Behaviour and the Brain 2:
Cognitive Neuroscience
Johannes Fahrenfort
[email protected]
Medical Faculty, room B563
1
Structure of this course
• Full-time course in period 6
• Exam is in the 4th week
• Book:
“Brain and behavior - a cognitive
neuroscience perspective”
• 3 tutorial meetings, obligatory
• Course runs 3 weeks:
‣ 3 lectures a week, 1 tutorial a week
‣ 8 lectures in total
+ Q&A session at the end
2
Schedule
When in doubt, consult https://rooster.vu.nl
3
Study materials
• Lectures roughly follow the book, but
lectures contain extra content
• Exam material is the book + lectures:
chapters 1-6 & 8–9 as well as the
contents from the lectures
• Exam: 40 multiple choice questions
4
Tutorial meetings
• Learn to reason about brain mechanisms
• Three tutorial meetings:
1. Identify lumping and/or splitting errors
2. Presenting on neural mechanisms (choose topic)
Re ecting on the studied material, implications
3. Presenting on neural mechanisms (choose topic)
Inventing a “brain-upgrade” (pros, cons, ethics)
• Writing a short summary report about
your presentation
5
fl
Tutorial meetings
• Tutorials are obligatory
• Missing at most 1 tutorial beyond your
control (funeral, wedding, illness): you need
to ask for permission and tutor will give
you a (rather big) replacement assignment
• Missing more than one tutorial: you cannot
pass the course
6
Rules of engagement
• Me:
‣ provide (hopefully) cool and interesting lectures
• You:
‣ keep up, keep up, read and learn the required chapters
‣ prepare for tutorial meetings
‣ coming unprepared to a tutorial may be counted as
missing a tutorial
7
Lecture recordings
• Lectures are on-campus, thanks for coming!
• Recordings from last year will remain on
Canvas but will not be updated
• Any updates I make to the lectures will not
be re ected in the recordings, but are still
exam material
8
fl
Final grade
• 75% is the exam grade (individual)
• 25% is the grade from the tutorials
(group work)
‣ 50% of the tutorial grade is the presentation
‣ 50% of the tutorial grade is your summary report
• Both exam and tutorial grade need to be
suf cient. Grades only count in current year.
9
fi
Lecture contents
1. Introduction: explaining the brain, modes of explanation
(Chapter 1 & Craver on Canvas optional reading)
2. History, neurons and action potentials (Chapter 3)
3. Architecture of brain & nervous system (Chapter 2)
4. Methods to investigate the brain (Chapter 1)
5. Vision (Chapter 5 & 6)
6. Perceptual organization and binding (Chapter 5 & 6)
7. Neural plasticity and memory (Chapter 4 & 9)
8. Attention & consciousness (Chapter 8)
9. Q&A
10
Overview of today
• The mind-body problem
• The modularity of the brain debate: speci ed
modules vs general purpose device?
• Explaining the brain:
‣ Box arrow models / functional explanations
‣ Reductionism
‣ Mechanistic explanations
11
fi
Cognitive Neuroscience
Mixture of:
• psychology (concepts, functions, cognition)
• philosophy (how can mental stuff be physical)
• biology (the brain is a biological machine)
• chemistry (neurotransmission)
• computer science / AI
(how does the brain compute / DL / LLMs)
12
What is the relationship between
thoughts/behavior and the brain?
univ. centre
response
stimulus
?
13
What is the relationship between
thoughts/behavior and the brain?
univ. centre
perception (seeing)
response
memory (recognizing)
stimulus decision making (thinking)
responding
14
What is the relationship between
thoughts/behavior and the brain?
univ. centre
response
“BLACK BOX”
stimulus
15
What is in the black box?
• Many modules / processes / subprocesses /
mechanisms involved
• Can we nd one to one mappings between
mechanisms in the brain and psychology
(mental ‘stuff ’)?
16
fi
The mind-body problem
• How can the body have a causal in uence
on the mind (and/or vice versa)?
Digesting Duck
• René Descartes (1596-1650)
‣ The body works as a machine
‣ Animals have no soul
‣ Reductionism / mechanistic thinking
‣ The soul causes thought
(situated in the pineal gland)
‣ Dualism
fl
The mind-body problem
Is the mind an epiphenomenon?
