Rembrandt self-portrait

Causation in
Complex Systems

George Ellis
University of Cape Town

ISC-PIF, Paris
November 2009
 1: The nature of causation
I will claim here that there are other forms of causation than
those encompassed by physics and physical chemistry, and these
are described quite well by Aristotle’s four types of causes .
A full scientific view of the world must recognise this fact, or
else it will ignore important aspects of causation in the real
world, and so will give a causally incomplete view of things

The key idea I will pursue is that as well as bottom-up

causation, top-down causation takes place in these structures,
due in particular to the crucial role of context in determining
the outcomes of lower level causation. Because of the existence
of random processes at the bottom, there is sufficient causal
slack at the physical level to allow all these kinds of causation
to occur without violation of physical causation.
fern
Fifth day of creation:
M C Escher
 Topics:
2: Complexity and Structure

3: Bottom-up and Top-down action

4: Five different kinds of Top-down action

5: The nature of causality

6: Predicting and intervening

 2: Complexity and Structure
True complexity, with the emergence of higher levels of order
and meaning, occurs in modular hierarchical structures
because this is the only viable ways of building up real
complexity on the basis of the underlying physics.

A hierarchy, with many layers of structure built upon each

other, represents different levels of abstraction, each built upon
the other, and understandable by itself. This is the phenomenon
of emergent order.

There will be a different description and vocabulary suitable at

each level of the hierarchy, related to the effective entities that
occur at that level.
The Hierarchy of Structure: 1
Sociology/Economics/Politics

Psychology

Botany/Zoology/Physiology

Cell biology

Biochemistry

Chemistry

Atomic Physics

Particle physics
 Hierarchy
• The key to handling complexity is hierarchical information
structure and analysis, and associated physical structuring

• A modular hierarchy represents a decomposition of the

problem into constituent parts, and into processes to handle
those constituent parts, each requiring less data and
processing, and more restricted operations, than the problem
as a whole.

• The success of hierarchical structuring depends on:

(a) implementing modules to handle lower-level processes,
(b) integration of these modules into a higher-level structure.

Grady Booch: Object Oriented Programming

 Modules and linkages
“We find separate parts that act as independent agents, each of which
exhibit some fairly complex behaviour, and each of which contributes
to many higher level functions. Only through the mutual co-operation
of meaningful collections of these agents do we see the higher-level
functionality of an organism. This is emergent behaviour – the
behaviour of the whole is greater than the sum of its parts” (and cannot
even be described in terms of the language that applies to the parts)

“Intra-component linkages are generally stronger than inter-component

linkages. This fact has the effect of separating the high-frequency
dynamics of the components – involving their internal structure – from
the low-frequency dynamics –involving interactions amongst
components” (this is why we can sensibly identify the components)
 Modularity: Abstraction
Abstraction and Labelling: Unable to master the entirety of a
complex object, we choose to ignore its inessential details,
dealing instead with a generalised idealised model of the object.
An abstraction denotes the essential characteristics of an object
that distinguishes it from all other kinds of objects.

It focuses on the outside view of the object, and so serves to

separate its essential behaviour from its implementation; it
emphasises some of the system’s details or properties, while
suppressing others. “Information has to be thrown away by the
billion bits all the time, because all the alternatives cannot be
examined” .

Key feature: Compound objects can be named and treated as units

by appropriate labelling. This leads to the power of abstract
symbolism and symbolic computation.
 Modularity: Inheritance
- Structuring of modular units with abstraction,
encapsulation with consistent interfaces, and
inheritance enables the modification of modules and
re-use for other purposes.

- Inheritance is the most important feature of a

hierarchy: it allows an object class to inherit all the
properties of its superclass, and to add further
properties to them (it is a `is a’ hierarchy).
- For example: cells specialise to neurons.

Efficiency and usability introduce the aim of reducing

the number of variables and names that are visible at
the interface, as is implied by encapsulation
 Modularity: Encapsulation
Encapsulation and Information Hiding: consumers of
services only specify what is to be done, leaving it to the object
to decide how to do it; this is an aspect of decentralisation of
control.

