Action CourseManual
Action CourseManual
Action CourseManual
Period 5
Action
Preface
The goal of the third year Bachelor’s course Action is to increase your understanding of the
cognitive and neural basis of human actions and the decision-making processes surrounding
them. We will highlight the close connections between motor cognition and higher order
decision-making. In this course we will build from the most basic levels of description to
highly complex and brain-wide systems. We start with studying how the cerebral cortex is
organised for generating actions in response to the environment, how the generation of
action is conditional upon learned rules as well as upon goals that we have set for ourselves,
and finally how emotional considerations enter our decision making. In this way we hope to
elucidate how the brain is organised for generating meaningful goal-oriented behaviour.
We hope that this course will provide you an inspiring learning experience!
Contents
Preface ..................................................................................................................... 2
Contents ................................................................................................................... 3
A. General Information ............................................................................................... 5
1. The Education Office and www.askpsy.nl ................................................................ 5
2. Regulations, including Code of Conduct & Education and Examination Regulations ....... 5
3. Attendance ......................................................................................................... 6
B. Course Information................................................................................................. 7
1. Course planning group ......................................................................................... 7
2. Course description ............................................................................................... 9
3. Intended learning outcomes ................................................................................ 11
4. Course schedule ................................................................................................ 12
5. Work format ..................................................................................................... 13
5.1. Education group meetings ............................................................................. 13
5.2. Lectures ...................................................................................................... 15
5.3. Literature .................................................................................................... 15
5.4. Accompanying practicals ............................................................................... 16
5.5. Repeat students only .................................................................................... 17
5.6. Exam .......................................................................................................... 17
6. Some practical guidelines. .................................................................................. 18
Getting started in the first meeting ....................................................................... 18
The structure of a typical education group meeting. ................................................ 19
Guidelines for doing the work at home. ................................................................. 21
C. Tasks .................................................................................................................. 22
Problem 0: Sneak preview. ..................................................................................... 22
Problem 1: Cortical control of voluntary movement.................................................... 23
Problem 2: Cognition. ............................................................................................ 25
Problem 3: The basal ganglia (again). ...................................................................... 27
Problem 4: The reward system. ............................................................................... 29
Problem 5: Success, failure, and the cerebral cortex. ................................................. 31
Problem 6: Outcome monitoring .............................................................................. 33
Problem 7: Adding affect. ....................................................................................... 35
Problem 8: Trying to make it work. .......................................................................... 37
D. Literature ............................................................................................................ 38
Task 1: Cortical control of voluntary movement. ....................................................... 38
Core Literature ................................................................................................... 39
Further Reading.................................................................................................. 40
Task 2: Cognition. ................................................................................................. 41
Core Literature ................................................................................................... 41
Further Reading.................................................................................................. 42
Task 3: The basal ganglia again. ............................................................................. 44
Core Literature ................................................................................................... 44
Further Reading.................................................................................................. 46
Task 4: Reward system .......................................................................................... 47
Core Literature ................................................................................................... 47
Further Reading.................................................................................................. 48
Task 5: Success, failure, and the cerebral cortex. ...................................................... 50
Core Literature ................................................................................................... 51
Further Reading.................................................................................................. 52
Task 6: Error monitoring. ....................................................................................... 53
Core Literature ................................................................................................... 53
Further Reading.................................................................................................. 54
Task 7: Adding affect. ............................................................................................ 56
Core Literature ................................................................................................... 56
Further Reading.................................................................................................. 57
Task 8: Trying to make it work. .............................................................................. 58
Core Literature ................................................................................................... 58
Appendix: Anatomical nomenclature ........................................................................... 59
1. Prefrontal cortex subdivisions. ............................................................................. 59
1.1. Lateral view................................................................................................. 59
1.2. Medial view ................................................................................................. 59
1.3. Ventral view ................................................................................................ 60
2. Cortical motor areas........................................................................................... 61
2.1. Lateral view................................................................................................. 61
2.2. Medial view ................................................................................................. 62
3. Specific delineations ........................................................................................... 62
3.1. Cingulate gyrus versus cingulate cortex .......................................................... 62
3.2. Anterior and middle cingulate cortex .............................................................. 63
3.3. Macaque principal sulcus translates into human inferior frontal sulcus. ............... 64
4. References ........................................................................................................ 66
A. General Information
1. The Education Office and www.askpsy.nl
The Education Office is responsible for the practical organisation and coordination of all
the education related activities within the Faculty of Psychology and Neuroscience (FPN),
for example the schedules.
Askpsy.nl is the website for FPN faculty information, FAQs, and contact options.
The EER and R&R apply to all students of a study programme equally, and are valid for
one academic year only.
Note that the 2022/23 course might have COVID-19 related restrictions and based on
the situation we might need to adjust our way of education. Please contact your tutor,
course coordinator, and/or mentor whenever you have questions or remarks relate to
the teaching format.
FPN regards behaviour in compliance with its core values as being of great importance. A
Code of Conduct has been developed to ensure a good and productive study
environment and to avoid undesirable and unwanted situations. A link to the Code of
Conduct can be found at https://www.maastrichtuniversity.nl/about-
um/organisation/codes-conduct-regulations
5
3. Attendance
There is an attendance obligation with respect to the tutorial group meetings and/or the
practicals. The document entitled rules and regulations of the education program define
the attendance requirement for the tutorial meetings and/or the practical meetings.
Please see art. 5.8 EER and Art 7 Rules and Regulations Bachelor in Psychology for the
complete attendance rules (www.askpsy.nl/regulations). If you miss more than the
allowed number of meetings to obtain your attendance obligation, you will have to take
the course again next year. Note: In some courses, it is possible to join a meeting in
another tutorial group. Joining a meeting in another tutorial group will however not be
registered as part of your attendance obligation. You need to pass all attendance
requirements in your own tutorial group.
X If you miss a meeting, you may join this meeting in another tutorial group.
However, this is only allowed if the tutor of the group you would like to join,
agrees that you will join the tutorial group you are officially not registered for.
If you miss a meeting, you may not join this meeting in another tutorial group.
6
B. Course Information
1. Course planning group
7
Judith Peters & Joel Reithler, practical B
coordinators
Department of Cognitive Neuroscience,
OXF55, room 1.018
[email protected]
Tel: 043-388 2472
8
2. Course description
During the 1960s, the term neuroscience became widely used to indicate a field of study
in which several disciplines collaborated to investigate the organization and function of the
nervous system. Active areas today include molecular genetics and cell biology, focusing
especially on unravelling the basic mechanisms of developmental and neurodegenerative
disorders. In addition, research into the dynamic aspects of neural representations of the
‘higher’ functions, the subject of cognitive neuroscience, has also expanded considerably
over the last decades. In the first two years of the bachelor programme you were already
introduced to basic and fundamental neuroscience, in particular during the courses ‘Body
and Behaviour’ (PSY1023), ‘Sensation and Perception’ (PSY1026), ‘Learning and Memory’
(PSY1028) and ‘Functional Neuroanatomy’ (PSY2029). Almost every task in this course
requires reactivation of knowledge you acquired in these courses. This means you will
often need to refresh your knowledge and consult the literature used in those courses.
This course will go into more detail and is not meant as an introductory course in
neuroscience!
NOTE: It is very important to review the relevant topics that were treated in these
previous courses.
The core issue of this course is the neural organization that allows the brain to generate,
‘actions’ or ‘acting’. In the end, all information processing will, or will have to, result in an
interaction with the environment, in ‘behaving’. The effective execution of purposive acts
plays a central role in our daily lives. Every day we perform thousands of motor acts, from
coughing or an eye blink to producing speech, mostly without being aware of it. We often
tend to forget how infants have to struggle to learn ‘simple’ skills such as walking, eating,
picking things up and talking!
We know now quite a lot about the physiological characteristics of the elements involved
in movement, such as muscles and neurons. Much less is known, however, about how
these elements are coordinated to result in goal-oriented behaviour. Maybe even more
problematic is the facts that many of our actions are not just reactions triggered by
external stimuli but instead are ‘spontaneously’ or internally generated. Like all other
animal species, we do not perceive our external world as a neutral, indifferent
environment. We look at it from the perspective of acting: the environment consists of
things “to-act-upon”. And this is strongly coloured by our needs.
Perception directs our actions almost continuously. Sensory systems are indeed very
closely linked to the motor system; they have not evolved for perception per se but to be
able to respond to the changes they detect in the environment. For human voluntary acts,
such as grasping an object, the somatosensory system and the visual system are most
important. You may have studied these functional systems in a previous course. In the
present course we will only touch upon them to the extent that they provide information
relevant for action planning.
There are by and large three classes of movements: reflexes, rhythmic movement patterns
and voluntary movements. The first two classes have been discussed in previous courses.
The third class is the subject matter of the present course. Voluntary movements consist
of complex movement patterns that have often been acquired and to a large extent
9
automated but are still willed, goal-oriented, consciously planned actions. Primates,
especially humans, have developed the ability to perform accurate manipulations with
their hands. This has resulted in an opposable thumb and little finger a relatively large
representation of the fingers and mouth in the motor and somatosensory cortex. This
ability to manipulate is linked to another recent achievement in evolution: a direct, fast,
monosynaptic connection between the cerebral cortex and the spinal cord, the
corticospinal tract. Manipulation of objects is regarded as an important factor in the
evolution of apes and man (cf. expressions like ‘to grasp the meaning’ and ‘to get a grip
on’!).
Acting always means choosing between different possible actions; therefore, action is
closely connected to decision-making. Our brains are highly specialized devices, evolved
and trained for selecting the most optimal way of reaching our goals given our current
internal state and the external situation. Therefore, any meaningful behaviour starts with
the selection of a goal, the selection of the appropriate strategy to pursuit the goals and
finally the selection of action patterns to realize the goal. It appears that similar neural
substrates might be involved for the selection at all these levels, and that the basal ganglia
play a central role in all of them. Several parallel neural loops originating in the cerebral
cortex and passing through the basal ganglia play a crucial role in the organization of ‘pure’
motor acts, in cognitive acts (planning, thinking, decision making) and in socio-emotional
acts. Each circuit is clearly sequential but they are also organized in parallel. They appear
to be critically involved in a large number of neuropsychiatric developmental and
neurodegenerative disorders, such as Parkinson’s disease, schizophrenia and ADHD, etc.
. Strikingly, all these disorders are characterized by motor symptoms plus cognitive and
socio-emotional problems. From this perspective, any meaningful behaviour must be seen
as a unity of motivational, cognitive, attentional and motor components.
We hope you will enjoy the exploration of the fascinating organisation and functioning of
the brain systems and circuitry underlying our actions!
10
3. Intended learning outcomes
to describe the role of (sub-)cortical structures for movement and action selection (incl.
prefrontal cortex);
to explain the relation between movement and cognition, and translate this knowledge to
motor/cognitive impairment in patients with Parkinson’s disease;
to distinguish between choosing an option based on expected reward and choosing an action
to bring closer the chosen option. To relate actions and decisions to the moral and social
context (i.e. social cognition, moral decisions, altruistic, and cooperative behaviour).
11
4. Course schedule
Note the irregularities in the schedule due to the holidays. Always check the online
timetables/Canvas for updates and last minute changes!
The practical sessions will be scheduled individually: check your personal timetable.
Week 2
Tu 25 Apr 13:30-15:30 Lecture# Cognition (P. Stiers)
We 26 Apr 08:30-15:30 Education groups Task 2
16:00-18:00 Tutor hour*
Week 3
Tu 2 May 08:30-15:30 Education groups Task 3
16:00-18:00 Tutor hour*
Th 4 May 16:00-18:00 Lecture# Basal Ganglia (P. Heckman)
Week 4
We 10 May 08:30-15:30 Education groups Task 4
16:00-18:00 Tutor hour*
Week 5
Mo 15 May 08:30-15:30 Education groups Task 5
16:00-18:00 Tutor hour*
Tu 16 May 13:30-15:30 Lecture# Neuro-economics (A.
Vostroknutov)
Week 6
Tu 23 May 08:30-15:30 Education groups Task 6
16:00-18:00 Tutor hour*
Fr 26 May 08:30-15:30 Education groups Task 7
16:00-18:00 Tutor hour
Week 7
We 31 May 08:30-15:30 Education groups Task 8
16:00-18:00 Lecture# Questions & answers (P. Stiers)
Week 8
Mo 5 June 09:00-11:00 Exam MECC Expofoyer
# Lectures are in the Maastrichtzaal (UNS40 B0.647), except on May 4 in the Akenzaal (UNS40
B0.673)
* Tutor hour is online and only open to one group member per group (see explanation below).
12
5. Work format
The Education group meetings will take place without the presence of a tutor. We
understand that this may worry you and make you feel insecure. However, your past
experience with problem-based learning and with two previous tutorless courses have
prepared you well for the task ahead. The fact that you have become experts in team-
work and group discussions should make you confident that you will be able to come to
terms with the literature presented in this course. In addition, there will be several
mechanisms in place to support you and your group in this process: a list of elaborated
learning goals per task, an annotated reference list, an online tutor hour on every
education group meeting day at 16:00h, a discussion forum, and a Q&A lecture after the
last education group meeting. Last, but not least, a “flying” tutor will pay a short visit
during each group meeting. You can read more about each of these mechanisms below.
To help with the coverage of the literature in meetings without the help of a tutor, some
changes are made to the structure of and activities during the education group
meetings. The major change is that the typical starting phase, of pre-discussing
problems in order to arrive at a set of learning goals, is skipped and replaced by a set of
elaborated learning goals that are made available beforehand to all students (see
separate Canvas module). This should reduce the uncertainty of having to delineate
learning goals without the steering hand of a tutor. Learning goals play a crucial part in
problem-based learning because they help you to decide what is relevant and what not
in the listed literature. The learning goals give you a clear indication of what issues and
topics need to be covered, without going into all the details of every paper. There are six
learning goals per task, each emphasising a particular aspect of the problem and/or the
literature. Together they cover the important aspects of the literature. They are also the
basis for formulating the essay question for the exam.
The reference list contains a note on each item listed. The note aims to give you a sense
of why this entry is included in the list, i.e., what about this paper is relevant for
understanding the task, or how it fits in the story. Sometimes it also highlights some
aspect of the text that is particularly relevant and what parts – if any – can be read
superficially, or sometimes even ignored all together.