General consensus: the mind is
what the brain does (physicalism)
19
Not a new idea, early neuroscience:
Phrenology
• Franz Joseph Gall (1758-1828)
• George Combe (1788 –1858)
Phineas P. Gage (1823–1860)
• Worked in railroad construction
• Personality change
after brain injury
• Actual personality changes
appear grossly overstated
Modularity of the brain?
• Or is the brain a homogenous general
purpose machine?
• Or do distinct regions execute distinct
functions?
• Phrenology says modular!
• Chomsky (language) and
Fodor (modularity of mind)
22
The features of a module
(according to Fodor): n a
n ’ i
ti
c gu o a ge
1. Modules are domain-speci c ‘f u n n
o a g l a
i t
n mm i n
k
2. Their operation is mandatory A rogra
p
3. They are informationally encapsulated
23
fi
The brain as a tabula rasa
(= empty slate)?
• Empiricists (such as David Hume):
the mind at birth is a “tabula rasa”
• Nature vs nurture debate
• The idea of modularity is closely linked to
that of innate knowledge
24
Modularity of the brain
• Many examples: language, vision, maps,
pathways etc
• That’s good, it helps us study the brain!
• Hopefully, the organization of the brain reveals
something about the architecture of the mind
• The interesting question is not where modules
are, but what they do and how they map onto
mind and behavior
25
What is an ‘explanation’ of a
function or (mental) phenomenon?
n ?
ep tio n ?
y ? e r c
en tio
m or u a l p A tt
M e Vi s
s ?
e ? s n es
u a g i ou n ?
g ns c tio
Lan Co Em
o
26
BREAK
27
What is an ‘explanation’ of a
function or (mental) phenomenon?
• Box-arrow cognition models (functional)
• Reductionist explanations
• Mechanistic explanations
28
Cognitive revolution (>1950)
Box-arrow models of cognition
29
Cognitive revolution (>1950)
Box-arrow models of cognition
30
Cognitive revolution (>1950)
Box-arrow models of cognition
31
Cognitive revolution (>1950)
Box-arrow models of cognition
32
Cognitive revolution (>1950)
Box-arrow models of cognition
33
Box-arrow explanations
• Apply some sort of functional analysis
(asking why?), often use mental terms in the
explanatory boxes
• ‘How possibly?’, but not ‘how actually?’
Engineers: “I don’t understand it if I can’t build it!"
• Box models are often underdetermined
(multiple different box models can explain the
observed phenomenon)
34
What is an ‘explanation’ of a
function or (mental) phenomenon?
• Box-arrow cognition models (functional)
• Reductionist explanations
• Mechanistic explanations
35
Reductionist explanations
Identity statements between higher and lower levels
X reduces to Y
X is the sum of the parts of Y
1.Anderson, P. W. More Is Different - Science 177, 393- (1972). 36
Reductionist explanations
Identity statements between higher and lower levels
Bridge laws
X reduces to Y Oppenheim & Putnam (1958):
X is the sum of the parts of Y The Unity of Science
1.Anderson, P. W. More Is Different - Science 177, 393- (1972). 37
Bridge laws?
Parietal cortex
V5/MT
1m Brain
V1 V3
lgn V2
V4
eye Temporal cortex
10 cm Systems
1 cm Maps
1 mm Networks
?
100 µm Neurons
1 µm Synapses
1Å Molecules
Churchland & Sejnowski (1992) 38
What is an ‘explanation’ of a
function or (mental) phenomenon?
• Box-arrow cognition models (functional)
• Reductionist explanations
• Mechanistic explanations
39
Mechanistic explanation
“A mechanism for a
phenomenon consists
of entities and
activities organized in
such a way that they
are responsible for
the phenomenon.”
The whole is more than
the sum of its parts
40
The whole is more than
the sum of its parts
(= nonreductive)
41
The whole is more than The whole is the same as
the sum of its parts the sum of its parts
(= nonreductive) (= reductive)
42
Mechanistic explanation
Emergent properties:
whole more than sum of parts
but no “spooky emergence”!
Memory
LTP, hippocampus,
grid cells etc etc
43
Mechanistic explanation
Emergent properties:
whole more than sum of parts
but no “spooky emergence”!