Encapsulation is when the internal workings are hidden from

the outside, so its procedures can be treated as black-box
abstractions. “No part of any complex system should depend on
the internal details of any other part”.

It involves information hiding – hiding all the internal aspects

of an object that do not contribute to its essential characteristics
[corresponding to coarse-graining in physics; the
accompanying loss of detailed information is the essential
source of entropy].
 Information Hiding:
- a physicist does not need to know the position and velocity of
each molecule in a gas; overall temperature and pressure will do

- a chemist does not need to know about nuclear binding forces

and quarks; nuclei and electrons can be treated as components of
molecules

- a biologist does not need to know all the internal workings of a

cell to understand its role in an organism

- a motor car driver does not need to know the details of how the
carburetor or battery works;

- an employee does not need to know the inner working of the

payroll department (who has to sign what forms); she just needs
to know when the cheque is coming
 Network structures
Network structures: represented as directed graphs
Motifs: building block patterns in complex networks
Uri Alon, An Introduction to systems biology:
Design principles of biological circuits (2007)

Maximal decentralisation of function, with coherence of goals

- Stafford Beer, Brain of the firms

Ross Ashby: Law of Requisite Variety stating that "variety absorbs

variety”, defines the minimum number of states necessary for a
controller to control a system of a given number of states.
• Arithmetic simplicity beneath metabolic network architecture
• Authors: W. J. Riehl, P. L. Krapivsky, S. Redner, D. Segre
(26 Oct 2009: arXiv:0910.4943v1 q-bio.MN)]

• Abstract: Metabolic networks perform some of the most

fundamental functions in living cells, including energy
transduction and building block biosynthesis. Is the evolution of
metabolism subject to general principles, beyond the
unpredictable accumulation of multiple historical accidents?
Here we search for such principles by applying to an artificial
chemical universe some of the methodologies developed for the
study of genome scale models of cellular metabolism. In
particular, we use metabolic flux constraint-based models to
exhaustively search for artificial chemistry pathways that can
optimally perform an array of elementary metabolic functions.
• Despite the simplicity of the model employed, we find that the
ensuing pathways display a surprisingly rich set of properties,
including the existence of autocatalytic cycles and hierarchical
modules, the appearance of universally preferable metabolites
and reactions, and a logarithmic trend of pathway length as a
function of input/output molecule size. Some of these properties
can be derived analytically, borrowing methods previously used
in cryptography.

• In addition, by mapping biochemical networks onto a simplified

carbon atom reaction backbone, we find that several of the
properties predicted by the artificial chemistry model hold for
real metabolic networks. These findings suggest that optimality
principles and arithmetic simplicity might lie beneath some
aspects of biochemical complexity.
 Multiple Realisation
A key feature is that higher levels will be realisable in
multiple ways through the entities at the underlying lower
levels:

- molecules in a gas
- electrons in a computer
- cells in a living body
- animals in an ecosystem
- people in an organisation

The higher level entity has an existence that does not

depend on the specific lower level realisation
(molecules in our body).
3: Bottom-up and Top-down action
Bottom-up action is when what happens at the higher levels is
controlled by what happens at the lower levels

- micro-physics underlies macro physics, e.g. kinetic theory of

gases, theory of solids (conduction, thermal capacity)

- physics underlies chemistry, e.g. nature of chemical bond

- protein folding and recognition is based on chemical bonding

- cells with their own internal function underlie all life,

- physics and chemistry underlie the functioning of the brain

- individual human behaviour underlies the functioning of society

 Level 2

Level 1

Bottom-up causation alone:

Micro forces determine what happens at the

higher levels

They are the foundation of higher level activity

 Bottom-up action
Allows the building up of a certain degree of complexity:

- Can lead to quite complex patterns occurring (sandpiles, reaction

diffusion equation, convection patterns, cellular automata, etc)

- Often guided by attractors in the possibility space:

One ends up there from a large basin of attraction,
irrespective of where you start in that basin

Restricted by fundamental constraints, e.g. to do with matter and

energy conservation

But cannot by itself generate genuine complexity, such as a single

living cell.
 Top-down causation
Top-down causation is when the higher levels of the hierarchy
causally effect what happens at the lower levels, in a coordinated
way. This occurs by higher level features setting the context for
lower level actions, the resulting constraints thereby creating new
possibilities.