13
4 – Online tutor hour
Education groups are scheduled in three time slots starting at 8:30, 11:00 and 13:30. In
the fourth time slot, beginning at 16:00, an online Zoom meeting starts, in which a tutor
is available, who can help you address some issues that arose during your education
group meeting, and that you were not able to resolve. This tutor hour is open to only
one student per group. Every group will elect for every tutor hour anew one student who
will represent the group at the upcoming tutor hour. This is preferably a different student
for every tutor hour meeting.
At the end of each education group meeting, the group formulates questions for their
representative to present during the tutor hour. These questions are discussed by the
people present during the tutor hour, with the support of the tutor. Every member
makes sure that the questions prepared by their group are addressed and is responsible
for informing their group on the next education group meeting of the result of the
discussion. To increase the efficacy of the tutor hour, we ask each group to submit their
prepared questions beforehand (e-mail to [email protected]). The
person chairing the tutor hour will then have the opportunity to group, similar questions
and lead a discussion intended to solve the issues.
A discussion forum for themes and issues related to the course is available at the above
address. The entries in the forum are organized per task. As already mentioned under
(2) above, one of the entries is called “Introduction”, and gives a more general
justification for the task and the literature. Next, there are some entries from previous
years of the course, dealing with confusions and issues that students over the years
keep struggling with. Some of these are posted and answered by the course coordinator
in anticipation of problems that students may encounter. Everyone can add new posts,
related to the course content. You can see this forum as a continuation of the
discussions in the education group meetings. Consider the entries as a means of testing
your knowledge and understanding of the course content. Writing a reply to a post is a
means to test your knowledge and an opportunity for allowing others to correct your
views, if they are inaccurate or incomplete. Course planning groups members will only
commit themselves to correcting views that are not in line with the literature studied and
that may compromise understanding of other students.
One lecture is scheduled following the last education group meetings. In this Questions &
Answers meeting question regarding the course content can be discussed. Again, the
aim is not to teach literature content, but to resolve ambiguities and confusions that may
have arisen. To avoid chaos and ensure that all relevant questions will be addressed, any
questions you would like to pose have to be send in beforehand
([email protected]), so that there is time to prepare the meeting. You
don’t have to wait until the last day before the lecture to submit questions.
7 “Flying” tutor.
During each of the three time slots for education group meetings, two tutors will be
travelling from group to group, to check on how the group is doing and offer some help if
needed. Because each tutor has 7 groups to attend to, the visit to each group can only
be short. The tutor can therefore not help you to structure and direct the discussions of
the literature. However, they can help with directing and stimulating the group process.
If your group is doing well, the visit can be short. If not, the tutor will help you diagnose
what the problem is and help you in finding solutions. The tutor can also help the
discussion of specific topics or aspects of the literature, where the discussion got stuck.
Lastly, the tutor will be responsible for attendance registration.
14
All these elements will be put in place to support you in assimilating the core literature
proposed for this course. Make sure that you see how each text fits in with the
elaborated learning goals, what parts of the text address with aspect of the learning
goals, and try also to understand how each text relates to the other texts in the task,
and to topics touched upon in other tasks. Several topics re-emerge in several tasks, as
subsequent tasks built on content of previous tasks and sometimes reach back to things
discussed earlier. The problem “0” (see section C) was included to help you see this
larger picture and the internal relationships between tasks. So, use it at the end of every
meeting to situate what was discussed relative to this larger picture.
5.2. Lectures
The course includes six lectures, five of which are presented on campus. One lecture, on
the motor system, is a recorded lecture that is available on line (see Canvas). The last
lecture, on May 31, is a “questions and answers” lecture, for which students can pre-
submit questions.
The language used in the lectures is English. The lectures may address topics from the
education group meetings and introduce additional knowledge. They are intended to
provide a broader and integrative perspective on the topics discussed in the group
meetings. In addition, they elaborate on the significance of these topics in the broader
field of neuroscience and clinics. They intend to deepen and broaden students’
understanding of the core literature and therefore are highly recommended.
5.3. Literature
There is no general textbook (or textbook chapters) available that covers all learning
objectives sufficiently. The literature for each problem will therefore mainly consist of peer-
reviewed articles published in international scientific journals. Most of the literature can
be accessed by searching the internet (PubMed and Google Scholar) and UM journal
subscription. This free accessible content will be made available to students via a
compressed file with pdf-copies of the papers. It is advised to use these pdf-copies, as
some of them contain additional comments and marks on what is relevant for the course.
You find an annotated list of the literature for each problem mentioned in section D at the
end of this manual.
For all problems, you are expected to read the “Core literature”. Of course, that doesn’t
mean that you need to know every detail of every paper in the core literature by heart.
Every piece of core literature will contain sections that are more relevant and sections that
are less relevant for the learning goals. It is your task to extract the information that is
essential for the learning goals. It will be helpful to take into consideration that the papers
were selected to elucidate the currently most accepted answer of the neuroscience
community with regard to a specific problem. Therefore, try to think of the common,
integrative story across the different papers listed as core literature for a particular
problem. This integrative story will help you in deciding the relative importance of
statements and findings in the various papers. For instance, in studying papers describing
empirical studies, understanding how the paper fits in with the global story will help you
in assessing what aspects of the methods and results are important and what aspects are
not. Also, not every section in the discussion of the results will be relevant for the problem
presented in the course. Be aware that authors write papers for the scientific community,
and build their arguments in order to convince their colleagues or opponents – instead of
explaining their views to university students. Therefore, not everything they write is
relevant for our educational purposes. One more point of caution is warranted here.
Although the ideas expressed in the core literature are well stablished, other people in the
field do express dissonant views, giving other interpretations to the same or additional
data. By selecting the papers, we took into account coherence, level of acceptance, and
15
comprehensibility. If you decide to explore the wider literature, you have to be prepared
for some level of controversy and inconsistency, and for a range of shades and nuances
in opinion. Don’t let them confuse you too much at this point – only the facts and views
in the core literature will be the subject matter of your examination.
In addition to the core literature, for each problem there are also papers enlisted for further
reading. These papers can help you to resolve uncertainties left after reading the core
literature – the same is true for the content of the lectures. As explained above, every
paper has a particular angle and was written with a particular purpose in mind, which
sometimes makes papers less then optimal for educational purposes. In order to broaden
your reference frame and promote integration and assimilation of the core literature, the
“further reading” literature and lectures provide additional information.
The further reading literature (in contrast to the content of the lectures) is NOT exam
material: your knowledge of their content will not be directly probed in the exam. However,
indirectly they might improve your understanding of the core literature and place it in a
larger context, which allows you to see better the relationship between content discussed
in the different tasks. Also, in many cases they will look at the same content as the core
literature, but from a different perspective and with a different emphasis. This may help
you to better grasp information that was not so clear from reading the core literature.
Lastly, further reading papers can discuss fields of application of the core course content,
and in that way will promote your understanding of why we are studying these topics.
Taking notice of the content of the further reading literature will surely improve your
performance on the exam. This is why it is advised to divide for each problem the papers
listed as further reading literature among your education group members and to give a
short overview of their content after having discussed the corresponding core literature
topics. You can then asses for yourself whether reading one or other of these extra papers
will improve your understanding of the core literature topics.
IPN3155: https://curriculum.maastrichtuniversity.nl/meta/434143/practical-group-
decisions
IPN3156: https://curriculum.maastrichtuniversity.nl/meta/434533/practical-neuronal-
basis-decision-making
Please note that these practicals are only remotely related to the topics of the course:
both practicals and the course are about decision-making. Apart from this, the practicals
and the course are independent. This means that the intent, content, and end result of
the practicals is in no way a help in passing the Action course exam. The reason for
linking the practicals to this course is mainly administrative. In line with this
independence, the evaluation of the practicals is independent from the evaluation of the
course: one can pass one and fail the other or vice versa, and these independent
evaluations perseverate into next year.
However, because the practicals are each linked to one version of the course, you can
only obtain a pass for the whole package, and cash in the 5 ECTs associated with the
course, if you obtained a pass for the evaluation and attendance of the practical and the
course.
16
5.5. Repeat students only
The changes to the course, this year, mainly relate to the changed procedures for
education group meetings and for the exam. No changes were made to the course
content.
5.6. Exam
The course exam will be administered on site at the MECC with TestVision. It will consist
of a mixture of question types, with around 10 multiple choice questions, around 10
short open questions and 1 essay question. Each question type defines a separate part
of the exam, and each of the three parts will have the same 1/3rd weight in the final
exam score. The exam questions will be formulated in English or Dutch, depending on
the specific track you follow. All students are free to write their answer in English or
Dutch.
In this course we work with “core” literature, and this means that you are only expected
to know and explain the content of the papers or text parts listed as core literature. See
section above on “Literature” for some guidelines on how to use the “further reading”
literature and lectures to support your understanding of the core literature. Therefore,
the exam will only probe knowledge of the core course literature and more general
themes from the lectures that overlap with the core literature, but NOT of the practicals.
So, keep in mind that your final grade for the course (i.e., IPN-PSY-3012-A-B) will NOT
be dependent on the grade for your practical report (although you do need to pass for
the practical to obtain a “pass” for the course as a whole).
Therefore, to pass the course as a whole you need to fulfil the following requirements:
17
6. Some practical guidelines.
The first session is focused on getting your education group organized and going. Before
you start, we would like you to consider some ground rules in this course. The absence
of a tutor allows you to
take charge of your learning environment. For some of you, this introduces uncertainties
that might make you feel unsafe. For this reason, we are going to spend some time
thinking about how you can improve collaboration in your education group, and
communicate with planning group members.
You can start with getting acquainted with your team members. Exchange names, where
you are from, and share how you have experienced your education groups thus far.
Now that you have learned who is on your team, it is time to agree on how you will work
your way through this course and put the decisions regarding this work process in a
team charter.
It is a good starting point to discuss first your expectations and fears regarding the task
ahead. The more you can bring your worries and uncertainties in the open, the easier it
becomes to design strategies to counter them. Think in particular about your
experiences with the course Functional Neuroanatomy (which is more comparable to
Action) and the tutorless courses; in your 2nd year.
After having brainstormed about what approaches might help in lifting your worries, it is
time to make decisions.
It is very important that your group will function as a team. Research on team
performance shows that high quality team charters—written plans for how the team will
manage teamwork activities—are positively related to team performance (Mathieu &
Rapp, 1999). Drafting a team charter increases team members’ knowledge of each
others’ strengths and weaknesses, helps to create shared expectations, and facilitates
the establishment of effective group practices for dealing with high and poor
performance. In the team charter, please report as a group on team roles, expectations,
and processes.
The scribe can complete a team charter during this first meeting. A template of the team
charter can be found in Canvas under Modules. The scribe can collect and enter
information in the template. The team charter should be sent by e-mail to every member
of the group, and to the course coordinator ([email protected]).
Please mind the file name format, e.g. gr12charter.xlsx. By having the team charter
available, others can check its contents – including the teachers. The team leader is
expected to update the charter when mutations occur.
As you will see in the template, we expect you to divide the following roles:
18
1. Chair.
2. Scribe/Co-chair.
4. Tutor hour representative (who will represent the group at the tutor hour at 16:00
that day).
You can combine roles as you see fit, and you can alternate some, while keeping other
roles fixed. It is preferable, however, to rotate the role of group representative at the
tutor hour, as it requires extra commitment (representing the group in an additional
online meeting).
In the team charter, your group can also verbalize expectations regarding meeting
attendance, participation quality, what to do with free riders, etc. This is the part where
you, as a group, decide how the tutorless meetings will be organized and what
procedures will be followed. It is important that you address these issues overtly at the
start of your learning journey. You are free to shape the meetings as the group sees fit,
but make sure that all the relevant literature issues have been discussed and all the
learning goals have been reached. All members are expected to conform to the
agreements you make.
The assignment presented in problem 0 is intended to help you with integrating the
content of all the course problems into an integrated whole. The reason for this preview
of the course content is that many students in previous years reported to have come to
understand this integration only very late, after the meetings had finished and they were
already studying for the exam. They thought this was unfortunate, because they felt that
understanding this structure during the course would have helped them to process the
literature. Moreover, tutors in the course had the impression that many students never
came to see how the different course problems aim to convey an integrated perspective
on how the brain is organized for meaningful behavior. The aim of this pre-discussion of
problem 0 is therefore to make the integrated perspective on action and the brain
explicit, in the form of a schematic chart that highlights the different steps in selecting
meaningful actions. While we explicitly ask you to discuss the scheme in the first
meeting, it is important to go back to this scheme in every course meeting – either at
the start, or near the end of the meeting – and discuss where in the scheme the current
task/problem situates. We hope that this helps you to see why you are asked to study
this literature and what is relevant and what not in it.
During each education group meeting you try, in collaboration with your team, to
formulate and discuss answers to the learning goals, based on what you have read in the
literature. Time and mental energy is limited. So, keep your contributions short, clear
and to the point. Below, you find some suggestions on how your group might structure
the 120 minutes of the education group meetings. Remember, the absence of a tutor
provides you the opportunity to take responsibility for your own learning. This
responsibility implies that you can decide with your team what is the best way to
proceed during the meetings. The structure below is just a suggestion -- you are free to
depart from it.
19
1 Feedback from the tutor hour.
You start with revisiting some of the challenges from the previous meeting. The person
representing the group at the previous tutor hour gives a report on what was discussed
their and how this reflects on the remaining issues from the previous task. Other group
members can add to or further discuss these issues, based on new insight or information
from a lecture, or literature read, etc.
2 Literature round.
Get an overview of the sources that have been read and by whom. PBL is all about
collaborative learning. Avoid the situation in which one or two students have to share all
their understanding of a source. Also, don’t expect a contribution to be flawless. Be
critical about what is presented and how it relates to what you have read. Starting with
an overview of who read what might be useful to know who you can rely on to support
you.
As a warming up, take a look at the schematic chart of the content of the course
presented in Problem “0”, that you discussed in your first meeting. Now that you have
read the literature, discuss what region of the chart is the focus of the literature that you
read. Try to delineate which notes in the chart are covered and which are not. Try also to
establish whether parts of the chart region covered by the current literature overlaps
with regions that were the focus of previous tasks, or the current region relates to
previous tasks.
