Machines in the ghost
1 like1,115 views
This document discusses the idea that natural minds are information-processing virtual machines that have been produced by evolution. It argues that to truly understand natural minds, we need to describe them with sufficient precision to enable the design of artificial minds. The key challenges are determining what concepts to use in theories of the mind and deciding whether producing artificial minds similar to natural ones will require new computing machinery. The document also discusses how virtual machines in computers can have causal powers and how psychotherapy can be viewed as debugging the "virtual machine" of the mind.
![of them (and thereby help to define the concepts), in order between theoretical concepts and observable phenomena is
to obtain deep explanations of what we can observe and new referred to as ‘symbol attachment’ in [3] and ‘symbol
more reliable and more precise predictions, regarding a wider tethering’ in [4]. Once this possibility is understood, the need
range of phenomena. In particular, in physics, chemistry and for so-called ‘symbol grounding’ (deriving all concepts from
biology, new theories have taught us that however useful our experience) presented in [5], evaporates. (The impossibility of
ordinary concepts are for ordinary everyday social interactions, ‘concept empiricism’ was demonstrated by Kant [6] over 200
they are often grossly inadequate: (a) for distinguishing years ago.)
all the phenomena that need to be distinguished, (b) for The ontology used by an individual or a community (i.e.
formulating precise and reliable predictions, and (c) for the set of concepts used to describe things in the world) can
describing conditions (often unobservable conditions) that are be extended in two ways, either by definitional abbreviation
the basis for reliable predictions. or substantively. A definitional abbreviation merely introduces
So it may be unwise to go on using the old concepts a new symbol as a short-hand for what could be expressed
of folk psychology when describing laboratory experiments previously, whereas substantive ontology extensions introduce
or field observations, when formulating descriptions of what new concepts that cannot be defined in terms of pre-existing
needs to be explained, and when formulating supposedly concepts. Such concepts are implicitly defined mainly by their
explanatory theories. But choosing alternatives to ordinary role in explanatory theories, as explained above. So substantive
language needs great care. Recently attempts have been made ontology extension always requires theory construction, as has
to give these concepts scientific status by using modern happened many times in the history of science and culture.
technology to identify precise brain mechanisms, brain states,
brain processes that correlate with the states and processes
described in ordinary language. Compare trying to identify the II. V IRTUAL AND PHYSICAL MACHINES
precise location in a particular country where some national
We can do for the study of mind and brain what was
characteristic or process such as religious bigotry, scientific
previously done for the study of physical matter and biological
ignorance, or economic inflation is located.
processes of evolution and development, if we understand that
The ‘design-based’ approach minds and brains are not just matter-manipulating, or energy-
Is there any alternative to going on using pre-scientific contexts manipulating machines, but information-processing machines.
in our descriptions and theories? Yes, but it is not a simple Minds are information processing virtual machines while
alternative. We need to adopt what Dennett [1] calls ‘the brains are physical machines, in which those virtual machines
design stance’, which involves constructing theories about how are implemented or some would say realised.
minds and brains actually work, which goes far beyond what In computing systems we also have virtual machines,
we either experience of their working in ourselves or observe such as running operating systems, firewalls, email systems,
in others. (My own early attempts at doing this 30 years ago spelling correctors, conflict resolution mechanisms, file
are presented in [2].) Moreover, our theories must account for optimisation mechanisms, all of which run on, i.e. are
the existence of many kinds of minds, with different designs. implemented in, the underlying physical hardware. There are
In particular we can’t just start defining precise new no simple mappings between the components of the virtual
explanatory concepts in terms of precise measurements and and the physical machines. There are often several layers
observations. Thinking that definitions come before theories is of implementation between high level virtual machines and
a common mistake. In the more advanced sciences, concepts physical machines. This provides enormous flexibility for re-
are used (e.g. ‘electron’, ‘charge’, ‘atomic weight’, ‘valency’, use of hardware, both in different systems using the same
‘oxidation’, ‘gene’, etc.) that cannot be defined except by hardware, and in one system performing different functions
their role in the theories that employ them. The structural at different times. It seems that evolution discovered the
relations within the theory partially define the concepts. That power of virtual machines before we did and produced brains
is because a theory is a formal system, and, as is familiar from implementing hierarchies of mental (and indirectly social)
mathematics, and made more precise in the work of Tarski, virtual machines.
the structure of a formal system determines which things The various components of a complex virtual machine
are possible models of that system. Usually there are many (such as an operating system distributed over a network
possible models, but the set of possible models can be reduced of processors) need not run in synchrony. As a result,
by enriching the theory, thereby adding more constraints to be timing relations between processes can change from time
satisfied by any model. This may still leave different models, to time. For this and other reasons, Turing machines do
of which only one, or a subset is intended. Such residual not provide good models for minds, as explained in [7].
ambiguities are reduced (but never completely removed) by Moreover if sophisticated memory management systems are
links between the theory and methods of observation and used the mappings between components of virtual machines
experiment, and practical applications associated with the and physical parts of the system can also keep changing.
theory. However those links do not define the concepts, If such complex changing relationships are useful in systems
because old methods of observation and old experimental we have designed and implemented we need to keep an open
procedures can be replaced by new ones, while the old mind as to whether evolution, which had several billion years
concepts endure. This loose, theory-mediated, connection head start on us, also discovered the power and usefulness of](https://image.slidesharecdn.com/machinesintheghost-120102113734-phpapp02/85/Machines-in-the-ghost-2-320.jpg)
![division, and giving counselling to enable a patient to values. Isn’t that what happens to humans? This point was
understand how he unintentionally causes family rows to made long ago in section 10.13 of [2]. So motivation in
escalate, differ in detail from the process of fixing a typical artificial virtual machines, including self-generated motivation,
bug in a computer program. For example the teaching and is not a problem in principle.
