Autonome Robotica
Experiments in Synthetic Evolutionary Psychology
Walter
de Back
Department of Philosophy &
Robotics Lab - Institute for Computing Sciences
Utrecht University
contact info
On these pages you find general information about doing experiments in synthetic
evolutionary psychology, written for students and experts. It is part of
a project to create a platform for research in synthetic evolutionary psychology,
in a share-ware simulator called Framsticks.
New: See the website of the newly established Virtual Life lab.
Synthetic Evolutionary Psychology (SEP) is the use of computer simulations
and autonomous (evolutionary) robotics to
investigate the history and structure of the brain and mind. The rationale
behind, and the methods of SEP are outlined in a recent paper by Dylan Evans & Walter
de Back, 2003 (forthcoming). See abstract.
Instead
of using analytical methods, in which data is extracted from already existing
systems, we propose to extend the methodology to include synthetic methods,
which involves contructing artificial systems. These artificial systems can
provide useful models for evolving minds and evolutionary histories that might provide evolutionary psychologists
with additional means to test their hypotheses about mental structure and
evolutionary trajectories.
A well-known
synthetic method that has successfully been deployed for this purpose before,
is game-theory. This has the main benefits of (1) forcing theorists to be
explicit, and (2) serving as a inference-machine (by exploring the workings
of a model beyond human capabilities). Game-theory has some serious drawbacks
as well, the main problem being the lack of embodiment and situatedness.
To overcome this problem, there is one kind of synthetic method we advocate
in particular: Evolutionary Robotics (ER). ER is an approach to autonomous
robotics in which neural network robot controllers are evolved by applying
artificial evolution.
The specific
project I propose is to use an A-life simulator Framsticks to create a platform
for research in synthetic evolutionary psychology. With the use of this program,
we can evolve neural network controllers for simulated autonomous robots by
open-ended artificial evolution.
Contents
1. What is Synthetic Evolutionary Psychology?
1.1 What is Evolutionary Psychology?
1.2 Synthetic approach in EP
1.3 What is Evolutionary Robotics?
What is Synthetic Evolutionary Psychology?
"Synthetic Evolutionary Psychology" is the name for the use of computer simulations and evolutionary robotics
to test hypotheses about the history of the (human) mind.
Evolutionary psychology (EP) is the branch of psychology
that approaches the human mind as a product of evolution. Actually, EP is not a branch
of psychology but a perspective through which to examine the subjects of
psychology - behaviour, cognition, intelligence etc. It is grounded in a
number of other disciplines: cognitive psychology, evolutionary biology,
neuro-psychology and anthropology. It examines how Darwinian evolution could
have produced the mental faculties which are exhibited by humans today. EP
builds upon ideas to be found in the discipline of sociobiology.
Some researchers in EP focus on the structure (modularity) of the mind,
others focus on the evolutionary history that shaped the mind. In any case,
hypotheses about the human mind that are derived from the general natural
selection theory remain 'mere hypotheses' until some kind of emperical testing
can be conducted. This is exactly what we propose and promote by coining the
term and field of interest: Synthetic Evolutionary Psychology.
The term is a combination of "Evolutionary Psychology" and "Synthetic Psychology".
The latter term was adopted from the subtitle of a booklet by Valentino
Braitenberg: "Vehicles" (1986). In this booklet, Braitenberg showed the
logic of the basic layout of brains based on two principles: (1) Complex
behaviour emerges from simple interaction between creature and its environment,
and (2) Mental states (such as beliefs and desires) are in the eye of the
observer and not a property of the creature itself. These ideas are the
foundations of the relevatively new science of Autonomous Robotics.
In Synthetic Evolutionary Psychology, we try to show the logic of evolution-designed
brains by synthesizing these creatures, evolutionary forces and brains.
In a recent article, Dylan Evans
(Bath University - UK) and Walter
de Back (Utrecht University, NL) promote this line of research in order
to subject hypotheses from evolutionary psychologists to empirical experimentation.