5 Literature discussion.
Going over all the literature in an understandable manner is not manageable, as not
everything in the papers is relevant. We advise you to organize the discussion by
learning goals. Go over the six learning goals one by one. Try to get a core
understanding of what you learned from reading the literature about the learning goal.
You can then try to add more specific findings from particular papers, lectures etc.
An effective discussion lets you get the maximum out of the time you have with your
group. Of key importance is a good, efficient style of reporting during the discussion.
Given that there sometimes is a lot of complex material, this can be a challenge.
Remember also that you only have some 15-20 minutes of time available per learning
goal. Make your life easier by preparing before the meeting, and help yourself and your
colleagues by sharing your information with them in a useful, yet brief way during the
meeting.
Here are some tips for preparing yourself before the meeting:
- Order and summarize the content per learning goal. Ask yourself for what learning goal
the content you are reading is important. Makes notes accordingly. After reading, ask
yourself what you can contribute for each learning goal.
- Work out what you want to say in the group meeting. Imagine you are giving a mini-
presentation on the material you read, just without PowerPoint. You do not want to
spew unsorted, uncommented facts read from the articles. Everyone can read these.
Focus on what the facts mean for the learning goal, and the task.
20
- Right down what remaining questions you have. You don’t understand something you
read? You cannot find the answer to some part of a learning goal? The group meeting is
the place to get you some answers.
6 Reflection on discussion.
Maybe after each learning goal, allow the “tutor hour representative” a moment to
evaluate how efficient the group was able to address the issues raised in the learning
goal: what you did manage to discuss, what aspects might need more attention, or left
you confused. See whether you are able to find some answers later on in the literature.
Alternatively, the tutor hour representative will write down the remaining issues for the
tutor hour at 16:00 on the same day. The Tutor hour representative is responsible for
sending in the remaining issues for further discussion in the tutor hour (via e-mail to
[email protected]).
At home, you do the heavy lifting: studying the listed literature and retrieve the
information that is relevant for the learning goals. Everyone is, of course, free to choose
their own approach to perform this task. However, here are some guidelines that may
help to improve the efficacy of the heavy work.
- Look at the “problem” for that task, as presented in section C of this course manual,
and try to answer the questions.
- Next, read the introduction to the task literature in annotated literature list in section D
of the course manual.
- Look at the learning goals before you start to read. In that way, you may already
recognize during the reading of the texts themes that are relevant for the learning
goals.
- Start reading the papers.
- First, read in the annotation to the reference list why the text is relevant.
- While reading, make structured notes, so that after reading you will have an overview
of the text’s content and structure. Also, write down themes that appear to be relevant
for the learning goals.
- After reading, go again through the learning goals and the literature introduction, and
see where the text relates to the learning goal topics and to the task as a whole.
- Make a short overview per learning goal of what aspect of which text is relevant for
each learning goal.
- Make notes of what in the texts you did not understand, so that you can ask for help on
these items, and so that at the end of the meeting you can easily check whether all
your questions have been answered.
21
C. Tasks
Imagine that you are at a supermarket, trying to decide whether to buy a pineapple or
bananas. In the diagram, you see a simplified flow-chart of the steps that take place
somewhere in your frontal lobe and that end in you picking up the bananas. Go through
the chart and try to explain what happens at each step (ignore for now the brain structures
mentioned on the left side, and focus on the central column in the chart).
[This chart is a road map for the course – come back to it during every meeting and try
to delineate to what part of the chart the literature you read relates to.]
22
Problem 1: Cortical control of voluntary movement.
In previous courses, you have had a look at all the CNS regions involved in the motor
control hierarchy, and how they support the spectrum from reflex to voluntary action.
This task explores the functional contribution of the cerebral cortex to motor control, and
more specifically the processes linking the perception of a particular stimulus to an action
A.
from
Patricia Garrido-Vásquez and Anna Schubö, 2014, Frontiers in Psychology.
What would happen in our brains when we are (visually) confronted with the object on
the left? Compare this with what happens when we see the object on the right?
Picture on the left was taken from https://www.pinterest.com/uxfails/ux-fails/, the one on the right from
http://www.satukyrolainen.com/affordances/
What kind of behaviour would you expect when people are confronted with the door and
the faucet depicted above? What (brain) processes would be needed before they can start
planning a movement towards them?
B.
There is a difference between reaching and grasping an object, versus using it as a ‘tool’.
23
These different ways of
handling objects can be
selectively impaired in
patients, e.g., with
ideomotor apraxia. In
normal persons, the Two
Action Systems model
suggests that functional
and neuroanatomic
distinctions between the
systems underlying
skilled use and grasping
could account for the
findings in the graph.
“One evening we took our patient, Mrs GP, to dinner with her family. We were discussing
the implication of her medical condition for her and her relatives, when, out of the blue
and much to her dismay, her left hand took some leftover fish-bones and put them into
her mouth (Della Sala et al., 1994). A little later, while she was begging it not to embarrass
her any more, her mischievous hand grabbed the ice-cream that her brother was licking.
Her right hand immediately intervened to put things in place and as a result of the fighting
the dessert dropped on the floor. She apologized profusely for this behaviour that she
attributed to her hand’s disobedience. Indeed she claimed that her hand had a mind of its
own and often did whatever ‘pleased it’.”
From: https://thepsychologist.bps.org.uk/volume-18/edition-10/anarchic-hand
24
Problem 2: Cognition.
In the previous problem, we saw how stimuli can be transformed into a motor act, and
how some stimuli are strongly associated with a particular motor act. Fortunately, most
people are capable of flexibly withholding or adjusting prepotent responses to certain
stimuli depending on whether the responses are appropriate or not in the current situation
– i.e., depending on whether they agree with the goals set by the current task rules.
A.
The Wisconsin card sorting test has been used to test the ability
of patients to flexibly shift rules. How would this work? What type
of patients might fail?
B.
The experimental task below has been used in monkeys.
25
Test 1
2800ms
C.
…
26
Problem 3: The basal ganglia (again).
A.
"Cats decorticated at birth move around gracefully and display goal-directed locomotion –
the uninformed observer would see no or little difference in movement pattern from that
of an animal with its cortex intact. Such decorticated cats can successfully search for food
and eat, and also use their forelimbs for exploratory movements to determine which path
to select from a platform to reach the floor". (Grillner ea, 2005, p. 365.2)
“Decortication” means that the entire cerebral cortex was surgically removed at birth – in
other words, the lesioned cats have no cerebral cortex. Yet, they move around and behave
almost indistinguishable from intact cats. It makes one wonder what exactly the neocortex
contributes to behaviour, and who is in charge in the brain and makes the decisions.
B.
It has long been controversial whether Parkinson’s disease is exclusively characterized by
motor symptoms or also by cognitive and emotional symptoms. Nowadays, the question
is not so much whether Parkinson-related cognitive and emotional changes exist, but
rather what they are and how they can be explained. Hayes and co-workers (1998)
performed a series of experiments in line with the “shifting hypothesis”. See the figure
below.
C.
Different parts of the basal ganglia serve different subregions of the prefrontal cortex.
The typical voluntary movement initiation problems in patients with Parkinson’s disease
27
relate to the motor domain, and are associated with dysfunction in the "motor loop"
through the basal ganglia, depicted in the left most part of the Figure below. Have a look
at the other loops depicted. How do these basal ganglia loops differ from the motor loop?
Based on what you know previous year’s courses, try to think of the symptoms that
emerge when these loops get dysfunctional as well?
28
Problem 4: The reward system.
The nucleus accumbens and the amygdala are both connected to the orbitofrontal cortex.
All three structures have to do with the processing of reward and punishment qualities
associated with stimuli or actions.
B.
In the 1990s Wolfgang Schultz started his search for neurons in the brain that signal
“reward”. He and his colleagues trained monkeys to press a lever to obtain fruit juice.
During training they measured the activity of dopamine neurons in the ventral midbrain
area of the monkeys brain and they found that some of these neurons showed a phasic
response (briefly increased their firing rate) when the juice was delivered. In a study using
different rewards (small or large amounts of juice) they found that the magnitude of the
phasic response corresponded to the magnitude of the reward. In another study the
reward would only be delivered when a light was on. When the light went on the monkey
knew he could press the lever and get a reward, but when the light was off pressing the
lever would not result in delivery of juice. Schultz found that after a few trials the neurons
no longer responded to the reward. Instead, the neurons showed a phasic response when
the light came on. Next Shultz and colleagues introduced uncertainty. They trained the
monkeys to discriminate between different lights each associated with a different
probability that the reward would be delivered. When the light predicted a 50% chance of
a reward the neurons responded to the light, but also to the reward. The neurons did not
respond to a reward that was predicted with 100% certainty.
C.
When regular consumers of alcohol and cannabis are exposed to marketing movies for
either alcohol or cannabis, their brain seems to show a response (compared to neutral
movies) that is comparable to when a reward is delivered (see Figure 2). It is as if watching
these movies is in itself rewarding for them. How is that possible? Try to explain this,
based on what you know about reward based learning and Schultz’ dopamine results
described above.
Surprisingly, when the same individuals were intoxicated with their preferred drug the
same marketing movies invoked a much weaker response. How can you explain this?
You need to take into account the effect of drug intake on dopamine, to understand this.
29
Figure 2: The upper panel shows BOLD activation (red) following cannabis and alcohol
marketing exposure (versus neutral) collapsed over all three groups while sober. The lower
panel shows the degree to which alcohol and cannabis intoxication reduces (blue) the BOLD
response to marketing exposure (compared to placebo).
30
Problem 5: Success, failure, and the cerebral cortex.
A.
Some of you may have seen the motion picture
"Dirty Mind", by Pieter Van Hees.
The picture tells the story of Diego Willems, a friendly and shy person who works as a
technician in his brother's not very successful stunt team. After an accident he wakes up
in the hospital and feels himself a different man. He has been transformed into Tony T. a
smooth-talking no-holds-barred stuntman. Tony's new personality changes his life
completely. He is extravert, always cheerful and goes for every challenge. He feels better
than ever and loathes the old timid Diego. Tony's lack of inhibition slowly but certainly
makes others turn against him.
Is this story mere fiction, or is there some reality to it? How do you know our personalities
are stored in the brain? Is it at a particular location that, when damaged, might give us a
new or altered personality? Or is it represented wide spread and therefore resistant to
brain damage? How does our personality, who we are, relate to action and decision
making?
B.
Antonio Damasio and co-workers established that people with damage to a specific part
of the prefrontal cortex have the tendency to make the wrong decisions in their lives. To
objectively measure this deficiency they designed the now well-known Iowa Gambling
Task. The task is illustrated in the Figure below. The task is still widely used in clinical and
research settings, because it illustrates accurately what problem these patients encounter.
Have a look at the Figure to grasp the concept of the task. Healthy people prefer to draw
cards from one of the two stacks of cards. Why is that? The patients, on the other hand,
have a preference for the other stack. What does this tell you about their behavioural
problem? Try to formulate hypotheses about the functional disability underlying this
problem.
31
C.
We choose between the options that are available in a particular situation based on what
we like most. One could say that such a decision is based on the value that each option or
object has for us. Such values are not stable but can change over time, sometimes quite
fast. For instance, at a dessert buffet, you might first pick your favourite dessert. But after
having eaten it you may choose for another dessert. The eating probably hasn’t changed
the sensory representation of the dessert, but at least something has happened with its
value, since on your second go, you preferred another dessert. How do you think the
brain knows that your preference changed, or when you keep eating your favourite
dessert, that at a certain point you had enough of this and want something else? And what
if you have to decide between very different things –for instance, whether you would prefer
a new car or a vacation in Hawaii.
32
Problem 6: Outcome monitoring
A.
33
C.
34
Problem 7: Adding affect.
A. Prospect theory
Imagine that you have to make a choice between participating in two types of lotteries
(without any initial investment). Circle the option you would choose:
a. You have a 50 % chance to win 1000 euro and a 50 % chance to win nothing
b. You get 450 euro for sure
And what would you choose if you had the following two options?
a. You have a 50 % chance to lose 1000 euro and a 50 % chance to lose
nothing
b. You lose 450 euro for sure
B.
In a simple reversal learning task participants first learned which of two stimuli was
rewarded (70% of the time; 25 cent gain) and which yielded a punishment (25 cent loss).
After every four correct choices the stimulus-reward contingency was reversed.
fMRI analysis aimed to visualize the brain activity that made participants shift their
strategy, by comparing the response to the feedback on trials just preceding a trial in
which a response scheme switch had occurred with the feedback activity on trials that did
not precede a switch. The feedback that induced a response scheme switch in the next
35
trial was significantly higher in several regions. One of them was anterior MCC (cf. Figure
A, left panel, white arrow). From what you learned in the previous task this should not be
a surprise. Try to explain this. Another region was lateral OFC (not shown). Try to explain
this with what you’ve discussed in task 5. As a test of how deep your understanding goes,
try to explain the effect observed in the ventromedial PFC (Figure B, white arrow). Here
the BOLD signal was higher in the opposite contrast – i.e., after feedback that preceded
non-switch trials compared to switch trials. To explain this, you need to take into account
that vmPFC is located in the default mode region of the medial PFC.
Hampton, A.N., Adolphs, R., Tyszka, J.M., O’Doherty, J.P. (2007). Contributions of the amygdala to
reward expectancy and choice signals in human prefrontal cortex. Neuron 55:545-555. Figure 4.
The really interesting observation in this study was that in two patients with bilateral
amygdala loss these activation differences did not occur (in none of the three locations
mentioned above). What can you learn from this regarding what amygdala contributes to
making the decision to switch strategy?
36
Problem 8: Trying to make it work .
A.
“You’re going to meet with a friend in a café, just before lunch. You arrive at the café
somewhat early. You hesitate to go in and consider going for a short walk instead.