In physical systems, structured interactions (e.g. electrons guided by

wiring between components in a computer) can lead to high level
behaviour of great complexity (e.g. the internet)

In social systems, accepted social frames and associated role models

can guide individual activities so that they add up to a coherent
complex emergent whole (e.g. an educational system).
 Level 2

Level 1

Bottom-up and top-down causation:

Additionally the higher levels control causal effects at the

lower levels. This allows inter-level feedback loops.

Claim: Emergence of genuine complexity is

possible only because of top down causation.
Reliable emergent higher level behaviour
Set initial higher level state: what transpires?

H1 H'1

L1 L'1
L2 L'2
L3 L'3

The lower level dynamics lead to coherent higher same-level

dynamics when the lower level dynamics acting on all the different
lower level states corresponding to a single higher level state, give
new lower level states corresponding to the same higher level state.
Examples: gas laws; inland revenue service, payroll office.
 No reliable higher level behaviour:
?? H'2
H1 H'1 H'3

L1 L'1
L2 L'2
L3 L'3

The lower level dynamics does not lead to coherent higher level
dynamics when the lower level dynamics acting on different lower level
states corresponding to a single higher level state, give new lower level
states corresponding to different higher level states.
Example: chaotic systems: weather; hairdresser; president of country.
 Three contexts of emergence:
1st: Evolutionary history of the universe and the world:
Once upon a time they did not exist!

2nd: Developmental history of each living being:

Once upon a time they were a single cell.

3rd: Functional nature of each complex object:

built up out of components that do not have the higher level
properties.

I suggest that in each case, it is top-down causation that enables

emergence of true complexity, which in turn enables more
complex forms of top-down causation.
 Fine tuning: Just Six Numbers [Martin Rees]

• 1. N = electrical force/gravitational force =1036

• 2. E = strength of nuclear binding = 0.007

• 3. Ω = normalized amount of matter in universe

= 0.3

• 4. Λ= normalised cosmological constant = 0.7

• 5. Q = inhomogeneous seeds for cosmic structures =

1/100,000

• 6. D = number of spatial dimensions = 3

4: Top-down action: five different kinds
Rather than referring just to top down causation, it is useful to
distinguish five different classes of top-down causation:

1. Algorithmic top-down causation

2. Top-down causation via non-adaptive information control
3. Top-down causation via adaptive selection
4. Top-down causation via adaptive information control
5. Intelligent top-down causation (the effect of the human mind on
the physical world)

The key technical concept in each case is equivalence classes: a large

number of lower level states can underlie a single effective high level
state. But the higher level state controls the lower level dynamics; the
corresponding lower level states form an equivalence class, with
respect to the higher level context, so the outcome is the same.
 1: Algorithmic top-down causation
Algorithmic top-down causation occurs when high-level
variables have causal power over lower level dynamics through
system structuring, so that the outcome depends uniquely on the
higher level structural, boundary, and initial conditions.

It occurs through determination of structure that controls micro-

interactions: which components are connected together in a
computer, which neurons are joined by synapses, which memory
locations are in what state. The lower level variables determine the
outcome in an algorithmic way from the initial and boundary
conditions as a consequence of the higher level structural relations

Examples. Algorithmic computational procedures in a digital

computer determined by the wiring and the software;
Bureaucratic procedures with interchangeable operatives.
 Example: Biological development
Reading of DNA codings: The central process of developmental
biology, whereby positional information determines which genes
get switched on and which do not in each cell, so determining their
developmental fate, is a top-down process from the developing
organism to the cell, largely based on the existence of gradients of
positional indicators (morphogens) in the body.

Without this feature organism development in a structured way

would not be possible, for each cell has the same genetic material.
Thus the functioning of the crucial cellular mechanism determining
the type of each cell is controlled in an explicitly top-down way; it
depends on the context

But it is algorithmic in that it is mechanical, given the context .