See also:
Abstract article:
"Evolutionary psychology is an approach to the study of the mind
based on principles drawn from evolutionary biology. In their research
so far, evolutionary psychologists have used many different methods, from
experimental manipulation of human behaviour in the laboratory to observation
of indigenous peoples and analysis of archaeological data. All
these methods may be called analytic, in the sense that they collect data
about already-existing systems and then analyse them. Here we
propose that evolutionary psychologists could benefit from extending their
methodological repertoire to include synthetic methods, which involve constructing
artificial systems. Such artificial systems can provide useful
models of evolved minds and evolutionary histories that might provide evolutionary
psychologists with additional means to test their hypotheses about mental
structure and evolutionary trajectories. One kind of synthetic method
that evolutionary psychologists have so far shown little interest in is evolutionary
robotics. We argue that, by ignoring this field, evolutionary psychologists
are missing out on a valuable research tool, and sketch out a research program
involving the use of robots to test evolutionary psychological hypotheses."
What is Evolutionary Psychology?
Simply
put: "Evolutionary psychology is the combination of two sciences -- evolutionary
biology and cognitive psychology".
from Introducing Evolutionary Psychology,
Dylan Evans & Oscar Zarate
"The goal of research in
evolutionary psychology is to discover and understand the design of the
human mind. Evolutionary psychology is an approach to psychology, in which
knowledge and principles from evolutionary biology are put to use in research
on the structure of the human mind. It is not an area of study, like vision,
reasoning, or social behavior. It is a way of thinking about psychology
that can be applied to any topic within it."
"In this view, the mind is a set of information-processing machines
that were designed by natural selection to solve adaptive problems faced
by our hunter-gatherer ancestors. This way of thinking about the brain,
mind, and behavior is changing how scientists approach old topics, and opening
up new ones."
Personally, I do not fully
agree with this definition. Viewing the mind as a set of information-processing
modules, is a concept derived from traditional cognitive psychology. In
my own view, the mind is not a problem-solving device. Rather, as is wonderfully
illustrated by Valentino Braitenberg, the brain is a complex structure of
sensor-motor couplings that enables creatures to engage in dynamics with
their environment (to be able to reproduce).
For me, then, Evolutionary
Psychology is about discovering how (ancient) environmental, social and
other evolutionary forces has shaped the human brain, to which we can ascribe
mental states, to become a human mind.
See also:
Synthetic approach in evolutionary psychology
Synthetic methods are contrasted with analytic ones. In the latter, one
acquires information from already-existing systems. In evolutionary psychology,
that can be from modern human beings (demographic data, wason selection task),
ancient humans (skulls, archeological sites), comparative ethology (chimpansees).
Synthetic methods, by contrast, studies and tests hypotheses by constructing
artificial systems. This can be (and is) done in several ways: game-theoretical
computer simulations, autonomous robotics (and evolutionary
robotics in particular).
First, we examine a classic example of a game-theoretical computer simulation.
Computer simulations are used in most sciences. As Dennett has put it: "Computer
modelling keeps the theorist honest", by forcing him to formalize and being explicit. Even a false
hypothesis is better then one that is so vague that it is not even false.
The best known example of computer simulation used in evolutionary psychology
is the Prisoner's Dilemma (PD) by Axelrod. The PD is a game-theoretical
approach to the emergence of cooperation and altruism.
"The game got its name from the following hypothetical situation: imagine
two criminals arrested under the suspicion of having committed a crime together.
However, the police does not have sufficient proof in order to have them
convicted. The two prisoners are isolated from each other, and the police
visit each of them and offer a deal: the one who offers evidence against
the other one will be freed. If none of them accepts the offer, they are
in fact cooperating against the police, and both of them will get only a
small punishment because of lack of proof. They both gain. However, if one
of them betrays the other one, by confessing to the police, the defector
will gain more, since he is freed; the one who remained silent, on the other
hand, will receive the full punishment, since he did not help the police,
and there is sufficient proof. If both betray, both will be punished, but
less severely than if they had refused to talk. The dilemma resides in the
fact that each prisoner has a choice between only two options, but cannot
make a good decision without knowing what the other one will do."