However, taking into account the cold and greyish weather and the hour approaching lunch
time, you decide to enter. The place is already crowded, even at this pre-lunch time. There
are tables with chairs to your left, but none empty. So you head towards the counter on
your right. Before you reach the one vacant bar chair there, a tall person with a funny
little head, coming from the left, sits on it and starts talking to a woman to his left. You
hold your step,
hesitant, look
around, to the left,
the right, until you
decide to take
position near the
right end of the
counter, close to
the window, to wait
until a place to sit
becomes
available.”
This example
description of
ordinary goal-
directed behaviour
brings together all
the elements and
topics that have
been discussed in
the current course.
The different tasks
of the course
pieced apart the https://marybloch.wordpress.com/tag/french/
underlying process
into subcomponents and studied each more closely. In order to help our understanding of
meaningful behaviour, we must also consider how the separate components integrate to
form a smooth-running clock work.
In this last pre-discussion you must have a go at it. Use the black board or electronic board
and together draw a flow chart of what happens in the brain of this person during the
events described. Use the chart from the sneak preview as a guide. Place on one side of
the board what is there in the environment and on the other side what is relevant from
the past experience of that person. Then fill the empty space in between with the
information flow, decision steps and brain structures that take place.
37
D. Literature
[A course on (goal-directed) action must start with the apparatus allowing us to generate
actions that bring about change in the environment. Therefore, the first problem is about
the global functional organization of the voluntary action system. We will not repeat the
discussion of the spinal reflexes and pattern generators, that were the subject of the first
year course “Body and Behaviour”. Moreover, we will not go into the nature of
representations in M1, the combination of cortical and cerebellar signals that make up the
output of M1, or the descending motor paths by which this output steers the muscles of
the body, because this was discussed in the second year course “Functional
Neuroanatomy”. Instead, we will look into how the cerebral cortex is organized to allow
execution of voluntary actions. This is not just the story of premotor and motor cortex,
because these areas are only involved in the last re-active step in a long sequence of
events that starts with (1) detecting opportunities in the environment and (2) keeping “in
mind” goals and expectations of the organism. So, voluntary action starts with the
posterior (or receptive) brain, and particularly the dorsal visual pathway (See the course
“From Sensation to Perception”), and it somehow requires taking into account the
organism’s past experience in this environment, as well as its current state of needs and
desires. In the first paper that you will read, Wong et al., (2015) provide a conceptual
framework for this integrated view on voluntary action. The authors take as their central
concept the “motor goal”, i.e., what you want to achieve in the environment. They then
discuss, on the one hand, the processes that are involved in establishing the motor goal
(what they call the “what” processes), and on the other hand, the processes involved in
achieving the motor goal with a specific motor output (what they call the “how” processes).
This course (as different from this task) will particularly elaborate on the former processes,
the decisions that are involved in selecting a motor goal.
The dorsal visual stream is the part of the system involved in processing and presenting
possible targets for action. This stream is the focus of the paper by Binkofski and Buxbaum
(2013). The stream was first described from the perspective of visual perception, and from
that perspective it is often referred to as the “how” pathway, because it is involved in
visuomotor integration. Note that this “how” perception pathway is on the “what”-side of
the motor goal in the framework of Wong et al. (2015). The pathway that represents
possible action targets terminates in the premotor cortex, where possible motor patterns
are stored. This makes it comprehensive why from the perspective of behaviour, possible
action targets are not merely stimuli, but are actually affordances for possible actions
(more about this in task 8). This also justifies calling these stimuli "action targets" instead
of just stimuli in space. In this exogenous information stream, a division is made between
a dorsal path involved in visually guided movements and a ventral path involved in skilled
tool-related movements (Binkofski, & Buxbaum, 2013).
However, what counts as a possible target for action does not only depend on what is
available in the environment, but also on what task the organism is engaged in. The task-
related issues are what we call the “cognitive” aspect of action, and you will learn more
about this aspect in the next two problems (tasks 2 and 3). Note that an action target is
not the same as the motor goal. Action targets are the sensory objects that guide our
actions (e.g., a red or a green light at a particular location). The motor goal is the change
in the environment that one intends to bring about (e.g., pressing the left button, or
stopping the car we are driving). Action targets need to be translated into motor goals, by
cognition. In addition to motor goals triggered by exogenous targets, the brain also can
select endogenously derived motor goals, which are available via the medial prefrontal
cortex. The final decision on which motor goal will be actually pursued via motor output
(and which goals will be suppressed) resides with areas in the midsection of the medial
(pre)frontal cortex (preSMA/SMA). This action selection centre is discussed in Passingham
et al. (2010), while Sumner et al. (2007) describe a patient-based study providing
empirical evidence for their hypothesis.]
38
Core Literature
Wong, A.L., Haith, A.M., Krakauer, J.W. (2015). Motor planning. Neuroscientist 21, 385-
398.
[Overview of the processes involved in goal-oriented (voluntary) action. Since an
action always aims at achieving a particular end-goal, the authors chose as the
central focus of the motor process the motor goal. This allows them to distinguish
between processes involved in selecting the motor goal and processes involved in
reaching that goal. The former processes take place in the dorsal attention system
with in the posterior brain the parietal attention regions in the intraparietal sulcus,
and in the anterior brain the Frontal Eye Fields (FEF) in the frontal lobe. The
posterior side is perceptual in nature, involved in representing possible action-
relevant stimuli in the environment. The FEF regions are involved in representing
the motor goal selected on behalf of these stimuli. In between these two types of
representations, somehow, the visual targets are translated into the motor goal.
This translation takes into account the appropriate task rules that stipulate which
object or target is most appropriate (or most valued->tasks 5, 7, and also 6) given
the current situation. Such conditional processes are implemented in the (lateral)
prefrontal cortex (the focus of task 2) and the medial motor cortex (preSMA/SMA),
where the optimal response is selected, as discussed in Passingham, Bengtsson, &
Lau, 2010), Sumner et al. (2007), and the task on error monitoring.
The processes on the execution side involve the selection of an appropriate abstract
motor program that will yield the desired end-effector movement (in ventral
premotor cortex), the detailing of the motor commands to perform the action [in
dorsal premotor and primary motor cortex] and, if necessary, the specification of
an accurate action trajectory [possibly involving the parietal cortex]. Motor
selection and movement specification rapidly translate a motor goal into a desired
movement. This is accomplished via the generation of a “feedback control policy”
(comparable to the inverse models, discussed in Functional Neuroanatomy), which
determines the complete movement trajectory given the current state of the limb,
the desired endpoint (e.g., the motor target) and the cost of bridging these two.
You can skip the last two sections, starting in page 393, left column “Motor planning
and reaction time”.]
Binkofski F, Buxbaum L (2013). Two action systems in the human brain. Brain and
language, 127(2), 222-229.
[Elaborates on the distinction between ventral and dorsal premotor cortex (cf.
Chouinard & Paus (2006; see 2nd year course “Functional Neuroanatomy” and
“further reading”) and Wong et al. (2015)). In this paper these subregions become
segregated endpoints of two visuomotor pathways, the grasp-system for visual
guidance of actions and the (tool) use system for skilled and tool-related actions.
Therefore, this paper describes the parietal bottom-up, exogenous, or
“affordances” part of the motor system.]
Passingham, R.E., Bengtsson, S.L., Lau, H.C. (2010). Medial frontal cortex: From self-
generated action to reflection on one's own performance. Trends in Cognitive
Sciences, 14, 16-21.
[This paper elaborates on the role of medial PFC in the generation of action, with a
particular plea for the role of pre-SMA as the critical substrate for the generation
of voluntary actions. They review this hypothesis in the light of new evidence and
argue that this hypothesis still holds. They further argue that the medial prefrontal
cortex is crucially involved in self-generated action and self-reflection. With this
they further substantiate the view (already in Gazzaniga et al., 2009) of a parietal
external and medial-prefrontal internal source of action control, which are kept in
balance from the pre-SMA. This
39
also expands on outcome monitoring and Rushworth et al. (2004)’s division of
labour between pre-SMA and CMA (cf. task 6).]”
Sumner, P., Nachev, P., Morris, P., Peters, A.M., Jackson, S.R., Kennard, C., Masud Husain,
M. (2007). Human medial frontal cortex mediates unconscious inhibition of
voluntary action. Neuron 54, 697-711.
[This paper describes an empirical study on patients with focal brain lesions to test
the causal contribution of SMA and supplementary eye fields (SEF) to the control
of action. They show that the response delaying effect of a subconscious action
prime on subsequent performance of the primed action – which can be seen as an
automatic (because unconscious) inhibition of possible actions afforded by the
environmental cues – is lost for manual actions or eye movements following
confined lesions to SMA or SEF, respectively. They see this automatic inhibition
function of SMA/SEF as an essential contribution to voluntary behaviour by freeing
person from the demanding affordances of the environment, while at the same
time allowing the environment to proposed potential lines of action. The study
therefore presents empirical evidence in support of the hypothesis argued for in
Passingham et al. (2010) that the medial motor areas are specialized for the
selection and initiation of voluntary behaviour.]
Further Reading
Grillner, S., Hellgren, J., Ménard, A., Saitoh, K., Wikström, M.A. (2005). Mechanisms for selection of basic motor
programs ‐roles for the striatum and pallidum. Trends in Neurosciences, 28(7), 364‐370.
[This is part of the core literature in Functional Neuroanatomy (IPN/PSY2029) task 2, on basal ganglia.
However, it gives an integrated view of how the brain stem contains a series of innately wired (instinctive)
motor patterns (e.g., locomotion, orienting, but also emotional expressions) that make up the
behavioural repertoire of an animal, and explains how basal ganglia together with cerebral cortex control
the execution of these programs. So, it allows bridging the two courses.]
Chouinard P & Paus T (2006). The Primary Motor and Premotor Areas of the Human. Cerebral Cortex. The
Neuroscientist 12(2), 143-152.
[This is also part of the core literature in Functional Neuroanatomy (2029), more exactly in task 1 on
organisation of the motor system. For our purpose it highlights the differentiation between M1 and
Premotor area (PMA), and within PMA is makes a clear distinction between PMdorsal and PMventral,
that aligns with the dorsal and ventral visuomotor pathways coming from the posterior brain (see core
literature, Binkofski & Buxbaum (2013). This division corresponds to conditional action
selection/execution (PMd) versus actions requiring object/tool manipulation, including mirror neurons
(PMv). Lastly, it nicely clarifies the differential role of direct spinal connections from M1 and PMA to
action.]
Nachev P, Kennard C, Husain M (2008) Functional role of the supplementary and pre-supplementary motor areas.
Nature reviews. Neuroscience 9(11), 856-869.
[This is a very interesting paper, full of detailed information, but somewhat hard-going. They attempt to
characterize the supplementary motor cortex, a term that they use to refer jointly to pre-SMA and SMA,
with supplementary eye field in between them. This cortical region roughly corresponds to the medial
PFC region discussed in Passingham et al., (2010) and the dorsal medial frontal cortex in Rushworth et
al. (2004; see task 6). It holds several somatotopic maps, and is involved in linking cognition to action.]
40
Task 2: Cognition.
[With some simplification, ‘cognition’ can be seen as the translation of the targets for
action that are available in the environment into the one action that will actually be
performed – i.e., it goes back to the traditional concept of stimulus-response mapping.
Not really surprising, it is the part of the prefrontal cortex just anterior of the
premotor/SMA cortex that is critically involved in this type of cognition. On the lateral side
of the brain the critical cortex is located in the (posterior) inferior frontal sulcus and on
the medial side it is located in the superior frontal gyrus, extending towards (and into) the
cingulate cortex (see ‘preSMA’ and ‘MCC’ in the appendix). This type of cognition is the
simplest form of goal-directed decision making, since the environmental cues signal what
responses are appropriate, based on the task rules that were learned previously. So, the
'correct' response is performed to comply with the requirements of the task (= the goal).
Although we are dealing here with the simples form of goal-directed decision making, it is
not yet clear how the brain accomplishes this stimulus-response mapping. That is not
because these parts of the prefrontal cortex has not been studied yet – on the contrary.
Neurons in this ‘cognitive’ cortex have been shown to respond to all different aspects of
the task (stimulus features, distractors, task rules, responses, rewards), and most neurons
do so only in response to combinations of these (e.g., the red stimulus if it is the target,
but not if it is distractor, etc.). The range of features that are traceable in this cortex
suggests that it receives brain-wide information, but only for the duration of a task trial.
A first and still valid attempt to make sense of this broad and dynamic coding of task
features is described in Miller (2000). The basic idea is that this cortex only stores tags of
the task elements, while the actual representation of these elements happens in remote
and specialized cortical areas (e.g., visual, motor, reward, rules…), and that it executes
computations (possibly together with the basal ganglia, see task 3) that lead to a decision
on what is the most appropriate response given the situation. It is obvious that these ‘tags’
need to be kept active for the duration of the decision and the execution of the correct
response. Because of this need, the ‘cognitive’ cortex of the inferior frontal sulcus has long
been thought to implement working memory. Passingham (Dick, not the Richard from task
1) & Sakai (2004) argue that this working memory hypothesis is too simple. The idea that
the cells in this cortex represent ‘tags’ of currently relevant task features implies that if
from one trial to the next the task-relevant features change, the remote areas to which
this cortex communicates must also change. Xu (2014) comments on a study that shows
just that. Lastly, Bunge et al. (2003) demonstrate that some neurons in this cortex are
selectively responsive to specific task rules regardless of the other features of the task.]
Core Literature
Passingham, D., Sakai, K. (2004). The prefrontal cortex and working memory: Physiology
and brain imaging. Current Opinion in Neurobiology, 14, 163-8.
[(Ventral) lateral PFC has been strongly implicated in working memory (WM), due
to the demonstration of continued cell activity during response delay periods. Here
it is argued that sustained activity is not specific to typical WM tasks, but can be
demonstrated to exist relative to any aspect that is relevant to a task (and it is not
confined to PFC). They argue that PFC’s sustained activity is related to the
“cognitive control” of all the elements that are required during task execution (in
line with Miller’s model), and that it doesn’t reflect so much the maintaining of
peripheral representations as the successful organizing of this information for
prospective use (i.e., the activity is outcome dependent)]
Miller, E.K. (2000). The prefrontal cortex and cognitive control. Nat Rev Neurosci 1, 59-
65.