therapy depend on kinds of self understanding that few
The myth that intelligence requires emotions
computer programs have at present. But in future it will
In recent years, especially following publication of Damasio’s
not be uncommon to find virtual machines, that, instead
[8] and Picard’s [9], much has been made of the alleged need
of being forcibly altered by a user editing rules, instead
to ensure that intelligent systems have emotions. I have argued
change themselves as a result of being given advice about
elsewhere that the arguments are fallacious for example in [10]
how to behave. That’s not even science fiction: it is not
[11] [12] [13] [14] and [15]. Some of these claims are merely
hard to achieve. This comparison should undermine three
poorly expressed versions of Hume’s unsurprising observation
assumptions: (a) the belief that the only way to understand
that without motives an intelligent system will have no reason
and fix problems with minds is to work in a totally bottom
to do anything: this is just a confusion between emotions and
up way, namely understanding and modifying brains, (b) the
motives.
belief that nothing we know about computers is relevant to
There is also a more subtle error. 26 years ago, [10]
understanding and repairing minds, and (c) the belief that
argued that intelligent machines need mechanisms of kinds
machines can only be programmed to do things, not advised,
that perform important functions and which in addition can
inspired, or instructed or cajoled into doing them.
sometimes generate emotional states and processes as side
Motivation in virtual machines effects of their operation, if they lack sufficient processing
So, states, events, and processes occurring in virtual machines power to work out what needs to be done. This does not imply
can have causal influences which alter both other things in that they need emotions,4 . just as the fact that some operating
the virtual machine and also the physical behaviour of the systems need mechanisms that are sometimes capable of
system (e.g. physical events in memory, hard drives, internal generating ‘thrashing’ behaviour does not imply that operating
interfaces, what is displayed on a screen, the sounds coming systems need thrashing behaviour. Desirable mechanisms can
from speakers, or various attached motors, etc.). However, we sometimes have undesirable effects. There can be particular
still need to understand the variety of kinds of causation in sorts of situations where an ‘alarm system’ detects a putative
virtual machines. need to override, modulate, freeze, or abort some other
In very simple computer models nothing happens until a process, and where because of shortage of information or
user gives a command, which can then trigger a cascade shortage of processing power, a rough and ready rule operates.
of processes. In more sophisticated cases the machine is It would be better if the situation could be fully evaluated
designed so that it initiates various activities from time and reasoned about (without any emotions), but if there is
to time, e.g. checking for email, checking whether disks inadequate capacity to do that in the time available then it
need defragmenting, checking whether current scheduling may be better to take the risk of false alarms, especially if
parameters need to be revised in order to improve processing the alarm system has been well trained, either by evolution or
performance. It can also be designed so that events initiated individual learning, and does not often get things wrong.
from outside trigger new internal processes. E.g. a user The need for therapy to undo bad learning
attempting to access a file can trigger a sub-process checking However the danger in having such powerful subsystems that
whether the user has the right to access that file. All of can change as a result of learning is that they may change in
these cases require the designers of the virtual machine to bad ways – as clearly happens in some humans. In some cases,
anticipate kinds of things that might need appropriate checking it may be possible to undo bad changes by re-programming,
or corrective actions to be performed. e.g. through discussion, advice, teaching, re-training, therapy,
But there is no difficulty in building a machine that acquires etc. In other cases help may be available only when it is
new competences while it is running, and also new conditions too late. Of course there are also cases where such systems
for exercising old competences. If the virtual machine go wrong because of physical damage, disease, malfunction,
architecture allows new goal-generators to be acquired, new corruption, etc., which may or may not be reversible, e.g.
strategies for evaluating and comparing goals, new values to chemical addictions.
be employed in such processes, then after some time, the
machine may have goals, preferences, intentions, etc. that were III. T HE VARIETY OF MENTAL VIRTUAL MACHINES
not given to it by anyone else, and which, as remarked in
There is not just one kind of mind. Insect minds are different
[2], can only be described as its own. It would be, to that
from minds of birds and monkeys. All are different in many
extent, an autonomous machine, even if the processes by which
ways from adult human minds. Human minds are different
such motives and motive-generators were acquired involved
at different stages of development: a newborn infant, a nine-
being influenced by things said and done by other people, for
month old crawler, a two-year old toddler, a four year old
instance teachers, heroic figures, and other role models. Indeed
talker, a 50 year old professor of psychiatry. Apart from
a machine might be designed specifically to derive motives,
values, preferences, etc. partly on the basis of such influences, 4 As argued in this slide presentation: http://www.cs.bham.ac.uk/
tempered by experimentation on the results of trying out such research/projects/cogaff/talks/#cafe04](https://image.slidesharecdn.com/machinesintheghost-120102113734-phpapp02/85/Machines-in-the-ghost-4-320.jpg)
![differences resulting from development and learning there are real robots, but that what the researchers were trying to get
differences that can be caused by genetic deficiencies, and by their systems to do fell very far short of what they implied
brain malfunctions caused by disease or injury. What’s more in their predictions and promises would be achieved, because
obvious is that the brains are different too. So we need to find a they had not analysed the requirements adequately.
general way of talking about different minds, different brains, Moreover, it could turn out that all of the currently
and the different kinds of relationships that hold between (a) understood models of computation are inadequate, and entirely
the minds, i.e. the virtual machines, and (b) the brains, i.e. the new kinds of virtual machine are needed, including machines
physical machines in which they are implemented. that grow their own architecture, as suggested in [3].
