
01:004
On the ‘real position = perceived position + vt’ premise the
following is a correct picture
But this assumption leads to contradiction
Consider an object moving at
Velocity = v
v
sensory delay = t
vt
R
E
A
L
Perception
trails behind
20 - 160 ms
FLE
Flash source (off)
vt
0
clock
0
FLE
vt
0
FLE
vt
0
FLE
vt
Either the above picture is completely wrong.
OR
It is partly wrong and there are two different t’s:
tm and tf, with tm < tf
No evidence for faster processing of motion
280
RT FLASH
270
RT MOTION
260
250
240
230
0
1.17
2.34
3.51
4.68
Velocity (deg/sec)
Nijhawan et al. (Visual Cognition, in press)
Lead of flash for perceived
simultaneity (ms)
No evidence for faster processing of motion
25
15
5
-5
-15
-25
1.17
2.34
Velocity (deg/sec)
4.68
Conclusion:
The above picture is wrong
and neutral monism is correct
Intuition and analysis suggest very different answers
concerning the functioning of the central nervous
system.
•Intuition suggests that there is a sharp boundary,
and a wide gap, between the sensory and motor
functions of the CNS.
•Intuition also suggests that there are “real” objects
out there in the world which are “perceived” by the
observer.
•A close look at the brain reveals that more
than 95% of neurons in our CNS can
neither be categorized as “sensory” nor as
“motor”.
•The nervous system development starts out
with both sensory (touch-proprioception)
and motor functions emerging from the same
cells.
•Many neurons in the somatosensory cortex
project to the motor neurons in the spinal
cord.
•Many neurons in the premotor cortex
respond to sensory stimulation.
The main divide between the sensory and
motor functions of the CNS seems to be
centered on the following intuition:
Sensory processes are geared to processing
objects in the “real” world, while motor
processes seem geared to moving limbs etc.
of the body to which these processes are
attached.
The INPUT vs OUTPUT divide
The INPUT vs OUTPUT divide has, I believe,
kept scientists from accepting that sensory and
motor processes are two facets of one
underlying process.
Factors maintaining the INPUT vs OUTPUT divide:
•OUTPUT: The brain “generates” commands
that move limbs.
•INPUT: Science is dominated by sighted
individuals, so the concept of “real objects”
is synonymous with “objects out there”. If
objects are “out there” then information must
travel from objects to our eyes. Hence the
concept of INPUT.
Here is the dilemma: Talking about
input in the absence of what it is an
input of is meaningless. If we talk
of an object as the source of the input
to the visual system (say), then we are
using the OUTPUT of the brain
—“the object” —to call it the source
of the input. Thus, when we use the terms
‘sensory INPUT’ in the traditional sense,
we are confusing OUTPUT with INPUT.
I argue that “real” objects are actually a result
of one branch (branch B) of the output of the
CNS. The other branch of the output (branch
A) is dominated by touch, proprioception and
the motor functions. Branch A is earlier both
in phylogeny and ontogeny than branch B.
Thus both sensory and motor functions are
outputs.
Speculations based on evolution, development and plasticity
The human nervous system development
Human development: 10 - 41 weeks
41 weeks
… the possibility of learning the significance
of the local signs which belong to our
sensations of sight, so as to be able to
recognize the actual relations which they
denote, depends, first, on our having
moveable parts of our own body within
sight.
-- Helmholtz
Functional significance of arm movements in neonates
10 - 24 days old
A. van der Meer, F. van der Weel & D.N. Lee, Science, 1995
Motor flash-lag
Voluntary movement of a rod in the
absence of visual feedback
(Nijhawan and Kirschfeld, 2003)
QuickTime™ and a
Photo - JPEG decompressor
are needed to see this picture.
The outcome
sensed position
of rod
Flash-lag Effect for Voluntary Limb Movement
Flash-lag Effect (cm)
12
10
8
6
4
2
0
PM
RJ
ME
Observer
MS
Motor Flash-lag anisotropy
Motor task
fixation point
Motor task
fixation point
Motor task
fixation point
Motor task
fixation point
Vision task
Flash
fixation point
Flash
Vision task
fixation point
Vision task
fixation point
Vision task
Flash-lag effect (cm)
Flash-lag effect (cm)
Motor task
12
12
10
10
8
6
4
2
Movement
toward
fixation
Movement
away from
fixation
8
6
4
2
Movement
toward
fixation
Movement
away from
fixation
vision + touch neurons in area F4 of premotor cortex
--Graziano, Hu and Gross, 1997
The visual receptive fields are anchored to a given body part,
and move with the body part irrespective of eye position
Many neurons in area F5 of premotor cortex respond when the
animal performs a particular action, and to visual stimulation
in the absence of overt action.
-- Rizzolati, Fadiga, Gallese and Fogassi, 1995
The intermediate system “controls” the kinematics
of limb movements and of “external” visual objects.
Intermediate net
Limb movement
Peripheral
input
Object motion
Peripheral
input