[This is an influential attempt to characterise the specific contribution of PFC to
complex behaviour. In this paper a simplified overview is given of the corner stones
of this model. More details can be found in Miller and Cohen (2001; see further
reading).
41
Meaningful behaviour is the output of a set of learned stimulus-response mapping
rules and extracting such rules that guide thought and action is a major function of
the PFC. This guidance he refers to as "cognitive control", in that PFC controls lower
level sensory, memory and/or motor operations for a common purpose. The PFC
doesn’t itself represent all the elements that make up a task, but keeps the rules,
stimuli and responses involved active by biasing their peripheral representations.
This keeping active is a form of working memory.
Less important for our purposes are the sections ‘Reward signals and rule
representations’ and ‘Practice and automaticity’. But be sure to look carefully at the
boxes 1, 2 and 3.]
Xu, Y. (2014). Inferior frontal junction biases perception through neural synchrony. Trends
in Cognitive Sciences, 18, 447-8.
[This is a summarizing highlight of the paper by Baldauf and Desomine (2014; cf.
further reading), providing a demonstration of the task-dependent switching in
communication of areas in PFC (i.e., inferior frontal sulcus area IFJ) with
downstream sensory areas (i.e., fusiform face area and parahippocampal place
area). Only 2 pages and easy reading.]
Bunge, S.A., Kahn, I., Wallis, J.D., Miller, E.K., Wagner, A.D. (2003). Neural circuits
subserving the retrieval and maintenance of abstract rules. Journal of
Neurophysiology 90(5), 3419-3428
[This early fMRI study by Bunge and colleagues is an empirical test of the ideas of
Miller about the role of the PFC in goal-directed behaviour. The authors designed
an experiment in which stimulus-response mappings rules had to be flexibly
adjusted based on contextual cues. Using an event-related design, they uncovered
the neural correlates underlying the representation and implementation of abstract
rules showing rule-type sensitivity in a distributed network including (ventral lateral
PFC, parietal cortex and pre-SMA. In the delay period between a cue on which rule
to apply and the actual presentations of stimuli on which to apply the cued rule,
these regions show higher BOLD signal for complex rules (match-to-sample of non-
match-to-sample) compared to simple rules (e.g., “press left” or “press right”,
which require no conditional mapping of stimuli to responses).
Note on terminology: The names used in this paper for the PFC regions are no
longer up to date. Their “ventrolateral” PFC is what is generally referred to as lateral
PFC or inferior frontal junction. Take note of the translational suggestions provided
in Table 1, page 3423, in the distributed pdf copy of the paper. Also, the paper
discusses functional nuances in various other regions which are beyond the scope
of the course. Sections of no interest are crossed out in the provided pdf copy.]
Further Reading
Miller, E.K., Cohen, J.D. (2001). An integrative theory of prefrontal cortex function. Annual
Review of Neuroscience 24, 167-202.
[The elaborated version of Miller (2000).]
D'Esposito, M., Postle, B.R. (2015). The cognitive neuroscience of working memory.
Annual Review of Pychology 66, 115-142.
43
Task 3: The basal ganglia again.
[While the previous task suggests that dedicated regions of the cerebral cortex form a
network for making all sorts of stimulus-response mappings, there is a growing body of
evidence suggesting that basal ganglia play a crucial role in this type of processing, i.e.,
in learning and activating the rules that govern a task and in selecting the appropriate
response patterns depending on the environmental cues.
Last year, you studied the role of basal ganglia in selecting motor patterns, and its
involvement in motor disorders such as Parkinson’s and Huntington’s disease. A refresher
on the organization, (in)direct paths and loops through the basal ganglia is provided by
Whichmann & Delong (2013). To understand the contribution of the basal ganglia to other
than motor aspects of action, you will first study its cortical-striatal-thalamic organization
in more detail in Haber (2016). It will become clear that the basal ganglia and the
prefrontal cortex form an intricate functional unit. While the striatum receives information
from all over the brain (even visual areas, for instance, project to the tail of the caudate
nucleus), the gating function that the direct and indirect paths exert over prefrontal-
thalamo-prefrontal communication is ideally suited to select based on this information (and
learned associations in it) the most appropriate behaviour for the current situation.
Moreover, the functional loops associated with different parts of the prefrontal cortex cover
all aspects of meaningful behaviour (motivation, cognition, attention and motor planning).
Therefore, these loops allow a competition between potential meaningful behaviours at
the level of each of these aspects, and allow to integrate these aspects of behaviour into
the selection of one meaningful behaviour (rather than one motor plan) over another. If
basal ganglia play this central role in behaviour, then there must be more to Parkinson’s
disease than motor selection and initiation problems. This is indeed the case, as is shown
by Cools et al., (2001) in a pioneering study showing that persons with this disease also
suffer from cognitive problems. The idea that basal ganglia are involved in cognition is
further elaborated in the final two papers. Lee et al. (2014) look at the basal ganglia from
an evolutionary perspective and note that this structure is most prominent in all
mammalian brains. Based on the evidence that basal ganglia are phylogenetically older
than the cerebral cortex, they develop the view that the basal ganglia are central to
decision making in general (remember the decorticated cats in Grillner et al. (2005) in
“Functional Neuroanatomy” that could still make meaningful choices). Lastly, Chatham
and Badre (2015) show how the selection/initiation function assigned to basal ganglia in
motor research can be extended to account for its active involvement in working memory
tasks.]
Core Literature
Wichmann T., DeLong, M.R., (2013). The basal ganglia. In: E.R. Kandel, J.H. Schwartz,
T.M. Jessell, S.A. Siegelbaum, A.J. Hudspeth (Eds.), Principles of Neural Science
(5th Ed.). London: McGraw-Hill Medical, pp. 982-998.
[Core literature is only pp. 982-985 left column, and section “Other Basal Ganglia
Circuits are involved in the Regulation of Eye movement, Mood, Reward, and
executive Functions”, p 990-991. This is a refresher on basal ganglia structures and
their functional organisation.]
44
Two observations made by Haber suggest that the loops were together rather than
separately: (1) the cortical inputs to the striatum from functionally different regions
of the prefrontal cortex strongly overlap in the dorsal-medial striatum, and (2)
there exists a spiral of connections between functionally matching topographical
parts of the striatum and midbrain dopamine region. These observations are
graphically summarized in Figures 5 and 6, respectively. Understanding these
figures is crucial. They present the evidence that the different loops through the
basal ganglia work together to produce meaningful, goal-oriented behaviour. Every
endogenously generated meaningful behaviour starts with a motivation (e.g., the
wanting of a cup of coffee), and trough the overlap of projections of cortex to
striatum and the spiralling connections between striatum and midbrain dopamine
neurons, this motivation triggers cognitive, attentional and motor options. These
different components together make up the complex and adaptive behaviour that
we call “getting a cup of coffee”.
The paper mentions a lot of very interesting details, which are beyond the aims of
the course. Hence, the crossed out text can be ignored. The paper focuses on the
striatum, whereas a comparable topographic organisation is also seen in the output
nuclei of the basal ganglia and their thalamic targets. Moreover, it is important to
realize that although Haber confines herself to talking about prefrontal-striatal
connections, the fact is that the whole cortex and subcortical structures (e.g.,
amygdala and hippocampus, but also for instance visual cortex, etc.) send collateral
information to the striatum. Hence, the striatum has information available on all
that is going on in the brain, to learn and/or recognize patterns, which it then uses
to control processing and the (pre)frontal cortex. These and other relevant
structures (e.g., the lateral habenula and Raphe nuclei) are also discussed in the
even more detailed paper by Haber & Knudson (2010) (see Further reading
literature)]
Cools, R., Barker, R.A., Shakina, B.J., Robbins, T.W. (2001). Mechanisms of cognitive set
flexibility in Parkinson’s disease. Brain, 124, 2503‐2512.
[This is an early paper showing that there is more to Parkinson’s disease then motor
problems. Knowing that dopamine modulation and learning play a crucial role in all
functional loops through the basal ganglia it is now self-evident that a disease of
the midbrain dopamine neurons should affect brain function and behaviour beyond
mere motor action selection and initiation. In the paper they try to show that people
with Parkinson’s disease have problems with switching from one task rule to the
other in situations when both rules could be used – i.e., when there are
environmental cues available triggering both task rules. This is evidence that the
basal ganglia’s role in behaviour extends beyond motor selection.]
Lee, A.M., Tai, L.H., Zador, A., Wilbrecht, L. (2014). Between the primate and 'reptilian'
brain: Rodent models demonstrate the role of corticostriatal circuits in decision
making. Neuroscience, 296, 66-74.
[The study by Lee et al. looks at the thesis that basal ganglia play a role in decision
making from a comparative perspective. It is remarkable when one looks at a cross-
section through the brain of a rat and compares it to a similar cut in a macaque
monkey or a human brain, that the basal ganglia is proportionally largest in the rat
followed by the monkey. This suggests that the basal ganglia are the
phylogenetically earlier structure and that the neocortex expanded later. Hence,
basal ganglia might be more vital in dealing with and behaving in the environment
than the cerebral cortex. They give an account of how basal ganglia might
contribute to decision making using a ballot box metaphor: activity in the direct
pathway would be ‘votes for’ a particular option, while activity in the indirect
pathway would be ‘votes against’ that option.]
Chatham, C.H., Badre, D. (2015). Multiple gates on working memory. Current Opinion in
Behavioural Sciences, 1, 213-31.
45
[Short and readable review on how the basal ganglia might contribute to working
memory. Because working memory is capacity limited, it is necessary to control
what goes into WM (input), what in it can be used (output) and what items can be
eliminated because they became obsolete (re-allocation). The authors see a clear
parallel between these working memory control processes and the fundamental
challenges faced by an animal’s motor system: activate action patterns that are
potentially suited for the given situation, execute one in particular when the time
is right, and release all when the situation changes. In action initiation these
functions are attributed to the gating of information between thalamus and cortex
by basal ganglia. Theoretical models posit that motor gating occurs via the
opposing circuit-level effects of the classes of direct path medium spiny neurons
that ‘open the gate’ by facilitating recurrent thalamocortical information flow (Go
cells), and of indirect path neurons that ‘close the gate’ (No-go cells) by blocking
this flow. It is hypothesized that through corticostriatal input the gate is opened for
the winning action pattern, whereas corticostriatal input to No-go cells prevents
other action patterns from becoming active. [This seems to be in line with Purves’
centre-surround model]]
Further Reading
Haber, S.N., Knudson, B. (2010). The reward circuit: Linking primate anatomy and human
imaging. Neuropsychopharmacology 35:4-26.
[This is a detailed account of the anatomy of the reward system, which makes it
difficult to see the forest from the trees. On the other hand it clarifies several
important issues about the organisation of the reward system. It follows somewhat
the same structure as Haber (2016), but goes into much more detail.]
Hélie, S., Ell, S.W., Ashby, F.G. (2015). Learning robust cortico-cortical associations with
the basal ganglia: An integrative review. Cortex 64:123-135.
[In Hélie et al., the hypothesis of basal ganglia involvement in cognition is discussed
more thoroughly, in relation to recent data. They discuss the available evidence for
different cognitive functions.]
46
Task 4: Reward system
[Goal-directed behaviour has everything to do with selecting goals and then selecting
actions to obtain these goals. A very simple form of a goal is obtaining a reward, and the
question that we will try to answer in this and the next two tasks is what happens if
rewards are added to a particular cognitive task – i.e., if instead of choosing one of two
stimuli based on their physical properties (e.g., their shape or colour), the choice is based
on their reward value. In the current task we introduce the core brain system for
processing reward. The reward system is basically the limbic or motivational loop of the
basal ganglia, comprising of ventral striatum, ventral pallidum, and the specific thalamic
nuclei that control the neural communication in the orbital and anterior medial prefrontal
cortex.
Before starting to think about reward-based decision making, however, you have to
become familiar with the basics of the neural processing of rewards. You start with a
discussion of the so called neural reward signal, which consists of a phasic release of
dopamine in the ventral striatum (and other structures of the motivation loop of the basal
ganglia) whenever an unexpected reward is encountered (Schultz, 2007). The dopamine
is released by the dopaminergic neurons, whose cell bodies are located in the midbrain
(substantia nigra pars compacta and ventral tegmental area), and whose axons go to the
subcortical and cortical structures of the motivation loop, as well as many structures
beyond those typical involved in reward-processing. The next paper investigates the range
of behavioural outcomes that can generate this phasic dopamine response in humans
(Daniel, & Pollmann, 2014) – in other words, the authors investigate what counts as a
reward for humans? The following paper studies in detail what effect the reward signal has
on the striatum (Willuhn et al., 2012). More specifically, the paper describes a study of
rats working for a dose of cocaine while the changes in dopamine release in the striatum
are being monitored over a period of weeks. It shows an evolution in the engagement of
sub parts of the striatum while the addiction develops, from the ventral striatum to the
dorsal striatum. It makes clear that while addiction starts out as a motivational problem,
over time it develops into a much bigger behavioural problem. Continued exposure to a
drug changes the connections in the cerebral cortex, first in the affective cortex (orbital
and ventromedial PFC), but later also in other cortical-basal ganglia loops.]
Core Literature
47
[In this paper the role of the phasic dopamine response in learning in humans is
investigated. First a summary is presented of the basic finding in animal studies on
the phasic dopamine response in the VTA/SNc and in the ventral striatum. This of
course overlaps with Schultz (2007). Then a discussion is presented of the
feasibility of measuring the phasic dopamine response using fMRI in humans.
Interesting discussion, but this is not part of the core literature. So, the most
important part of this paper is section 4, where an overview is provided of the type
of feedback that can elicit a phasic dopamine response in ventral striatum. The
range covered is from primary reward, over social rewards and cognitive feedback
(information on performance, rather than a reward), and includes even internally
generated confidence estimates that appear to function as rewards. The
summarized results are very suggestive of a central role of the human
dopaminergic system in a wide range of learning and decision making tasks.]
Willuhn I., Burgeno, L.M., Everitt, B.J., & Phillips, P.E. (2012). Hierarchical recruitment of
phasic dopamine signalling in the striatum during the progression of cocaine use.
Proceeding of the National Academy of Sciences U.S.A., 109, 20703-8.