 Localised
gene expression

Segment properties

Positional expression of genes

leading to segments …
 Positional
Information

Gene
Expression
fly

Specific
proteins

Developmental
outcome
Example: Neuronal processes

• The brain is made of interconnected neurons

• Information flows: dendrites to nucleus to axon to synapse
and on to another neuron: This is an algorithmic process.
 neurons

The outcome depends on the connections between

neurons.
These structural relations form a network, that has to be
specified in addition to the properties of the neurons.
Different connections will give different outputs.
 2: Top-down causation via non-adaptive
information control

In non-adaptive information control,

higher level entities influence lower level entities so

as to attain specific fixed goals through the
existence of feedback control loops

Information flows underlie their functioning

The goals determine the outcome

 Feedback control systems and
information
Feedback control (cybernetic systems):

Error message
Controller

Comparator
System State Goals

Examples - the temperature of a shower

- the speed of a steam engine
- a thermostat
It is top-down action because if you disconnect the parts it
won’t work. The system is connected so as to give the outcome.
 The role of goals and information
The series of goals in a feedback control system are causally
effective higher level entities.

They embody information about the system’s desired behaviour or

responses – living systems are goal seeking (‘teleonomic’)

These goals are not the same as material states, for they are
desired rather than actual states, although they will be represented
by material states and systems that will make them causally
effective through such representations

A complete causal description must necessarily take them into

account. They exist as emergent properties of the system – they are
not embodied in any component on its own.
 The role of goals in dynamics
The nature of causality is different when feedback control systems
are guided by goals

Standard Physics
(physics, equations of state, initial conditions) (outcomes)
(initial conditions) (outcomes)

Feedback control systems

(physics, physical structure, goals) (outcomes)
(goals) (outcomes)

• The outcome of a feedback control system is determined by the

goals rather than the initial data
Top-down causation via non-adaptive information control

Homeostasis in the human body:

• Body temperature
• Blood Pressure
• Normal heart rate
• Transport across cell membranes
• Maintenance of resting potential in neurons

- each is governed by implicit goals, embodied in the physical

structure of the body:
‘the human body has literally thousands of control systems in it’
[Guyton]

- They have been built in through the adaptive process of

evolution and so embody images of environment. They are
constant across individuals, time, and place in specific species.
3. Top-down causation via adaptive selection
Adaptive processes take place when many entities interact and
variation takes place in the properties of these entities, followed
by selection of preferred entities that are better suited to their
environment or context.

Higher level environments provide niches that are either

favorable or unfavorable to particular kinds of lower level
entities; those variations that are better suited to the niche are
preserved and the others decay away.

A selection agent or selector accepts one of the states and

rejects the rest on the basis of fitness criteria guiding adaptive
selection. This selected state is then the current system state
that forms the starting basis for the next round of selection. A
different environment will lead to a different outcome.
 Adaptive Selection:
generation of adapted states
with new information encoded

System Selection Agent

State

Variation
Preferred Fitness
System
Criteria
States
State

Environment:
Niches
 Top-down action by adaptive
selection: evolution
Development of DNA codings (the particular sequence of bases in
the DNA) through an evolutionary process which results in
adaptation of an organism to its ecological niche.

This is a classical case of top down action from the environment

to detailed biological microstructure - through the process of
Darwinian adaptation based on random mutations, the
environment (along with other causal factors) fixes the specific
DNA coding. There is no way you could ever predict this coding
on the basis of biochemistry or microphysics alone. You can’t
even ask the appropriate questions in their languages.

This is the way new information comes into biological processes.

It is unpredictable because a random element enters.
 Environment

Animal

DNA sequence

Through natural selection, top-down action from the environment

codes information about appropriate responses to the environment
into the detailed base sequence in the animal’s DNA
Ecological context Coding: CAGTCCTA…

The DNA double helix with complementary base pairs

Neural Networks Training of artificial neural nets to perform a
specific task (say letter recognition) determines the interaction
weights in the network.

The niche is a particular set of letters to be recognised. The fitness

criterion is correct pattern recognition, and the adaptive process is
the training of the neural network.

This is a form of top-down causation from the pattern to be

recognized (a high-level concept, as it is defined in terms of the
relation between the elements) to the low-level property of network
weights.