Multiple levels
44
Example: oxygen metabolism
LUNGS
Valves
Deoxygenated blood
Oxygenated blood
Organs
Valves
BODY
Muscles Oxygen ATP
Nutrients
Mitochondria
Catabolism
Molecules
Energy
45
Explanatory power of mechanisms
LUNGS
Valves
Deoxygenated blood
Mechanistic
Oxygenated blood
Organs Levels
Valves
BODY
Muscles Oxygen ATP
Nutrients
Mitochondria
Catabolism
Molecules
Energy
46
Explanatory power of mechanisms
LUNGS
Valves
Deoxygenated blood
Oxygenated blood
Organs
potentially reciprocal
(circular) interlevel
Valves
dependence!
BODY
Oxygen Nutrients
Muscles ATP
“Where” is
Mitochondria
Catabolism
metabolism?
“Where” is
Molecules
Energy
the mind?
47
Computer analogy
Where in a computer is
the ‘word processor’?
?
48
What level of description matters?
• Phenomena may not have a “fundamental”
level of explanation, but require a multi-
level understanding
• Some levels can be so far removed from
the phenomenon that it is hard to see how?
they are related (despite being essential)
49
How to identify
mechanisms?
50
Etiological causal relevance
Does not identify a mechanism….
Detect effect
Manipulation
51
How to identify mechanisms:
establishing constitutive relevance
Memory
LTP, hippocampus,
grid cells etc etc
52
How to identify mechanisms:
establishing constitutive relevance
Top-down
experiment
Memory is impaired Detect effect Manipulation Memorize words
Connections
Interfere with LTP Manipulation Detect effect get strengthened
Bottom-up through LTP
experiment
53
Mutual manipulability
measure Top-down
experiment
manipulate
behavior/ Detect effect Manipulation behavior/task
mental mental
manipulate Manipulation Detect effect measure
lesion/ Bottom-up correlation
stimulation experiment structure/connect.
54
How to investigate brain mechanisms
• Correlational methods • Stimulation methods
‣ fMRI ‣ Electrophysiology ‣ TMS
‣ MEG/EEG ‣ Brain stimulation (Pen eld)
‣ eCog ‣ tDCS
• Connect. / struct. methods • Lesion/inhibition methods
‣ DTI ‣ Lesions ‣ Cooling
‣ VBM ‣ Strokes
‣ Tracers ‣ Tumors
55
More about these in lecture 4
fi
What is the function under
investigation?
• Does a given (mental) phenomenon (or
cognitive function) map onto a distinct brain
mechanism?
• Phenomena such as motion detection, depth
perception, working memory, change blindness,
and pitch perception may not t onto a single
distinct mechanistic process or structure of
the brain
56
fi
Two types of errors
• Lumping errors describing two or more
mechanisms in the brain as a single cognitive function
‣ “Memory” is not one single process in the brain
(as it is in the computer)
‣ “Consciousness” may not be one thing
“We have now come to believe that memory is not a single or
‣ “Attention”
unitary faculty of themay notasbewas
mind, one thing
long assumed. Instead, it is
•composed
Thingsofmaya variety of distinctasand
be treated dissociable
distinct processes
while and to
referring
systems. Each system depends on a particular constellation of
the same thing:
networks in the brain that involve different neural structures, each
of which plays a highly
‣ “Attention” mayspecialized
share the role
same within the system.”
mechanism with “working
memory” etc Daniel Schacter
57
Two types of errors
• Lumping errors describing two or more
mechanisms in the brain as a single cognitive
function
‣ “Memory” may not be one thing
• Splitting errors: describing a single mechanism
in the brain as multiple cognitive functions
‣ “Attention” may be constituted by the same
mechanism as “Imagery” as “Working memory” etc
58
Filler terms often hide failures
of real understanding
• Activate • Inform
• Cause • Inhibit
• Control • Modulate
• Encode • Process
• Excite • Recognize
• Filter • Represent
r …
• Generate fo • Regulate
sm
• In uence a ni • Store
c h
m e
A 59
fl
What is an ‘explanation’ of a
function or (mental) phenomenon?
• Box-arrow cognition models (functional)
Try to answer the ‘why’ question
• Reductionist explanations
Require hard-to-come-by bridge laws
• Mechanistic explanations
Try to answer the ‘how’ question, use mutual
manipulability to determine constitutive relevance
60
Preparation tutorial 1:
• Think about cognitive functions or brain
mechanisms that you think (could be) subject to
splitting or lumping errors.
• Find one or more articles about these functions
or brain mechanisms (can also be wikipedia).
• Can you identify lumping and/or splitting errors?
61
Questions
62