Even virtual machines have architectures
We are talking about a type of complex system with many IV. W HAT IS INFORMATION ?
concurrently active parts that work together more or less There are many questions still to be answered about the
harmoniously most of the time but can sometimes come concept of information used here. What is ‘information’? What
into conflict. These parts are organised in an information- is an information-user? What is involved in understanding
processing architecture that maps onto brain mechanisms in something as expressing a meaning or referring to something?
complex, indirect ways that are not well understood. So we Is there a distinction between things that merely manipulate
should ask questions like this if we wish to do deep science symbolic structures and things that manipulate them while
studying a particular kind of mind: understanding them and while using the manipulation to derive
• What sorts of component parts make up the architecture of new information? In how many different ways do organisms
this sort of mind? acquire, store, derive, combine, manipulate, transform and use
• What are their functions? information? How many of these are, or could be, replicated
• For each such component, what difference would it make if in non-biological machines? Is ‘information’ as important a
it were modified or removed, or connected in a different way concept for science as ‘matter’ and ‘energy’, or is it just
to other components? a term that is bandied about by undisciplined thinkers and
• Which parts of the architecture are involved in various popularists? Can it be defined? Is information something that
processes that are found to occur in the system as a whole? should be measurable as energy and mass are, or is it more
• Which parts are connected to which others, and how do like structure, which needs to be described not measured (e.g.
they interact? the structure of this sentence, the structure of a molecule, the
• What kinds of information do the different parts acquire structure of an organism, the properties of a toroid)?
and use, and how do they obtain the information? We currently have the paradoxical situation that philoso-
• How is the information represented? (It could be phers who are good at conceptual analysis are badly informed
represented differently in different subsystems). about and often have false prejudices about computing
• Can the system extend the varieties of information contents systems, whereas many software engineers have a deep but
that it can make use of (extend its ontology), and extend the unarticulated understanding, which they use very well when
forms of representation that it uses? designing, implementing, testing, debugging, maintaining,
• What is the total architecture in which they function, and virtual machines, but which they cannot articulate well because
how is it made up of sub-architectures? they have not been taught to do philosophy.
• How are the internal and external behaviours se- Information-theory is not about information!
lected/controlled/modulated/coordinated? The word ‘information’ as used (after Shannon) in so-called
• Can conflicts between subsystems arise, and if so how can ‘information theory’ does not refer to what is normally
they be detected, and how can they be resolved? meant by ‘information’, since Shannon’s information is a
• What mechanisms in virtual machines make those processes purely syntactic property of something like a bit-string, or
possible, and how are they implemented in brains? other structure that might be transmitted from a sender to
• In how many ways can the different virtual machines either a receiver using a mechanism with a fixed repertoire of
individually, or through their interactions go wrong, or produce possible messages. Having that sort of information does not,
dysfunctional effects? for example, allow something to be true or false, or to
• How did this virtual machine evolve, and what does it contradict something else. However, the more general concept
have in common with evolutionary precursors and with other of information, like ‘mass’, ‘energy’ and other deep concepts
contemporary animal species? used in scientific theories, is not explicitly definable. That is
Answering these, and similar, questions requires a long term to say, there is no informative way of writing down an explicit
investigation. One of the reasons why optimistic predictions definition of the form ‘X is Y’ if X is such a concept. All you’ll
regarding imminent successes of Artificial Intelligence (or end up with is something circular, or potentially circular, when
more recently robotics) have repeatedly failed is that the you expand it, e.g. ‘information is meaning’, ‘information is
phenomena we are trying to explain and to replicate, are far semantic content’, ‘information is aboutness’, ‘information is
more complex than anyone imagines. I.e. the main reason is what is expressed by something that refers’, ‘information is a
NOT that the wrong programming languages, or the wrong difference that makes a difference’ (Bateson), and so on.
models of computation were used (e.g. symbolic vs neural), But that does not mean either that the word is meaningless
or that the test implementations used simulations rather than or that we cannot say anything useful about it. The same is true](https://image.slidesharecdn.com/machinesintheghost-120102113734-phpapp02/85/Machines-in-the-ghost-5-320.jpg)
![of ‘energy’. It is sometimes defined in terms of ‘work’, but though U has never encountered S or anything with similar
that eventually leads in circles. So how do we (and physicists) information content, for instance when U encounters a new
manage to understand the word ‘energy’? sentence, diagram or picture for the first time. Even before
The answer was given above in Section I: we understand any user encounters S, it is potentially usable as an information
the word ‘energy’ and related concepts, by understanding their bearer. Often, however, the potential cannot be realised without
role in a rich, deep, widely applicable theory (or collection U first learning a new language, or notation, or a new theory
of theories) in which many things can be said about energy, within which the information has a place, and which provides
e.g. that in any bounded portion of the universe there is a substantive ontology extension for U, as discussed in [16].
scalar (one-dimensional), discontinuously variable amount of This may be required for growth in self-knowledge too.