Decision making is a property of the network rather than of any

single cell.
 Neuronal plasticity is developed adaptively

‘The initial set of relatively non-specific synaptic connections

are refined to produce a precise pattern of connectivity’ [Wolpert]
This kind of refinement takes place in all processes involving neural
plasticity. This is adaptively guided by neurotransmitters on the
basis of our primary emotions [Neural Darwinism].
4. Top-down causation via adaptive control
Adaptive information control takes place when there is adaptive
selection of goals in a feedback control system, thus combining
both feedback control and adaptive selection.

The goals of the feedback control system are irreducible higher

level variables determining the outcome, but are not fixed as in
the case of non-adaptive feedback control; they can be
adaptively changed in response to experience and information
received, in the context of the local environment. They vary
with the individual and with time.

The process is guided by fitness criteria for selection of goals.

Top-down causation via adaptive control
Associative learning in animals, such as Pavlovian conditioning:
- animal response to a stimulus such as a sound, which is taken as
a sign of something else and causes physical reactions
implemented by motor neurons.

The training is causally effective by top-down action from the

brain to cells in muscles. The fitness criterion is avoidance of
negative stimuli. How do you demonstrate this top-down
causation? - change the conditioning, and the response is different.

It is not always predictable: hidden internal variables may change

the outcome (spontaneous alternation in rats; monkey and
bananas) and emergent behaviour may occur.
 5: Intelligent top-down causation
- The effect of the human mind on the physical world.

Intelligent top-down causation is the special case of feedback

control with adaptive choice of goals, where the selection of
goals involves the use of symbolic representation and
manipulation to investigate the outcome of goal choices.

The key feature of this higher level of causation, is its use of

language and abstract symbolism such as mathematics for
reasoning as well as communication, so enabling high-level
culture and technology to arise through coordination and
planning.

Probably only occurs in humans (Deacon: The symbolic species).

Aircraft Design: Plans for a Jumbo Jet aircraft result in billions of
atoms being deployed to create the aircraft in accordance with
those plans. This is a non-trivial example: it costs a great deal of
money to employ experts in aerodynamics, structures, materials,
fuels, lubrication, controls, etc. to design and then to manufacture
the aircraft in accordance with those plans

The plan itself is not equivalent to any single person’s brain state:
it is an abstract hierarchically structured equivalence class of
representations (spoken, drawn, in computers, in brains, etc.) that
together comprise the design.

It is clearly causally effective (the aircraft would not exist without

it). It could not occur without language and mathematics, as well
as thesocial systems in which it is embodied.
Boeing 747-400
A timetable for an airline determines when aircraft fly in a more
or less reliable way. It results in an aircraft flying on a particular
path at a particular time, resulting in particular patterns of
atmospheric pollution through specific molecules.

How do you demonstrate top-down causation? - change the

timetable and different patterns of pollution will result.

Physics can describe the material out of which the timetable is

made and the ink markings on the paper; it cannot comprehend
the causal chain by which this leads to particular aircraft flying
at particular times. The relevant variables (the entries in the
timetable) belong to an irreducible equivalence class of abstract
entities coding information that controls what happens in the real
world. They operate through social convention.
 The nature of goals
Conscious Goals in human activity:

• our actions are governed by hierarchically structured goals at

all structural levels in society
• these may be explicit or implicit, qualitative or quantitative

• they are not physical quantities

• they can be represented in many ways, so are effectively an
equivalence class of representations

• they are adaptively formed in response to experience: learning

takes place in particular contexts
• the mind responds to the meaning of symbols in the relevant
social context
 Hierarchical structure: 2
Cosmology Sociology
Astronomy Psychology
Geology Physiology
Materials Biochemistry

Chemistry
Physics
Particle Physics

Hierarchy of causal relations

* The right hand side involves goals & conscious choices

The Effectiveness of Consciousness
Dimensions of consciousness:
- rationality and understanding
- feelings and intentions
- social systems/constructions, e.g. laws/money

• Concepts are not the same as brain states

- They can be represented in many different ways
• These are all causally efficient: they effect the
nature of physical objects in the world
• These function are based in neuronal structure
 5: The nature of causality
The key point about causality in this context is that simultaneous
multiple causality ( inter-level, as well as within each level) is
always in operation in complex systems.