[Here we look at the effect of the “reward signal” on the striatum and behaviour.
In this study the reward signal is artificially manipulated by administering cocaine
to rats whenever they activate the right key in their cage. Cocaine is a drug that
slows down the re-uptake of dopamine in the striatum (after it has been delivered
by the midbrain dopaminergic neurons). Therefore, activating the correct key leads
to an artificially enhanced reward signal. As you can imagine this procedure is
highly addictive. So, basically, the experiment shows the dynamics of the
nigrostriatal network in the course of developing a cocaine addiction. Initially the
ventral striatum is involved, particularly during the phase of feedback-based (i.e.,
cocaine reinforced) learning. After acquiring the addiction, the feedback is no longer
necessary, and in that phase the dorsal striatum is engaged instead of the ventral
striatum. This means that the motivational addiction (ventral striatum controlled)
is now replaced by a behavioural addiction (dorsal striatum controlled) – i.e., the
habit of performing behaviour that increases the chance of getting cocaine.]
Further Reading
Evers, E.A., Stiers, P., Ramaekers, J.G. (2016) High reward expectancy during
methylphenidate depresses the dopaminergic response to gain and loss. Social,
Cognitive and Affective Neuroscience. doi: 10.1093/scan/nsw124.
[This study sheds light on the relationship between the role of tonic dopamine in
behaviour regulation (which is affected in Parkinson’s disease) and the delivery of
phasic dopamine in response to rewards and action outcome. The tonic dopamine
level in the striatum is psychopharmacologically increased by administering
methylphenidate or placebo while participants perform a reward task in the MRI
scanner.]
Berridge, K.C., (2007). The debate over dopamine’s role in reward. The case for incentive
salience. Psychopharmacology, 191, 391–431.
[A (lengthy discussion of the role that dopamine plays in motivation. The common
view is that dopamine is the signal of the brain that modifies synapses (LTP) in the
striatum to the effect that learning has taken place (i.e., a causal role in learning).
Berridge argues that reward is a much more complex entity in which he
distinguishes three different subcomponents: liking, learning and wanting. If an
animal learns that a stimulus signals that a reward is going to come (i.e., classical
conditioning: it learns to predict the emergence of the unconditioned stimulus), it
learns just that. The fact that it starts to like the predictive stimulus is based on a
separate mechanism (opioid receptor system in n accumbens). The fact that it will
generate actions to obtain the reward is the consequence of another system that
makes the animal want the reward (mesolimbic dopamine system). [This division
48
is reminiscent of Rushworth’s insistence that there are separable mechanisms for
selecting objects, and for selecting actions to obtain them. It also fits with the
massive anatomical connections between MCC and dorsal striatum (Chouinard &
Paus 2006), whereas ventral striatum (involved in liking and wanting, which are
object centred) is connected with OFC and vmPFC]. In the paper, first, an argument
is developed that DA is neither necessary, nor sufficient for learning. Then the
hypothesis is proposed that DA is essential for learning to want stimuli after having
learned that they are predictive. It is because of this that it so perfectly follows the
learning curves of conditioning. It provides also framework for apathy, addiction,
and maybe even Gilles de la Tourette, as disorders of motivation. It also provides
support for the idea that competitions for action are (also) motivational
competitions.]
[This impressive paper is too long and complex. So either we mark selected
sections as core literature or we find a more simple paper.]
49
Task 5: Success, failure, and the cerebral cortex.
[While the previous task looked at the main subcortical structures (midbrain dopamine
neurons and (ventral) striatum) involved in processing unexpected rewards for the sake
of future action selection, this task focuses on the cortical part of the reward system and
its role during reward-oriented decision making. The organization of the prefrontal cortex
for dealing with rather ‘dry’ forms of cognition was discussed in tasks 2 (but also in tasks
1 and 3): action decisions guided by formal rules of the task are associated with the lateral
prefrontal cortex and the medially located preSMA and SMA. In reality, however, many of
our action decisions are directed at obtaining specific goals. While the domain of goals is
extremely wide, particularly in humans, a well-defined and easy to study goal is obtaining
a reward. The main difference in the behaviour paradigms from task 2 and the current
task is that now not the rule, but the participant’s value frame determines which is the
correct option. In a typical cognitive paradigm, the task rule defines the meaning or
relevance of the stimuli and, when two or more stimuli are presented, it also dictates which
stimulus is the correct one, i.e., the one that needs to be chosen or responded to. In a
typical reward-based decision task (for instance, a two-armed maze in a rodent
experiment), several stimulus options are presented simultaneously, each with their own
reward value. The participants have to consider each of these options, weight the value
that each has for them, and decide which they prefer. This is a subjective choice, which
appeals to the self-centred value of the stimuli, which participants acquired in the previous
trials or sometimes even during their life experience before the experiment. In the current
task 5, we investigate how the brain is organized to make such subjective decisions. You
will learn that the mammalian brain has evolved specialized regions for the updating of
stimulus-reward associations based on experience, and for selecting stimuli based on such
stored information.
The first paper is a short review paper presenting the overall organisation of the brain for
reward-based decision making (Lee et al., 2007). The paper is not very detailed, but helps
to make clear that making such decisions is still a cognitive endeavour, requiring the lateral
and medial neural hardware that we already discussed, in addition to the anterior medial
and orbital prefrontal cortex, necessary for making the subjective decisions. The rest of
the literature in this task focuses on the anterior medial and orbital cortex. The first
important thing to know about this value-related part of the cerebral cortex is that it is a
subpart of the anterior node of the default mode network. This overlap becomes clear from
reading the selected pages from Gazzaniga’s handbook, which investigate the
specialization of this anterior default mode node for self-centred value processing. This
overlap has an important functional implication, namely that during value-based decision
making this part of the cortex is DE-activated, as is the case for the entire default mode
network during engagement in cognitive processing. This means that difficulty of making
value-based choices is not reflected in the amount of activity increase, but in the amount
of activation decrease.
Now that these links with the global organisation of the brain are established, the third
paper looks at the neural underpinning of making a reward-based decision. Its conclusion
is that this type of decision making is best seen as a two-phase processes, where first a
decision must be made about which object one prefers, and next what action is best suited
to obtain the preferred object. Moreover, the selection of the preferred object requires the
collaboration of two separate regions, the orbital prefrontal cortex for storing information
about the values (or rewards) associated with objects and the ventromedial prefrontal
cortex for comparing currently available objects based on their stored values, to make a
decision about preference. The importance of the ventromedial prefrontal cortex in reward-
based, or economic, decisions is a main focus of the lecture on Neuro-economics”. The
importance of the orbital cortex as a storage and updating device of object-reward
associations is demonstrated in the last paper, where the effect is investigated of short
term changes to the reward value of objects (in this case foods) on the activity in this part
of the cortex (O’Doherty, et al., 2000).]
50
Core Literature
Lee, D., et al. (2007). Functional specialization of the primate frontal cortex during decision
making. Journal of Neuroscience, 27, 8170-8173.
[This instructive review describes the larger picture. It divides the prefrontal cortex
in three regions and discusses each’s contribution to goal-directed behavior: The
lateral PFC represents the options and cues in the environment, which it keeps
activated even through delay periods when the cues disappeared (see task 2). The
orbital frontal cortex predicts the rewards expected from different options and cues
based on learned stimulus-reward associations stored here (this task). Lastly, the
medial PFC (cingulate cortex) represents the possible action to obtain the predicted
rewards (task 2), and their value (effort, chance of success …) (task 6). This paper,
as are many in this field, is inspired by economics (see also the lecture), and the
authors talk a lot about “utilities” which refers to the quantified expression of
potential costs and benefits associated with alternative actions and choices.]
Gazzaniga, M.S., Ivry, R.B., Magnun, G.R. (2009). Social cognition. In Cognitive
Neuroscience. The biology of the mind, 3rd ed., chapter 14. pp. 599-633.
[Only the first pages are core literature, from 599-606. This is a general
introduction to the contribution of ventromedial prefrontal cortex to behavior. This
part of the PFC belongs to the default mode network, as it constitutes one of the
two central nodes of that network, namely the affective node, whereas the posterior
cingulate cortex is the memory-related node (see the task on the limbic system in
the previous year course Functional Neuroanatomy). So, this affective node of the
default mode network is also the core regions involved in goal-oriented decision
making. After reading this subsection of Gazzaniga, this will no longer be a
surprise. The text explains that this part of PFC is involved in self-referential
information processing. This is of course the core business of goal oriented decision
making, since the value comparisons underlying the selection of goals relates to
values of the choice options for the self, i.e., for me.]
Rushworth MF, Kolling N, Sallet J, Mars RB. (2012). Valuation and decision‐making in
frontal cortex: one or many serial or parallel systems? Curr Opin Neurobiol. 2012
22(6): 946-955. [Partial: 946-949]
[This paper presents a general overview of how the orbital and medial prefrontal
cortex is organized for making reward-based decisions, and it discusses the distinct
contributions of three subparts of this cortex. The first two relate to choosing
objects or options in the environment. The lateral orbital cortex is involved in
learning and updating of stimulus-reward associations. This cortex is active when
new reward cues are learned (i.e., during feedback) or when the value of existing
stimulus-reward associations changes (such as in sensory-specific satiety (cfr
O’Doherty et al., 2000)). The medial orbital and ventral medial prefrontal cortices
(cf. Gazzaniga et al., 2009) play an essential role at the moment of deciding. This
cortex is involved in weighting the values of the available stimulus options and
choosing the one with the highest subjective value. The third of the three
substructures is according to these authors not involved in stimulus selection but
in action selection – i.e., in deciding what action to perform in order to obtain the
chosen stimulus (and associated outcome). This cortex is the subject matter of the
next task. Somewhat confusingly, the authors refer to this structure as the
“anterior cingulate cortex”, but actually it is only the midsection of the cingulate
cortex that is involved in action-outcome association and decision making (see task
6, and the Appendix). They also discuss (in the pages beyond p. 949) an alternative
(or rather parallel) functional role of this midcingulate cortex (making foraging
decisions rather than simple dual decisions) which is very interesting, but not what
we are here interested in. So, you can ignore the foraging part of the story.]
51
O'Doherty, J., Rolls, E.T., Francis, S., Bowtell, R., McGlone, F., Kobal, G., Renner, B., Ahne,
G. (2000). Sensory-specific satiety-related olfactory activation in the human
orbitofrontal cortex. Neuroreport 11:893-7.
[Sensory specific satiety place an important role in the argumentation that lateral
OFC (through amygdala signaling; see Murray (2007, in task 7) is involved in
storing and updating of stimulus-reward associations. Single OFC cells have been
shown to vary their output intensity with the level of satiety for a particular food,
and this doesn’t affect the representation of pleasantness of other foods in the
same area. Here they show in humans that BOLD activity in the (lateral)
orbitofrontal cortex decreases for the odor of a food after having eaten it to satiety,
whereas no such decrease is observed for the odor of the food not eaten. (A
decrease of BOLD activity with satiety was also observed in amygdala in some
participants, pointing in the direction of Murray (2007)’s conclusions in task 7, that
amygdala is the source of this changed affective value of the food.)]
Further Reading
Rushworth MFS, Noonan MP, Boorman ED, Walton ME, Behrens TE, 2011. Frontal cortex
and reward-guided learning and decision-making. Neuron 70, 1054-69.
[An elaborated discussion of the contribution of prefrontal regions to rewad-based
decision making. The maxi-version of Rushworth et al. (2012; see task 8).]
Dunn, B.D., Dalgleish, T., Lawrence, A.D. (2006). The somatic marker hypothesis: a
critical evaluation. Neuroscience and Biobehavioral Reviews 30, 239‐271.
[Brain damage to the orbital prefrontal cortex is associated with severe problems
in making weighted value-based decision, that can be rightly describes as
personality change. This somatic marker hypothesis was developed by Damasio
and colleagues to account for the clinical symptoms of these patients. This paper
provides an exposition and critical evaluation of this theory.]
52
Task 6: Error monitoring.
[In this task you will study the contribution of the medial prefrontal cortex (PFC) to goal-
directed decision making. We encountered the medial PFC in task 1, where we learned
that a medial sub-region of the superior frontal gyrus, around Brodmann areas 6 and 8,
was involved in the endogenous control of action (Passingham, Bengtsson & Lau, 2010).
We also encountered the medial PFC in task 2, where we learned that it is part of the
cognitive network in which the preSMA/SMA is involved in the selection of actions, which
means that its activity rises whenever a choice has to be made between competing
responses. Lastly, in task 5, this subpart of the medial frontal cortex was implicated in the
model proposed by Rushworth and colleagues to capture the neural mechanisms
underlying goal-oriented decision making. They argued for a distinction between two
processes, the first being selecting the preferred object in the environment and the second
selecting the most appropriate response to obtain that object. The former was attributed
to the anterior medial PFC and the latter to the midsection of the medial PFC (which they
unfortunately called the ACC, but consult the appendix to resolve terminological
confusion). The midsection of medial PFC, therefore, can be seen as a region dedicated to
motor control of the environment. It is important to note that this motor-dedicated section
of the medial PFC is larger than preSMA/SMA, and covers most of the medial frontal cortex
superior to the corpus callosum (see the midcingulate cortex or MCC in the appendix).
The literature for task 6 starts with a paper that demonstrates the functional dissociation
between the two separate zones of the medial PFC, in a study of people with selective
brain damage (Camille, Tsuchida, & Fellows, 2011). The other three papers zoom in on
the midsection of the medial PFC. Rushworth et al. (2004) argue that this midsection must
be further divided into a dorsal/posterior subsection (i.e., the SMA/preSMA), involved in
action selection, and a more anterior and ventral region called the rostral cingulate motor
area (rCMA, or also the anterior midcingulate cortex or aMCC), which is involved in the
updating, weighting and evaluation of action-outcome associations. This means that
anterior midcingulate region plays a crucial role in processing the outcome of our actions.
Both positive and negative outcomes provide information on the usefulness of our actions
and it is of crucial importance to store such outcomes for later decisions. Notice the
similarity with the OFC-vmPFC distinction encountered in task 5, where lateral OFC stores
OBJECT-reward associations and the edial OFC/vmPFC uses this stored information to
make real time decisions on OBJECT preferences. Now the aMCC stores ACTION outcome
associations and the SMA/preSMA uses this information to select in real time the most
appropriate ACTION.