it, that its totality is conserved, that it can be transmitted in A user with appropriate mechanisms has potential to
various ways, that it can be stored in various forms, that it can derive infinitely many distinct items of information from a
be dissipated, that it flows from objects of higher to objects small structure, e.g. infinitely many theorems derivable from
of lower temperatures, that it can be used to produce forces Peano’s five axioms for arithmetic. Physically quite small
that cause things to move or change their shape, etc. etc. As objects can therefore have infinite information content, in
science progresses and we learn more about energy the concept combination with a reasoning mechanism, though limitations
becomes deeper and more complex. The same is happening of the implementation (e.g. amount of memory available) may
with ‘information.’ constrain what is actually derivable. It follows that physical
An implicitly defined notion of ‘information’ structure does not constrain information content, unless a type
We understand the word ‘information’ insofar as we use it in of user is specified.
a rich, deep, and widely applicable theory (or collection of Information processing in virtual machines
theories) in which many things are said about information, Because possible operations on information are much more
e.g. that it is not conserved (I can give you information complex and far more varied than operations on matter and
without losing any), that instead of having a scalar measure energy, engineers discovered, as evolution had ‘discovered’
of quantity, items of information, may form a partial ordering much earlier, that relatively unfettered information processing
of containment (information I2 is contained in I1 if I2 is requires use of a virtual machine rather than a physical
derivable from I1), and can have a structure (e.g. there are machine. E.g. digital electronic calculators can perform far
replaceable parts of an item of information such that if those more varied tasks than mechanical calculators using cog-
parts are replaced the information changes but not necessarily wheels.
the structure), that two information items can share some It seems to be a basic law that increasing usable information
parts (e.g. ‘Fred hates Mary’ and ‘Mary hates Joe’), that it in a virtual machine by making implications explicit requires
can be transmitted by various means from one location or the physical implementation machine to use energy. Similarly
object to another, that it can vary both discontinuously (e.g. (as suggested by Jackie Chappell), using greater information
adding an adjective or a parenthetical phrase to a sentence, content requires more energy to be used: e.g. in storage,
like this) or continuously (e.g. visually obtained information sorting and processing information. So biological species able
about a moving physical object), that it can be stored in to acquire and process vast amounts of information must be
various forms, that it can influence processes of reasoning near the peak of a food pyramid, and therefore rare.
and decision making, that it can be extracted from other
information, that it can be combined with other information Causal and correlational theories of meaning are false
to form new information, that it can be expressed in different It is often thought that learning to understand S as referring
syntactic forms, that it can be more or less precise, that it can to X, requires an empirical discovery that there is a causal
express a question, an instruction, a putative matter of fact, relation between S and X, such as that occurrences of X always
and in the latter case it can be true or false, known by X, or often cause occurrences of S to come into existence, or
unknown by Y, while Z is uncertain about it, etc. etc. such that the occurrence of X is shown empirically to be a
reliable predictor of the occurrence of S. That this theory
Information is relative to a user, or potential user is false is shown by the fact that you have no reason to
Whereas energy and physical structures simply exist, whether believe that occurrences of the word ‘eruption’ are correlated
used or not, information in a physical or virtual structure with or reliable predictors of eruptions, and moreover you can
S is only information for a type of user. Thus a structure understand the phrase ‘eruption that destroyed the earth 3000
S refers to X or contains information about X for a user years ago’ even though it is impossible for any such correlation
of S, U. The very same physical structure can contain or causal link to exist, since the earth was not destroyed then.
different information for another user U’, or refer to something Further we can use concepts that refer to abstract entities
different for U’, as shown by ambiguous figures, and also whose existence is timeless, such as the number 99, or the
written or spoken languages or notations that some people shortest proof that 2 has no rational square root. So empirical
understand and others do not. This does not make information correlations and causal influences are impossible in those
inherently subjective any more than mountains are subjective cases.
because different people are capable of climbing different
mountains. (Indexicality is a special case, discussed below.) Information content determined partly by context
The information in S can be potentially usable by U even It is sometimes thought that artificial minds would never](https://image.slidesharecdn.com/machinesintheghost-120102113734-phpapp02/85/Machines-in-the-ghost-6-320.jpg)
![be able to grasp context-sensitive information. For example, parts cannot access. When we ask ‘Did X know that P?’ it
an information-bearing structure S can express different is not clear whether the answer must be ‘yes’ in cases where
information, X, X’, X”, for the same user U in different some part of X made use of the information that P. E.g. human
contexts, e.g. because S includes an explicit indexical element posture control mechanisms use changes in optical flow. Does
(e.g. ‘this’, ‘here’, ‘you’, ‘now’, or non-local variables in that mean that when you are walking around you know about
a computer program). Indexicality can make information optical flow even though you don’t know that you know it?
incommunicable in the sense that the precise information Your immune system and your digestive system and various
content of one user cannot be transferred to another. (Frege, metabolic processes use information and take decisions of
for example, showed that one user’s use of the word “I” has many kinds. Does that mean that you have the information, that
a sense that another person is incapable of expressing.) A you know about the information, that you use the information?
corollary is that information acquired by U at one time may Some people might say yes others no, and some may say
not be fully interpretable by U at another time, because the that it depends on whether you know that you are using the
context has changed, e.g. childhood ‘memories’. In [17] it was information.
argued that such indexicality accounts for the ‘ineffability’ of Likewise there are different parts of our brains that evolved
qualia. However this does not usually prevent the intended at different times that use different kinds of information
function of communication from being achieved. The goal of (even information obtained via the same route, e.g. the
communication is not to replicate the sender’s mental state, or retina or ear-drum, or haptic feedback). Some of them are
information content, in the receiver, but to give the receiver evolutionarily old parts, shared with other species (e.g. posture
information that is adequate for some purpose. control mechanisms), some newer and possibly some unique
Many structures in perceptual systems change what to humans, (e.g. human face recognition mechanisms, and
information they represent as the context changes. Even if mechanisms that can learn to read music and other notations).