Any attempt to characterise any partial cause as the whole (as

characterised by the phrase `nothing but') is a fundamentally
misleading position. Indeed this is the essence of fundamentalism:
claiming a partial truth to be the whole truth.

This is important in regard to claims that any of physics,

evolutionary biology, sociology, psychology, or whatever are able
to give total explanations of any specific properties of the mind.
Rather they each provide partial and incomplete explanations.
 Causality: Bottom-up and top-down explanation
There are always multiple levels of explanation that all hold at the
same time: no single explanation
- so one can have a top-down system explanation as well as a
bottom-up explanation, both being simultaneously applicable

e.g. Why aircraft fly [Russell Ackoff]

- the bottom up view: kinetic theory/Bernoulli’s law
- the top down view: it was designed that way
- the same level view: the pilot is flying it to fulfill the timetable
- topmost: it makes a profit for the company

They are all simultaneously true and relevant!

It won’t fly unless they all apply at the same time.
 Causality: Aristotle’s kinds
The material cause: “that out of which”,
The efficient cause: “the primary source of the change or
rest”
The formal cause: “the form”, “the account of what-it-is-
to-be”.
The final cause: “the end, that for the sake of which a
thing is done”

We can adapt Aristotle’s categorisation to the

hierarchical context considered here, by seeing
the Material Cause as the lower level (Physical) cause,
the Efficient Cause as the same level (Immediate) cause,
the Formal Cause as the immediate higher (Contextual)
cause,
the Final Cause as the ultimate higher level cause
Francis Crick famously said

"You, your joys and your sorrows, your memories and your
ambitions, your sense of personal identity and free will, are in fact
no more than the behavior of a vast assembly of nerve cells and
their associated molecules".

But nerve cells and molecules are made of electrons plus protons
and neutrons, which are themselves made of quarks .. so why not

"You, your joys and your sorrows, your memories and your
ambitions, your sense of personal identity and free will, are in fact
no more than the behavior of a vast assembly of quarks and
electrons"?

And these themselves are possibly vibrations of superstrings.

So why did Crick stop at the level he chooses? Undoubtedly because
that is the level he best understood and was familiar with!

• Indeed each scientist will perceive as fundamental the level they

happen to work on and understand deeply in causal terms, so they
usually assume that causality at that level is real. And that is
reasonable, if they are all real, as I take to be the case (a table is
still a table even though it is made of atoms, for example; and the
atoms are also real, as are the neutrons and protons).

• Crick’s dictum either applies to all levels except the (unknown)

bottommost one, or to none. If it applies to all levels, Crick’s
molecules are no more real than memories and ambitions; but he
assumes the molecules are real, so his position is inconsistent.

• There is no reason to privilege molecules or cells in the hierarchy

of structure. If we accept molecular reality, then we should also
acknowledge the memories and ambitions as real too, for that is
then the only consistent position.
All Truth is shadow except the last, except
the utmost; yet every Truth is true in its
own kind. It is substance in its own place,
though it be but shadow in another place
(for it is but a reflection from an intenser
substance); and the shadow is a true
shadow, as the substance is a true
substance.

Isaac Pennington (1653).

6: Predicting and intervening
KEY QUESTION:

What level to work at?

The most powerful interventions are at the highest

levels

Multiple realisability at lower levels: will return with

new plans for same intention

True security: Turning an enemy to a friend

-else they may always re-arise (Al Qaeda)
1. Algorithmic top-down causation

Prediction: Calculations can in principle predict

Possibility space:
Understand attractors and basins of attraction

Affecting it: Change viability/mode of lowest levels:

e.g. cut off supplies,

drugs,
poison
2. Top-down causation via non-adaptive information
control

Goals determine outcome

Bottom up: Can try to set system outside its viable

parameter levels

Same level: Try to change the goals, if possible

Else adapt them to your purposes

(coopt predictable behaviour)
3. Top-down causation via adaptive selection

Unpredictable element; prediction difficult or impossible

But can try to understand basins of attraction or

convergence in possibility space

- Convergence due to limits on how things can be done:

Conway Morris

Change of context changes adaptive behaviour (alters

niches)
4. Top-down causation via adaptive information control

Again unpredictable: only meta-cause constant

Can try to change environment so that adaptation goes the

way you want

e.g. Pollution and rivers [People or Penguins]

5. Intelligent top-down causation (the effect of the human
mind on the physical world)

Key elements: Purpose and meaning, search for

understanding
- e.g. seeing distant person: scan for purpose

That is the best level to intervene

BUT understand The myth of rational behaviour

The real roots of behaviour: emotion and intuition

- these are the guides of rational behaviour:
Determine choices of what to pay attention to
 The Mind

Rationality Faith, Hope

The individual mind: Each of Emotions, Rationality,

Faith/Hope,, Aesthetics, Ethics, and Telos are causally
effective. They are modulated by the society in which we live:
they cannot be understood in isolation.
We have the freedom to choose how to act in this context.
 The Mind

Rationality Faith, Hope

Emotion
Primary Secondary:
Genetic/biological Social/cultural

The individual mind: Each of Emotions, Rationality,

Faith/Hope,, Aesthetics, Ethics, and Telos are causally
effective. They are modulated by the society in which we live:
they cannot be understood in isolation.
We have the freedom to choose how to act in this context.
 The Mind

Perception, Expectations

Rationality Faith, Hope

Intuition, Imaginatiom

Emotion
Primary Secondary:
Genetic/biological Social/cultural

The individual mind: Each of Emotions, Rationality,

Faith/Hope,, Aesthetics, Ethics, and Telos are causally
effective. They are modulated by the society in which we live:
they cannot be understood in isolation.
We have the freedom to choose how to act in this context.
 Ethics, Aesthetics, Meaning

Perception, Expectations Society

Rationality Faith, Hope
Intuition, Imaginatiom

Emotion
Primary Secondary:
Genetic/biological Social/cultural

The individual mind: Each of Emotions, Rationality,

Faith/Hope,, Aesthetics, Ethics, and Telos are causally
effective. They are modulated by the society in which we live:
they cannot be understood in isolation.
We have the freedom to choose how to act in this context.
 7: Conclusion
Other kinds of entities than simply physical exist, as well as other
kinds of causation.

If we ignore them we will have a causally incomplete view of the

universe.

In particular, top-down processes underlie the possibility of truly

complex emergent phenomena such as the workings of the mind
and culture.

Intervention must understand levels of causation/control, and their

interactions (top down, same level, bottom up).
Womans head (leonardo)
Social environment: Genetic inheritance:
- effect of society - effect of biology

Mind,
underlying
consciousness
and personality

Personal Choice:
- causal effectiveness of consciousness/will

The nature-nurture issue: three main factors

that contribute to the development of the mind
 8: Transcendence
This multi-faceted nature of causation allows various
non--physical entities to be causally efficacious in the
physical world, through their effect on the human
mind.

Some of these can be said to be of a transcendent

nature: they are of a different order than, and
independent of the nature of, the physical matter out
of which we are made.

These features are NOT emergent, they are eternal and

unchanging. It is our understanding of them that is
emergent and developing with time.
1: Mathematics Comprehension and Utilisation is a case of
top-down causation from a Platonic world of mathematical
abstractions to the human mind, being realized in details of
neuronal connections, and then into the real world where it is
causally effective in terms of creating patterns on paper and
through underlying physics, engineering, commerce, and
planning in general

Major parts of mathematics are discovered rather than invented

(e.g. irrational numbers). They have an abstract rather than
embodied character; the same abstract quantity can be
represented in many symbolic and physical ways, and these
representations form an equivalence class. They are
independent of the existence and culture of human beings.