"Game theoretical approaches to model behaviour, as described above, can
help to illuminate the benefits of adopting certain strategies in particular
(social) situations. However, it is not very effective in pointing out how
these strategies are implemented in the situated, embodied minds of natural
animals. Despite, for example, the many attempts to confirm the fact that
animals play Tit-for-Tat, hardly any evidence for it has been accumulated.
This is caused by the inherent simplicity of the games. Generally, in game
theory, the benefits and costs for actions are precisely defined, whereas
in nature these values are in the eye of the beholder and immeasurable.
Furthermore, these games view just one isolated behaviour observed in nature
and studied it as an all-or-nothing concept (e.g. to co-operate or not),
whereas behaviour in real animals is heavily related to other behavioural
activities and the environment. Investigating the origins of
behaviour from a game theoretical perspective is much like studying chess
in order to investigate intelligence, as is done in traditional AI (Hemelrijk
1997) . Both explore cognitive reasoning abilities without providing the
behavioural fundaments on which it is based.
Since the late 1980s, a growing number of cognitive scientists have become
critical of the traditional focus of their discipline on abstract reasoning,
arguing that minds are always situated in bodies and worlds, and cannot
be understood apart from them (Clark 1997) . The body, mind
and environment are continuously involved in intricate complex dynamics,
which adds to the complexity of understanding minds by analytic reverse engineering.
Autonomous robotics offers the unique possibility of building an artificial
mind inside an artificial body that operates inside a real environment. This
provides psychology and cognitive science with a powerful tool for testing
hypotheses through synthesis of artificial autonomous agents (Pfeifer and
Scheier 1999)."
To recap, we promote the use of computer simulations in evolutionary
psychology because of the fact it forces one to be explicit. However, computer
simulations of the (game-theoretical) kind described above have serious
limitations. Therefore, we are especially interested in the use of autonomous
robotics (and, even more specific, evolutionary robotics)
to test hypotheses from evolutionary psychology. This provides us with the
unique opportunity to build or evolve brains for autonomous embodied systems
(such as ourselves) guided by evolutionary forces.
See also:
What
is Evolutionary Robotics?
Before answering this
question, let's start with autonomous robotics. Autonomous robotics is an
interdisciplinary study to construct and program robots that are able to
operate in the real (physical) world without human intervention. To do this,
they have to be able to autonomously sense their environment and react to
the (changes in the) environment in appropriate ways. These robots are equipped
with artificial 'brains': programs that control the way in which sensations
are converted into actions.
These programs can be implemented very differently. In vacuum tubes ###Grey
###
by an engineer putting all the appropriate knowledge and reasoning-rules
into the robot. Or, as Brooks proposed, However, it is also possible
to replace these hand-coded programs with Neural Networks. These networks
can be trained by experience. This is known as adaptive robotics.
One step further towards true self-organisation is to let these neural
networks be optimised through genetic algorithms. A neural network can be
coded onto an 'artificial DNA-string' composed of zeros and ones. A genetic
algorithm starts out by randomly generating such strings (that code for
neural networks). Then, all of these random network are tested in robots
as controllers. Most of these robots wouldn't do anything good, but some
always do better then others. The better are selected for reproduction.
This process of testing, selection and reproduction is continued until we
have robots controlled by neural networks that meet our predefined criteria.
This method of making 'brains' for autonomous robots is known as evolutionary
robotics.
So, to answer the question above: evolutionary robotics is a self-organisationary
method for making brains for autonomous robots by applying artificial evolution
on neural network controllers.
Drawbacks
A well-known and
important difference between artificial and natural evolution is that the
former is goal-directed, while the latter is open-ended. Artificial evolution
is a technique originally devised to optimise parameters that is inspired
by the optimisation property observed in nature. Natural evolution, in contrast,
does not converge to a single solution and then stop: it is open-ended and
continuous. There are no optimal solutions in nature because the problems,
arising from the biological context which includes many co-evolving creatures,
are constantly changing. Evolutionary psychologists are interested in the
mechanisms that result from natural, and therefore open-ended, evolution.
A synthetic approach to evolutionary psychology should attempt to replicate
this process.