what is on your retina is unchanged after you rapidly turn A deep feature of at least the human architecture, and
your head 90 degrees in a room, the visual information will perhaps some others also, is that sensors and effectors
be taken to be about a different wall – with the same wallpaper providing interfaces to the environment can be shared between
as the first wall. Many examples can be found in [18]. different subsystems, as described in [20]. E.g. your vision
Sometimes U takes S to express different meanings in mechanisms can be shared between: posture control, visual
different contexts because S includes a component whose servoing of manipulation actions, mechanisms involved in
semantic role is to express a higher order function which reading instructions, and affective mechanisms that appreciate
generates semantic content from the context, e.g. ‘He ran aesthetic qualities of what is seen. Your walking mechanisms
after the smallest pony’. Which pony is the smallest pony can be shared between a subsystem concerned with moving
can change as new ponies arrive or depart. More subtly what to the door and also a subsystem concerned with social
counts as a tall, big, heavy, or thin something or other can vary communication, e.g. flirting by walking suggestively. We can
according to the range of heights, sizes, weights, thicknesses describe such architectures as using ‘multi-window’ perception
of examples in the current environment. and ‘multi-window’ action, whereas current artificial systems
There are many more examples in natural language that lead mostly use only ‘peephole’ perception and ‘peephole’ action,
to incorrect diagnosis of words as vague or ambiguous, when where input and output streams from each sensor or two each
they actually express precise higher order functions, applied effector are go along channels of restricted functionality.
to sometimes implicit arguments, e.g. ‘big’, ‘tall’, ‘efficient’, Sometimes the sharing is concurrent and sometimes
‘heap’, or ‘better’ (discussed in [19]).5 sequential. Conflict resolution mechanisms may be required
Information content shared between users when concurrent sharing is impossible. Some of the
Despite the above, it is sometimes possible for X to mean the mechanisms that detect and resolve conflicts may be
same thing to different users U and U’, and it is also possible inaccessible to self-monitoring (discussed later), so that an
for two users who never use the same information bearers individual may be unaware of important decisions being taken.
(e.g. they talk different languages) to acquire and use the same Much philosophical, psychological, and social theorising
information. This is why relativistic theories of truth are false: misguidedly treats humans as unitary information users,
although I can believe that my house has burned down while including Dennett’s intentional stance and what Newell refers
my neighbour does not, one of us must be mistaken: it cannot to as ‘the Knowledge level’.
be true for me that my house has burned down but not true
for my neighbour. Truth is not ‘for’ anyone. Meaning depends Just the beginning of an analysis of ‘information’
on the user. Truth does not. The analysis of the concept of ‘information’ presented
here amounts to no more than a small fragment of the
V. I NFORMATION - USING SUBSYSTEMS full theory of types of states, events, processes, functions,
An information-user can have parts that are information mechanisms and architectures that are possible in (virtual
users. A part can have and use some information that other and physical) information-processing machines. I doubt
that anyone has produced a clear, complete and definitive
5 This idea is developed in the context of Grice’s theory of communication,
list of facts about information that constitute an implicit
with implications for the evolution of language, here: http://www.cs.
bham.ac.uk/research/projects/cosy/papers/#dp0605 Spa- definition of how we (the current scientific community well-
tial prepositions as higher order functions. educated in mathematics, logic, psychology, neuroscience,](https://image.slidesharecdn.com/machinesintheghost-120102113734-phpapp02/85/Machines-in-the-ghost-7-320.jpg)
![biology, computer science, linguistics, social science, artificial brains, add only more of the same. However, in humans and
intelligence, physics, cosmology, ...) currently understand the many other animals there additional kinds of information and
word ‘information’. information processing.
E.g. there’s a great deal still to be said about the
Brains are needed for more than movement
molecular information processing involved in development of
Most organisms manage without brains. Why not all?
an individual from a fertilised egg or seed, and in the huge
One function is resolving conflicts between different parts
variety of metabolic processes including intrusion detection,
responsible for decisions that could be incompatible, e.g.
damage repair, transport of materials and energy, and control
decisions to move one way to get to food or to move another
by hormones and neurotransmitters. It may be that we shall
way to avoid a sensed predator. This requires coordination,
one day find that far more of the brain’s information processing
possibly based on dynamically changing priorities. A different
is chemical than anyone dreams now is possible.
kind of requirement, discussed later, is doing more complex
A more complete theory would provide a more complete
processing of information, e.g. in order to acquire a better
implicit definition of the concept ‘information’ required
understanding of the environment, or to acquire something
for understanding natural and artificial systems (including
like a terrain map, or to plan extended sequences of actions.
far more sophisticated future artificial systems). A hundred
In organisms with many complex parts performing different
years from now the theory may be very much more deep
functions it may also be necessary to coordinate internal
and complex, especially as information processing machines
changes, for example all the changes involved in reaching
produced by evolution still seem to be orders of magnitude
puberty in humans, or during pregnancy, where many internal
more complex than any that we so far understand.6
changes and external behaviours need to be coordinated.