Jean-Pierre Changeux and Alain Connes: Conversations on

Mind, Matter, and Mathematics
 The basic geometrical features

R = radius
H = hypotenuse

Y= A = area
height

X = base
A = π R2

π = 3.1415926535897932 ……..
H2 = X2 + Y2
(Pythagoras) (universal constant)

The same results will be discovered near Alpha

Centauri or the Andromeda Galaxy
Mandelbrot set

Mandelbrot set
2: A second example is language:

Terence Deacon has plausibly argued that any language system

must obey necessary semiotic constraints on universal
grammar;
these too are discovered by humans rather than being created,
because they are based in the underlying logic of what is
possible in symbolic representation of the real world.

“Semiotic constraints delimit the outside limits of the space of

possibilities in which language has evolved within our species,
because they are the outside limits of any symbolic form of
communication”
[Deacon: “UG and Semiotic constraints”, in Language Evolution,
Ed Christiansen and Kirby]
“So perhaps the most astonishing implication of this
hypothesis is that we should expect that many of the
core universals expressed in human language systems
will of necessity be embodied in any symbolic
communication system, even one used by an alien
race on some distant planet!” [Deacon].

3: Logic itself is a third example

It is universal and discovered.
- underlies all the rest (e.g. mathematics, semiotics).
- we take it for granted in all our thought and embed it
in our language.
4: Physics Theories: Maxwell’s theory of electromagnetism (an
abstract entity, described by Maxwell’s equations) led to the
development of radio, cell phones, TV, and so on. It is shown to
be true by experiments and by its technological outcomes.

Maxwell’s theory is not the same as any single person’s brain state.
It can be represented in many ways and formalisms
These various representations together form an equivalence class, as
they all lead to the same predicted outcomes.

The abstract theory has altered physical configurations in the real

world, and hence is causally effective. It is an irreducible higher
level variable (it cannot be derived by coarse-graining any lower
level variables) representing the nature of physical reality. It is
based in the nature of matter but is itself a product of the human
mind.
5: The functional pre-requisites of a society
- A common language and associated set of understandings
- A social organisation centered around agreed social roles
- A method of sharing out and controlling available resources
- A system of production and consumption of resources
- A set of agreed norms and values to govern behaviour
- A set of sanctions for those violating these norms
- A method of training youth to be adult members of society

All societies must tackle these same set of problems and devise
means of dealing with them.
There is a limited set of ways of doing so. One or other of them
will be discovered by each society.
6: Ethical values and Morality

The highest level of intention [values/ethics] is causally effective:

this is the choice of criteria for what are acceptable goals,
and so controls all lower level goal choices

Ethics

Cosmology Sociology
Astronomy Psychology
Geology Physiology
Materials Biochemistry

Chemistry
Physics
Ethics is causally effective in the physical world
9: Causal openness for higher causes is there because

1: The functioning (according to the laws of physics) of the

parts of given nature is determined by context.

Higher level purposes can conscript physics to its ends by

changing the operating context/altering context dependent
constraints. Structure and boundary conditions crucially
affect outcomes. This is an essential part of top-down and
same-level action.

Example 1: a computer operated as music system or word

processor.

Example 2: muscle cells being used in football or in playing

music.
Causal openness for higher causes is there because

2: We do not just have invariantly functioning parts

assembled in different ways and so operating in
different contexts.

The nature of the parts – the way they function – is also

affected by context in a top-down way. Indeed they are
often adapted to their higher level function.

Example 1: Living cells (through developmental biology).

Example 2: Neutrons in an atomic nucleus.

Example 3: Humans in society

Causal openness for higher causes is there because

3: Chance (statistics associated with coarse graining, random

boundary conditions) means physical outcome in
biological systems is not uniquely determined by physics
alone.

This provides the openness needed for Darwinian selection

processes to choose outcomes that satisfy higher level
goals and values.

This top-down mechanism may be far more prevalent than

recognized up to now in the developmental and functional
contexts as well as the evolutionary context.

Example: Adaptive immune system.

4: Quantum Uncertainty means the physical outcome is
not uniquely determined even in principle.

This is not always negligible at the macro level,

despite occurring at a micro-level.

Example: evolutionary history on earth is influenced

by cosmic rays, that are emitted subject to quantum
uncertainty

For biological functioning, see P W Glimcher,

“Indeterminacy in brain and behaviour”.
Annual Review of Psychology, 56: 25 (2005) .