Because of its role in monitoring the outcome of selected actions, this part of the medial
PFC has a direct control over the reward system: when a response error is detected – i.e.,
a selected action did not yield the expected rewards – the reward system needs to be shut
down. This most likely happens via the lateral habenula, which has an inhibitory control
over the midbrain dopamine neurons in substantia nigra and ventral tegmental area (VTA).
The action outcome monitoring network is elucidated in an fMRI study described in
Ullsperger and von Cramon (2003). The last paper explores the clinical relevance of these
insights into the functional contributions of the medial PFC, more specifically for psychiatric
disorder, such as major depression disorder and obsessive-compulsive disorder. Both can
be seen as disorders in the processing of behavioural errors.]
Core Literature
53
medial OFC/vmPFC and a group of patients (n=4) with focal lesions in the
midcingulate cortex [what Rushworth would call ACC], with a larger group of
healthy control participants (n=17). Participants perform two reversal learning
task, one in which stimulus-reward associations are learned and re-learned and the
other in which action-outcome associations are relearned. In line with Rushworth’s
model, the first group is selectively impaired on the stimulus selection task
compared to the healthy controls, while the second group is selectively impaired
on the action selection tasks.]
Rushworth, M.F.S., Walton, M.E., Kennerley, S.W, Bannerman, D.M. (2004). Action sets
and decisions in the medial frontal cortex. 8:410-7.
[The paper argues for a functional separation in the motor-related section of the
medial frontal cortex that is reminiscent of the distinction that you encountered
between the lateral OFC and the medial OFC/vmPFC. They discuss literature
supporting the idea that the (pre)SMA is specialized for action selection and more
anterior cingulate motor area for action-outcome learning and representation. This
means that the aCMA stores and updates information about actions and their
outcome, during feedback, while the (pre)SMA uses this information to make an
acute decision about the best action for the current situation. (Note the link here
with what you learned about (pre)SMA in task 1.)]
Ullsperger, M., von Cramon, D.Y. (2003) Error monitoring using external feedback: specific
roles of the habenular complex, the reward system, and the cingulate motor area
revealed by functional magnetic resonance imaging. Journal of Neuroscience
23:4308–4314.
[The reversal learning paradigm used by Camille et al., (2011) critically relies on
the processing of feedback- i.e., information after the trial (getting a reward yes-
or-no – and thus on learning from the outcome of an action. In the fMRI study by
Ullsperger et al. the making of errors is raised to the level of a new task paradigm
in order to reveal how the brain processes error. In this way they demonstrate in
a data driven manner the outcome monitoring network, from midcingulate cortex
up until the subcortical habenula nucleus. They show that this network is closely
related to the reward system and the assessment of reward prediction errors.]
Taylor, S.F., Stern, E.R., Gehring, W.J. (2007). Neural systems for error monitoring.
Recent findings and theoretical perspectives. Neuroscientist 13:160-72.
[Easy introduction to the outcome monitoring function of the medial prefrontal
cortex, with important excursion into its clinical relevance. They see a connection
between medial PFC dysfunction and major depression disorder (MDD) and
obsessive compulsive disorder (OCS). This interpretation of these disorders rest on
the idea that medial PFC is involved in action error monitoring, and that both
disorders result from abnormal responses to behavioural outcome error: In MDD
negative outcomes overcast positive outcomes, leading to negative mood; in OCS
the abnormal obsessive repetitive rituals are thought to originate as a means to
prevent the occurrence of negative outcomes. In there more theoretical account
the reaffirm the distinction put forward by Rushworth at al., 2004.
Read until p. 167 (i.e. skip pages 168 and further).]
Further Reading
Wessel, J.R., Danielmeier, C., Morton, J.B., Ullsperger, M. (2012). Surprise and error:
common neuronal architecture for the processing of errors and novelty. Journal of
Neuroscience 32:7528-37.
[Broadening of the functional characterization of the outcome monitoring network
(Ullsperger & von Cramon, 2003) to processing unexpected outcomes in general –
not just error monitoring.]
54
Rushworth, M.F.S., Buckley, M.J., Behrens, T.E.J., Walton, M.E., Bannerman, D.M. (2007).
Functional organization of the medial frontal cortex. Current Opinion in
Neurobiology 17:220-227.
[A further elaboration of Rushworth et al. (2004).]
Paus, T. (2001). Primate anterior cingulate cortex. Where motor control, drive and
cognition interface. Nature Reviews Neuroscience 2:417-424.
[A readable review on the extent of "motor" cortex in the medial prefrontal cortex,
delineating the cingulate motor areas]
Procyk, E., Wilson, C.R., Stoll, F.M., Faraut, M.C., Petrides, M., Amiez, C. (2014).
Midcingulate Motor Map and Feedback Detection: Converging Data from Humans
and Monkeys. Cerebral Cortex. Webpublished. doi:10.1093/cercor/bhu213.
[A recent "redo" of the Tomas Paus paper, very accurate, but less readable. Ii is
an interesting attempt to link feedback-related response to a specific cingulate
motor region.]
55
Task 7: Adding affect.
[The previous two tasks introduced the brain’s specialized structures for selecting objects
in the environment based on their (dynamically changing) values, and for selecting actions
to obtain these goals and to monitor the outcome of the selected actions. These structures
were revealed by studying brain activity during the execution of tasks in which rewards
(and punishments) for actions were manipulated. Such tasks are typically used in the field
of Neuro-economics. It was also discovered, within this field, that people (and animals) do
not always act “rational”, i.e., maximize their gains and limit their losses. Other factors
then pure gains and losses play a role in what will be decided. A good example is the
experiment discussed in last years’ Functional Anatomy course, where monkeys were
confronted with a plexi-glass container in their cage, in which a reward was visible. In line
with economic theory, the animals directly approached the container and tried to obtain
the reward. However, when the investigators additionally placed a plastic snake on top of
the container, the monkeys behaved irrational and didn’t come near the container. Clearly,
not only gains and losses are taken into account when deciding on actions. In the case of
the monkeys, a strong emotional factor – the fear of snakes – has a lot of impact in
weighing the values of approaching or ignoring the reward. One could say that the fear
down-regulates the reward value of the reward. But even without such a strong emotional
factor, emotion plays a role, for instance, in terms of the emotional value of winning or
losing. And, in general, it is not hard to see that the affective impact of events that occur
around us have an effect on the decisions that we make.
The papers in the current task investigate how this affective value of events finds its way
into the decision making process discussed so far. In the experiment mentioned above,
monkeys could be made to behave “rationally” by lesioning their bilateral amygdala. Since
amygdala is strongly linked with emotion, it is no surprise that amygdala turns out to be
the crucial neural substrate in “irrational” decision making. Studies show that in addition
to its role in other affective processes, the amygdala also tags events, objects, options in
the environment with affective value, both positive and negative. It sends this affective
tag to the cortical regions that are involved in storing and updating the associations
between events and outcomes, namely the lateral part of the orbital frontal cortex (lOFC)
for stimulus-reward (or value) associations, and to the anterior midcingulate cortex
(aMCC; also often refer to as (dorsal) ACC) for action-outcome associations. In other
words, the behavioural value of an object is not only determined by the reward that is
associated with it, but also by the affective value that it has for the organism. This affective
value is assessed by the amygdala, and processes in the same pathway that processes
reward value for decision making. So, in opposition of the well-known dual process models
of irrational decision making developed in economic theory, neuroscientific evidence
suggests that there is only one decision making pathway and it processes both types of
values.]
Core Literature
Murray, E.A. (2007). The amygdala, reward and emotion. Trends in Cognitive Sciences,
11, 489-497.
[This paper attempts to establish the contribution of amygdala to decision making.
To clarify its role, she first resolves two major misconceptions about the amygdala,
one being that amygdala is involved only in negative emotions, and the other that
amygdala is critically involved in stimulus-reward association learning. She
concludes that reward and emotion are not identical: although the amygdala is
essential for processing the emotional aspects of reward, including its valence
(positive or negative) and its relative value (e.g. good versus superb), many other
aspects of reward processing are handled outside the amygdala. After this
clarification Murray develops a theory of the contribution of amygdala to reward
processing – and hence, reward-based decision making – which can be summarized
as that amygdala tags the options and cues encountered in the environment with
56
affect and sends this value information to the lateral part of OFC (which in parallel
receives information about the identity of objects and rewards from sensory
systems). In OFC this affective value is integrated with other aspects of reward,
and stored for later decision making in the ventromedial prefrontal cortex (cf. task
5). The discussion is somewhat lengthy (you don’t need to know all the details),
but the model is nicely summarized in Figure 3.]
De Martino, B., Kumaran, D., Seymour, B., Dolan, R.J. (2006). Frames, biases, and
rational decision-making in the human brain. Science, 313(5787), 684‐687.
[This is a famous study in neuro-economics, in which they tested a typical prospect
theory experiment based on the behavioural paradigm of loss aversion in an MRI
scanner. The framing effect refers to the observation that people tend to deviate
from a pure rational or economic strategy in gambling tasks depending on whether
the outcome is formulated as a loss or as a gain. The result shows that amygdala
is responsible for this “irrational” behaviour. Therefore, this is an illustration in
humans of the model put forward by Murray (2007). See also:
https://www.youtube.com/watch?v=v-qdfG7Z9qU.]
Hampton, A.N., Adolphs, R., Tyszka, J.M., O’Doherty, J.P., (2007). Contributions of the
amygdala to reward expectancy and choice signals in human prefrontal cortex.
Neuron, 55, 545-555
[Study of the changes in brain activity in two patients with focal bilateral amygdala
lesions in a reversal learning task (in which learning is mainly driven by ventral
striatal mechanisms, as argued in Murray (2007)). The discussion doesn’t go very
deep and is therefore not important, but the results presented confirm the
contribution of amygdala to the decision to shift strategy. ]
Further Reading
Paton, J.J., Belova, M.A., Morrison, S.E., & Salzman, C.D. (2006). The primate amygdala
represents the positive and negative value of visual stimuli during learning. Nature,
439, 865–870.
[Physiological study in Monkeys providing evidence that single cells in amygdala
respond to the valence of stimuli independent of their identity.]
57
Task 8: Trying to make it work.
[Throughout the different tasks, various aspects of the neurobiology of goal-directed action
selection have been discussed. Now it is time to look for the forest instead of the trees.
Below are two papers in which the authors attempt to develop an integrated picture of the
working of the prefrontal cortex. Of course, there are many other attempts available.
These two are important because they emphasize the cortical-subcortical basis of the
integrated functioning of the brain.]
Core Literature
Krauzlis, R.J., Bollimunta, A., Arcizet, F., Wang, L. (2014). Attention as an effect not a
cause. Trends in Cognitive Sciences 18, 457-64.
[Attention is usually seen as a causal agent that exerts its effects on the posterior,
sensory brain structures. This effect is often compared to a filter which allows only
a limited subset of sensory signals to reached later stages of processing. In other
words, brain structures at later stages of processing generate ‘priority’ signals that
bias competition for representation in sensory cortices, imposing top-down control
on these sensory cortices. However, within the larger framework of goal-oriented
behaviour, attention is not a causal agent, but an effect: the biasing influence of
attention arises from and is steered by circuits involved in value-based motor and
non-motor decision making. From this larger perspective, making good decisions
requires making good estimations of states. States are seen as conglomerates of
learned associations between inputs, outputs, and prior knowledge. Each state is a
template for a particular situation and dictates what are relevant stimuli, actions
and rules in it. Hence, a state also dictates what to attend to. The authors
hypothesize that estimating the most appropriate state for the current situation,
and then making it the dominant one, involves the basal ganglia.]
58
Appendix: Anatomical nomenclature
You might encounter inconsistent use of anatomical terms in the literature. This is the
nomenclature we will adhere to in lectures, education group sessions and exam questions.
The lateral side of the prefrontal cortex is typically subdivided into three gyri – inferior,
middle and superior – by its two major sulci that run from anterior to posterior: the inferior
(IFS) and the superior (SFS) frontal sulcus. Note that the superior frontal gyrus extends
into the medial wall and continues until the cingulate sulcus. At the dorsal and of the major
frontal sulci are two important cognitive structures: At the junction of SFS and precentral
sulcus are the frontal eye fields (FEF), and at the junction of IFS and precentral sulcus is
“IFJ” or inferior frontal junction.
Two other structures discussed in the course are the “lateral” prefrontal cortex (LPFC),
covering the banks and the depth of the posterior part of IFS (fuchsia ellipse), and the
“dorsolateral” prefrontal cortex (DLPFC), more anterior on the surface of the middle frontal
gyrus (MFG, blue ellipse).
The "medial wall" is the cortical surface within the longitudinal fissure. Its three major
subdivisions relate to the three lobes that contribute to the medial cortex. Thus we can
speak of the medial (pre)frontal, medial parietal and medial occipital cortex.
59
In addition to these lobar divisions, a major division of the medial wall originates from the
cytoarchitectural discontinuities in the cingulate gyrus.
- The cingulate gyrus is the cortical rim between the corpus callosum and the cingulate
sulcus, and it is curled all around the corpus callosum (see red region in figure above).
- Posteriorly it extends into the retrosplenial cortex and then becomes the enthorinal cortex
in the medial temporal lobe. This entire ring-shape cortex (pink in the figure below) is the
limbic cortex.
According to the influential model of Vogt ea (2005), the cingulate cortex can be divided
into three sections, referred to as the anterior, middle and posterior cingulate cortex --
ACC, MCC and PCC. The division between the PCC and MCC coincides with the border
between Brodmann areas 23 (pcc) and 24 (mcc), with most of the PCC covering the
parietal part of the medial wall. The distinction between the ACC and MCC also coincides
with cytoarchitectonic changes but within Brodmann area 24, and functionally differentiate
the affective ACC from the more motor selection related MCC (e.g., comprising a.o. the
cingulate motor areas)
Within the ACC a further subdivision is made between the pre-genual part (pACC) and the
subgenual part (sACC). These terms refer to the relative position of the cingulate cortex
relative to the genu (“knee”) of the corpus callosum. The MCC is also subdivided, into an
anterior part and posterior part.