Lewis Wolpert wrote, in an Observer book review, March
VI. B IOLOGICAL INFORMATION PROCESSING 24, 2002: ‘First the only function of the brain from an
All living things, including plants and single-celled evolutionary point of view is to control movement and so
organisms, process information insofar as they use sensors to interaction with the environment. That is why plants do not
detect states of themselves or the environment and use that have brains.’7 This often quoted, but grossly misleading, claim
information either immediately or after further processing to reflects a widespread current focus on research that models
select from a behavioural repertoire. The behaviour may be brains as sensory-motor control mechanisms, e.g. [21]. Even
externally visible physical behaviour or internal processes. if the original function of brains was to control movement,
While using information an organism normally also uses up much of what human brains do has nothing to do with control
stored energy (usually chemical energy), so that it also needs of movement – for example explaining what is observed,
to use information to acquire more energy. predicting future events, answering a question, and doing
There are huge variations both between the kinds of mathematics or philosophy.
information contents used by different organisms and between So we see that information in organisms may be implicit
different ways in which information is acquired, stored, and transient, implicit and enduring, explicit and capable of
manipulated and used by organisms. The vast majority of multiple uses, used only locally, used in controlling spatially
organisms use only two kinds of information: (a) genetic distributed information processing, and may vary in kind of
information acquired during evolution, used in replication, abstraction, in the ontology used, in the form of representation
physical development and maintenance of the individual, and used and in the kinds of manipulation that are available.
(b) transiently available information used by online control Varieties of explicit information
systems. I call the latter ‘implicit’ information. There may Only special conditions bring about evolution of mechanisms
also be some less transient implicit information produced by that create, store and use explicit, that is enduring, re-usable,
gradually adjusted adaptive mechanisms. information structures that can be used in different ways
Since the vast majority of species are micro-organisms (e.g. forming generalisations, making predictions, building
the vast majority of information-using organisms can use terrain maps, forming motives, making plans, remembering
only implicitly represented information; that is to say they what happened when and where, communicating to other
use only information that is available during the transient individuals, etc.) As explained in [22], this requires new
states of activation produced by information being acquired architectures involving mechanisms that are not so directly
and used, and the information is represented only transiently involved in the sensorimotor relationships and their control.8
in activation states of sensors, motors and intervening There may have been intermediate stages of evolution
mechanisms, and also in parameters or weights modified by in which non-transient information was stored only in the
adaptive feedback mechanisms. environment, e.g. in chemical trails and land-marks used as
The short term transient implicit information in patterns of sources of information about the presence of food or predators,
activation and the longer term implicit information in gradually or about the routes followed by con-specifics (pheromone
changing adaptive mechanisms together suffice for most living trails), or about location relative to a nest or an enduring source
things – most of which do not have brains! In some animals
7 http://observer.guardian.co.uk/science/story/0,
6 For a draft attempt to answer the question ‘What is infor- ,673268,00.html
mation?’ see http://www.cs.bham.ac.uk/research/projects/ 8 See ‘Sensorimotor vs objective contingencies’ http://www.cs.
cogaff/misc/whats-information.html bham.ac.uk/research/projects/cosy/papers/#dp0603](https://image.slidesharecdn.com/machinesintheghost-120102113734-phpapp02/85/Machines-in-the-ghost-8-320.jpg)
![of food. (Some insects can use land-marks so this capability individual B believes is the case, but the functional roles
probably evolved a long time ago.) This is possible only for of those two forms of representation will be different.9 In
animals in an environment with stable structures, unlike some particular, the former will be ‘referentially transparent’ and the
marine environments. latter ‘referentially opaque’, and the conditions for truth of the
Mechanisms that evolved to use external enduring two beliefs are quite different – something young children do
information may have been precursors to mechanisms not learn in the first few years.
using internally stored explicit information. There probably The ability to think and reason about the mental contents
were many different evolutionary transitions, adding extra of another has much in common with the ability to think and
functionality. In previous papers (e.g. [23]) colleagues and reason about one’s own mental contents as far as mechanisms
I have emphasised three main categories of competence and formalisms are concerned, though obviously different
requiring different sorts of architectural components, namely sensor mechanisms are involved, i.e. external and internal.
reactive, deliberative and meta-management capabilities, but However the evolutionary benefits are very different. It is
there are many intermediate cases and different sorts of not clear which evolved first. Probably both types of meta-
combinations of cases that need to be understood – not only semantic competence co-evolved, each helping to enrich the
for understanding how things work when everything functions other.
normally, but also in order to understand the many ways things Incidentally, nothing said here implies that any organism
can go wrong, including both the consequences of physical has full self-knowledge. On the contrary, self-knowledge
malfunctions and also the consequences of dysfunctional will inevitably be associated with a bottleneck with limited
processing of information in virtual machines. capacity – a point that has relevance for psycho-analysis.
Compare [25].
Deliberative mechanisms
A very small subset of organisms (and some machines) have Perhaps a more widely-shared ability to formulate internal
a ‘deliberative’ information-processing capability insofar as questions or goals was an evolutionary precursor to meta-
they can construct a set of information structures of varying semantic competence, since formulating a yes-no question
complexity (e.g. plans, predictions, theories), then compare (‘Will it rain today?’) requires an ability to represent
their merits and de-merits in relation to some goal, and propositions that are capable of being true or false without
produce new information structures describing those tradeoffs, a commitment to their truth value. The same is true of having
then select one of the alternatives and make use of it. That desires and intentions. So the ability to represent ‘X thinks
ability to represent and reason about hypotheticals was one I will eat him’ and to reason about the consequences of X’s
of the first aspects of human processing modelled in AI thinking that (e.g. X will run away) may have arisen as a
subsystems, but it was probably one of the last to evolve, modification of the ability to represent one’s own goals (‘I
and only a very tiny subset of organisms have it in its richest will eat X’) that have not been achieved, plans that have not
form, described in [24]. yet been carried out, tentative predictions that have not yet
It is now fashionable to contrast that early ‘symbolic’ AI been tested, and questions that have not yet been answered.