The large gyrus (purple region in figure above) dorsal to the cingulate sulcus is the medial
aspect of the superior frontal gyrus. In value and reward-related neuroscience the rostral
part of the extra-cingulate cortex is often referred to as the ventral medial prefrontal
cortex (VMPFC) -- implicated in goal-based decision making -- and the dorsal medial
prefrontal cortex (DMPFC) -- implicated in social cognition.
The orbital prefrontal cortex can grossly be divided in three major divisions (Uylings ea,
2010), a lateral, central and medial part (fuchsia, yellow and green, respectively, in
figure). Each of these three divisions also show cytoarchitectonic differences in the
anterior-posterior dimension, but there is no sharp boundary between these subparts, but
rather a transition zone.
60
2. Cortical motor areas
61
2.2. Medial view
3. Specific delineations
The cingulate gyrus is the stretch of cortex between the corpus callosum inferior to it and
the cingulate sulcus superior to it. So, it is the name of a gyrus or a visible part of the
cortical surface in the traditional sense of neuroanatomy. However, a large part of the
cerebral cortex is not visible at the surface, but hidden in the depths of the sulci. While
not visible, this hidden cortex is nonetheless functionally relevant, and researchers need
62
to be able to refer to it to describe their study results. For that reason they divide the
cortical sheet in regions that do not always align with the traditional neuroanatomical
division schemes. The term “cingulate cortex “ is such a functionally defined term. Since
research has shown that the cortex inside the cingulate sulcus has functional relevance
that aligns with that of the cingulate gyrus, custom has grown to refer to the cortex on
the gyral surface and in the sulcus with one term: the cingulate cortex. This custom is
illustrated in the figure below (Left panel from Cole, Yeung, Freiwald, & Botvinick, 2009;
right panel from Vogt, Nimchinsky, Vogt, & Hof, 1995).
The term ACC (“anterior cingulate cortex”) is often very loosely used both by people who
study cognition and by people within the field of affective neuroscience. Only in review
articles with a wider perspective the need is felt to use a more precise terminology, such
as the subdivision between anterior and middle cingulate cortex, which is inspired by the
neuroanatomical studies of Brent and Leslie Vogt (Palomero-Gallagher et al., 2009; Vogt
& Palomero-Gallagher, 2012). This division is not only supported by neuroanatomical data,
but is also functionally meaningful, in the sense that the anterior portion of the cingulate
cortex is often referred to as the ‘A’ffective cortex, whereas the midsection is often referred
to as the ‘M’otor part of the cingulate cortex.
While this terminology is sound and clear, unfortunately it is often not followed in the
existing literature. This is usually for historical reasons, and it poses a challenge to the
reader of the many papers on the medial prefrontal cortex. In several of the papers used
as core literature in this course, the term ACC or sometimes dorsal ACC is used to refer to
what actually is the motor section of the cingulate cortex (so ‘MCC’). These are usually
people doing monkey research (‘ACC’) or cognitive neuroscientists (‘dACC’). To avoid the
confusion, you should ask yourself whether the person is talking about motor-related
functions or more affective and value-related functions. In the latter case, they are
justified to speak of ACC, but in the former case you better translate the terms used to
MCC.
This distinction is one of the central themes in the current course: goal-directed action is
based on two successive decisions, the first being which of the presented options do I
prefer (value based, hence affective and self-centered, hence ACC), and what action will
a select to obtain that option (motor related, based on previous action outcome monitoring
and therefore MCC). That is a recurring distinction in most of the current scientific
literature and the dissociation is nicely demonstrated in the paper by Camille et al., (2011).
This model of goal-directed decision making is more general discussed in Rushworth,
Kolling, Sallet & Mars (2012).
It should be point out, however, that usually when the discussion is about the affective
part of ACC the larger aspect of anterior medial cortex is meant. This extends beyond ACC
proper and includes the ventromedial prefrontal cortex (vmPFC), which is itself not
rigorously defined but certainly more ventral-anterior than the sub-genual and pre-genual
ACC (sACC and pACC in the figures). This term is a used rather vaguely to refer to any
activation that is related to reward expectations and reward related choice values, which
often also involve some ACC activity (in the pre-genual ACC area). Moreover, in Macaque
monkeys the part of the cortex is shifted more downward and towards the medial side of
the orbital cortex. This is why in monkey studies on reward-based object selection they
speak about medial orbital/ventromedial PFC. This larger affective cortex coincides with
one of the two major nodes of the default mode network, which should be evident from
the chapter in Gazzaniga, Ivry, & Magnun (2009, p. 599 ff.).
Similarly, one should also be aware that when we talk about the motor-functions of the
medial PFC, that this is not confined to the midcingulate cortex. For, voluntary action
selection is typically associated with preSMA and SMA, which are located superior to the
63
MCC proper. This is what Rushworth et al. (2004) refer to as the superior frontal gyrus
(SFG), as opposed to the more inferior and anteriorly located MCC (which they
misleadingly call “ACC”). So, they talk about ACC and SFG, but they should say MCC and
(pre)SMA. The distinction between these two sub-regions of MCC mirrors the distinction
between the lateral orbital cortex and vmPFC for the processes related to object selection:
keeping track of the changing values of things and making concrete decisions in a choice
situation. The motor equivalent is of course actions instead of objects: the anterior MCC
is involved in updating outcomes of actions, while the (pre)SMA is involved in making
acute decisions about actions in an ongoing situation.
Some clarification might be useful regarding the correspondence between the so called
“lateral prefrontal cortex” (LPFC) in macaque monkeys and in humans. Several of the
papers discussed were published almost two decades ago, around the turn of the century,
when functional imaging was still young, and nomenclature was dominated by macaque
monkey research. In the Macaque monkey the cortical region of the prefrontal cortex
mostly studied as the LPFC is the cortex in the posterior principle sulcus. In the figure
below it can be seen that this part of the macaque cortex maps onto the cortex found in
the interior frontal sulcus in the human brain. There is a considerable preservation of the
ordering of the fields across the two species, although functionally there are certainly
unresolved mapping issues. One example is the frontal eye fields, which are identified as
8Av in the monkey brain (e.g., Amiez, & Petrides, 2009; Hutchison et al., 2012). However,
in fMRI studies of the human brain the frontal eye fields are repeatedly identified at the
junction of the superior frontal sulcus and the precentral sulcus (the light blue field in
panel B of the Figure) (Stiers et al., 2006; Amiez, & Petrides, 2009; Hutchison et al.,
2012). There is also inconsistency with respect to the numbering of fields, which is broadly
inspired by the Brodmann numbering system. In the influential but relatively coarser
Walker parcellation scheme for the macaque brain (Walker, 1940) the entire cortex in the
principle sulcus is identified as area 46, which gives rise to confusion, because area 46 in
the human brain is located more anterior on the surface of the middle frontal gyrus (Goulas
et al., 2012, See yellow cluster in panel B of the Figure; Rajkowska, Goldman-Rakic,
1995). The extrapolation of neurophysiological results from macaque studies then wrongly
suggests that this dorsolateral part of the human cortex is the human homologue of the
principle sulcus in the macaque monkey brain.
While most cell recording studies in the macaque brain have only a limited spatial window,
recent fMRI studies in the macaque monkey allow for a more complete comparison of the
organization of the monkey and human brain for cognition. Such studies confirm that the
posterior part of the inferior frontal sulcus cortex, central to human cognition, maps to the
posterior part of the principle sulcus cortex in the macaque brain, and particular the ventral
bank of the principle sulcus. Nakahara, Hayashi, Konishi, & Miyashita (2002) collected
fMRI data while monkeys and humans performed an adaptation of the Wisconsin card
Sorting task and found in both species strategy switch related activity in the ventral
prefrontal brain, associated with the anterior bank of the inferior precentral sulcus in
humans and the inferior prefrontal convexity (possibly area 44) in the macaque data.
Neither species showed activity in the inferior frontal or principle sulcus, respectively. More
recent studies of monkeys performing typical cognitive tasks in the fMRI scanner do report
activity in this region, however, in addition to the more inferior lateral activity clusters
reported by Nakahara et al. (2002). Zinke (2013) administered a relayed matching to
sample task to monkey and human subjects and found during the delay period in two
monkeys increased activity in the mid-to-inferior arcuate sulcus (8Av/44) and the posterior
principle sulcus (9/46v) and in humans in the inferior precentral sulcus and the posterior
inferior prefrontal sulcus, particularly at the junction between precentral sulcus and inferior
frontal sulcus. Moreover, Premereur, Janssen and Vanduffel (2018) found increased
64
activity in the ventral prefrontal convexity, extending into the ventral bank of the posterior
principle sulcus and in the anterior bank of the inferior arcuate sulcus in monkeys during
the switch trials compared to stay trials in a response switch task.
The inferior frontal junction (IFJ) is located at the junction between the inferior frontal
and the precentral sulcus. It is located in the frontal lobe, but it is not motor cortex;
therefore, it is "prefrontal" region. IFJ is the most posterior part of the cortex in and around
the inferior frontal sulcus, i.e., the cortex involved in what we call 'cognition' (=keeping
active the task-relevant information and rules for stimulus-response mapping). The more
complex a task becomes, the more the activation around inferior frontal sulcus will extend
from IFJ to more anterior cortex.
The dorsal-ventral division in the lateral PFC is not yet well established. So, it is not
worth while at the level of this course to spend much time on this topic. Many different
attributions have been proposed (spatial versus objects, simple versus complex, visual
versus auditory, etc.). For the purpose of this course it is sufficient to consider the lateral
prefrontal cortex as a whole, i.e., the cortex on the banks and the bottom of the posterior
principle sulcus in macaque monkeys and around the posterior inferior frontal sulcus in
humans. The neurons that are active during delay periods are found in this lateral
prefrontal cortex (both ventrally and dorsally).
65
the four numbered activity centers defined by Stiers, Mennes & Sunaert (2010) and
presented in subpart F.
D) Group brain activity increase across 10 human participants performing a delayed
matching to sample task similar to (C). The same activity contrast is as in (C) is overlaid
on a partially inflated human grey matter-white matter surface reconstruction of the left
hemisphere (from Zinke, 2013, Figure 3.11). Light blue dots and numbers are similar to
those in (C) and (F).
E) Brain activity of two macaque monkeys during performance of a response modality
switching task (hand or saccade response) (Premereur, Janssen & Vanduffel, 2018, Figure
2). Shown is the conjunction across two monkeys of switch trials versus stay trials, overlaid
on a flat map of the left grey matter-white matter surface. Open stars mark the local
maxima in the conjunction map. Light blue dots and numbers are as in (C) and (F).
F) Spatial distribution of activity foci that co-localize across 4 different cognitive tasks.
(Stiers, Mennes & Sunaert, 2010, Figure 2). Each spherical dot marks the center of mass
of these activity foci across 10 participants. Orange dots mark foci with a significant
multiple demand response (increased activity with task difficulty in all four tasks), while all
non-orange dots lack this property. Dark blue dots indicate foci whose BOLD signal relates
to the motor response, while green dots are systematically deactivated in each task. The
yellow dot marks the position of the human frontal eye fields. The light blue numbers serve
to identify the activity foci relevant for the present discussion across the two species (see
(C), (D) and (E)). The dot with number '1' marks the inferior frontal junction, at the junction
between the inferior frontal and the precentral sulcus. The dot numbered '2' is in both
species located in the ventral part of area 9/46. Dot number ‘3’ is more ventrally located
and may in both species fall within area 44. The dot numbered ‘4’ marks the frontal eye
fields. Note that the activity foci attributed to the frontal eye fields in the monkey and
human brain do not agree with the identification of cortical fields, which only in the monkey
corresponds to area 8Av.
4. References
Amiez C, Petrides M. (2009). Anatomical organization of the eye fields in the human and
non-human primate frontal cortex. Prog Neurobiol. 89(2):220-30.
Cole MW, Yeung N, Freiwald WA, Botvinick M. (2009). Cingulate cortex: Diverging data
from humans and monkeys. TICS 32: 566-574.
Goulas A. Uylings HBM, Stiers P. (2012). Unravelling the intrinsic functional organization
of the human lateral frontal cortex: A parcellation scheme based on resting state fMRI.
Journal of Neuroscience 32:10238-10252.
Hutchison RM, Gallivan JP, Culham JC, Gati JS, Menon RS, Everling S. (2012). Functional
connectivity of the frontal eye fields in humans and macaque monkeys investigated
with resting-state fMRI. J Neurophysiol. 107(9):2463-74.
Palomero-Gallagher N, Vogt BA, Schleicher A, Helen S. Mayberg HS, Zilles K. (2009).
Receptor architecture of human cingulate cortex: Evaluation of the four-region
neurobiological model. Human Brain Mapping 30:2336-55.
Petrides M, Pandya DN (2006) Efferent association pathways originating in the caudal
prefrontal cortex in the macaque monkey. J Comp Neurol. 251:227–251.
Premereur E, Janssen P, Vanduffel W. (2018). Functional MRI in macaque monkeys during
task switching. J Neurosci, 38: 10619-10630.
66
Rajkowska G, Goldman-Rakic PS (1995) Cytoarchitectonic definition of prefrontal areas in
the normal human cortex. I. Remapping of areas 9 and 46 using quantitative criteria.
Cereb Cortex 5:307–322.
Vogt BA, Nimchinsky EA, Vogt LJ, Hof PR (1995). Human cingulate cortex: surface
features, flat maps, and cytoarchitecture. J Comp Neurol. 359:490-506.
Vogt BA, Palomero-Gallagher N. (2012): Cingulate gyrus. In: Mai, JK, Paxinos G. (Ed.).
The Human Nervous System. Amsterdam: Elsevier. pp 943–987.
Walker AE. 1940. A cytoarchitectural study of the prefrontal area of the macaque monkey.
J Comp Neurol. 73:59–86.
Stiers, P., Mennes, M., & Sunaert, S. (2010). Distributed task coding throughout the
multiple demand network of the human frontal-insular cortex. NeuroImage, 52, 252-
262.
Zinke W. (2013). Exploring cognitive states of the macaque monkey with functional fMRI.
PhD dissertation, University of Breme, Germany.
67