work with biologically inspired AI work. But that is just silly, Old and new in the same architecture
since all of the work was biologically inspired insofar as it was It is not always remembered that besides having such
an attempt to model processes that occur in human beings. A sophisticated (and possibly unique) capabilities for explicit
human making a plan or proving a theorem is just as much a manipulation of information, humans share many information
biological organism as a human running, jumping or catching processing capabilities with other species that lack the
a ball, even though more species share the capabilities required distinctively human capabilities. For instance many animals
for the latter activities. Biology is much richer than some have excellent vision, excellent motor control, abilities to
researchers seem to realise. cooperate on certain tasks, hunting capabilities, the ability
Meta-semantic competences to avoid predators, nest-building capabilities, as well as
Another relatively rare biological information-processing the information-processing involved in control of bodily
capability involves the ability to refer to, reason about, or functions. The notion that somehow human cognition, or
care about things that themselves contain, or use information. human conscious processes can be studied and replicated
For instance when you discover what someone thinks, and without taking any account of how these more general animal
when you worry about someone’s motives you are using mechanisms work and how they interact with the distinctively
such meta-semantic competence. Humans can also apply this human mechanisms, may lead both to a failure to understand
meta-semantic competence to themselves (though probably humans and other animals and also to designs for robots that
not at birth: the architecture needs time to develop). Having don’t perform as required.
this ability requires a more complex architecture than merely Some of the ideas developed with colleagues in Birmingham
being able to refer to things without semantic content (e.g. about the different components in such a multi-functional
physical objects and processes). That’s partly because having architecture are reported in the paper by Brigitte Lorenz and
meta-semantic content involves representing things that are Etienne Barnard presented at this conference and will not be
treated as true in only a hypothetical, encapsulated, way. The 9 J. Barnden’s ATT-META project has developed a way of making that
very same form of representation may be used for what an distinction which is also related to the ability to think metaphorically. http:
individual A believes is the case and for what A thinks another //www.cs.bham.ac.uk/∼jab/ATT-Meta/](https://image.slidesharecdn.com/machinesintheghost-120102113734-phpapp02/85/Machines-in-the-ghost-9-320.jpg)
![developed here. From this standpoint it is almost always a to run, if an interpreter is available. Moreover if it were
mistake to ask questions like ‘How do humans do X?’ Instead necessary, then substantive ontology extension during learning
we should ask ‘How do different subsystems in humans do and development would be impossible.
X?’ (e.g. X = control actions, interpret visual input, learn, store Nobody is yet in a position to say what the prelinguistic
information, react to interruptions, generate motives, resolve languages do and do not share with human communicative
conflicts, etc.) And we should expect different answers not languages. In particular, it may be the case that the
only for different subsystems, as in Trehub’s [26], but also main qualitative competences required for human language
for humans at different stages of development, in different use already exist in these pre-linguistic competences in
cultures, with and without brain damage, etc. young children and other animals, and that the subsequent
evolutionary developments related to human language
were mainly concerned (a) with developing means of
VII. P RE - LINGUISTIC COMPETENCES
generating adequately articulated external behaviours to
It is often assumed that there is a massive discontinuity communicate their structures, (b) improving perception of
between human linguistic competence and other competences such behaviours, and (c) extending internal mechanisms (e.g.
(e.g. as argued by Chomsky in [27]), though there are many short term memories) that are equally useful for sophisticated
who hotly dispute this (e.g. Jablonka and Lamb [28]). But internal information processing (such as planning) and for
people on both sides of the dispute make assumptions about communication with others.
human language that may need to be challenged if we are If so, information (unconsciously) acquired, used and stored
to understand what human minds are, and if we wish to for later use in early childhood may have much richer
produce working human-like artificial minds. In particular it structures and deeper semantic content than has previously
is often assumed that the essential function of language is been thought possible. Whether it includes the kind of content
communication. But, as argued in [29], and more recently required to support psychoanalytic theories remains open.
in [30] many of the features that make such communication
Non-auditory communication
possible (e.g. the ability to use varieties of information with
It is easier to achieve a large collection of perceptually
rich and varied structures and with compositional semantics,
discriminable signals by using independently movable fingers,
and non-communicative abilities to check whether some state
hands, mouth, eyes, head than to do it all by modulating
of affairs matches a description, to notice a gap in information,
an acoustic signal – especially as that would interfere with
expressed in a question) are also requirements for forms
breathing. So, if some animals already had hands for which
of information that are involved in perception, planning,
they were learning to produce (and perceive, during controlled
expressing questions, formulating goals, predicting, explaining
execution) large numbers of distinct manipulative competences
and reasoning.
concerned with obtaining and eating food, grooming, climbing,
From that viewpoint, communication between individuals
making nests, fighting, threatening, etc., then perhaps the
in a public medium was a secondary function of language
first steps to communicative competence used structured
which evolved only after a more basic kind of competence
movements, as in that involved producing and perceiving
evolved – the ability to use an internal, non-communicative,
structured sign language, rather than vocal language.
language in mental states and processes such as perceiving,
There is much suggestive evidence, including the fact that
thinking, supposing, intending, desiring, planning, predicting,
humans find it almost impossible not to move hands, eyes,
remembering, generalising, wanting information, etc. and in
facial muscles, etc. when talking, even when talking on the
executing intentions or plans.
phone to someone out of sight. More compelling is the ease
It is clear that human children who cannot yet talk, and
with which deaf babies are able to learn a sign language, and
many animals that do not use an external language can
the reported fact that some children with Down syndrome seem
perceive, learn, think, anticipate, have goals, threaten, be
to learn to communicate more easily if sign language is used.
puzzled, play games, carry out intentions and learn and use
Most compelling is the case of the Nicaraguan deaf children
facts about causation. I know of no model of how any of that
who invented their own highly sophisticated sign language,
can be done without rich information processing capabilities
leaving their teacher far behind. See [32].10
of kinds that require the use of internal languages with
An implication of this is that the human ability to develop
compositional semantics. But I know of no detailed model
linguistic competence does not depend on learning a language
of what those ‘prelinguistic’ languages are, how they evolved,
that is already in use by others. Language learning then
how they develop, how they work, etc.
appears to be a process of collaborative, creative problem-
Fodor, in [31], postulated a ‘language of thought’ (LOT) solving, which may be constrained by the prior existence of a
which was supposed to be innate, available from birth and social language, but need not be.
capable of expressing everything that ever needs to be said In [33] Arbib proposed that action recognition and imitation
by any human being, but he left most questions about how was a precursor to the evolution of language, but the arguments
it worked unanswered and was not concerned with non-
human animals. Moreover he supposed that external languages 10 http://www.indiana.edu/∼langacq/E105/Nicaragua.
are translated into the LOT, whereas there is no reason html ‘A Linguistic Big Bang’, by Lawrence Osborne The New York Times
October 24, 1999. A five minute video including examples of the invented
to believe such translation is necessary, just as compiling language is available here http://www.pbs.org/wgbh/evolution/
to machine code is not necessary for computer programs library/07/2/l 072 04.html](https://image.slidesharecdn.com/machinesintheghost-120102113734-phpapp02/85/Machines-in-the-ghost-10-320.jpg)
![are somewhat different, and do not share our hypothesis in context demands, in collaborative problem solving rather
[29] that the existence of a rich non-communicative language than in a process of information transfer. (This is related to
preceded the evolution of communicative language, though Grice’s maxims of communication, and to current concerns
the commentary by Bridgeman [34] makes a similar point, with ‘situatedness’ in language understanding, perception
emphasising pre-linguistic planning capabilities. and action. http://www.cs.bham.ac.uk/research/
Once something like this form of structured communication projects/cosy/papers/#dp0605)
had developed, enabling information that was previously If the contents of linguistic communication between
only represented in internal languages to be shared between individuals make such use of the context then it may be
individuals via a public language, the enormous advantages equally true for thought processes within an individual. In that
of having such a communicative competence, both in solving case, much of the information content of your mind is outside
problems that required collaboration and in accelerating you. This may be one of the ideas people who emphasise
cultural transmission, might have led to an unusually rapid the importance of ‘embodiment’ are getting at. However,
process of selection for changes that enhanced the competence everything said here could apply to minds in individuals
in various ways, e.g. developing brain regions specialised at with simulated bodies located in simulated environments. The
storage of ever larger vocabularies and more complex rules for important thing is what sorts of structures, processes and
construction and interpretation of action sequences. causal interactions occur within the individual’s information
Another development might have been evolution of brain processing architecture and how they are related to its
mechanisms to support construction and comparison of environment. Whether that environment is physical or another
more complex symbolic structures, e.g. more deeply nested virtual machine environment makes little or no difference to
structures, and structures with more sub-structures. the kind of mind discussed here.
The nature of linguistic communication From sign language to sound language
If linguistic communication evolved from collaborative non- Use of sign language has many problems. As has often been
linguistic activities, controlled by sophisticated internal noted, it can work only when the sender’s hands are free,
languages, this must change our view of human language. and when the receiver can see the sender, thereby ruling out
It is often thought that linguistic communication involves a communication in many situations where visual attention must
process whereby an information structure in one individual be used for another task, or where there are obstructions
gets encoded in some behaviour which is perceived and to vision, or no light is available. So perhaps our ancestors
decoded by another individual who then constructs the same started replacing hand signals with sounds where possible.
information internally. However, in collaborative, creative, This could then lead in the usual way to a succession of
problem solving the shared physical and task contexts provide evolutionary changes in which the vocal mechanisms, auditory
much of the information, and need not be communicated. All mechanisms, and the neural control systems evolved along
that is needed is whatever suffices to produce desired results, with physiological changes to the vocal tract, etc. But the
without copying an information structure from one brain to ability to learn sign languages remains intact, even if not used
another. by most people.
If a mother hands a child a piece of fruit, the child may If all this is correct then it may have deep implications
thereby be triggered to form the goal of eating it, without the for clinical developmental psychology as well as for
mother having had the goal of eating it. Moreover, the mother understanding precisely what the effects of various kinds of
does not need to anticipate the precise movements of bringing brain damage are.
the food to the mouth nor the precise chewing and swallowing
movements produced by the child. Those details can be left
to the child. More generally, context allows communication to VIII. A RCHITECTURES THAT GROW THEMSELVES
be schematic rather than concrete and detailed.
This is obvious when A asks ‘Where are my keys?’ and Much research in AI and robotics is concerned with what
B replies ‘Look behind you?’. B may not know exactly sort of architecture an intelligent system needs, and various
where the keys are except that they are on the table behind rival architectures are proposed. If the architecture needs to
A. A turns round and sees the keys, getting the precise grow itself, some of this research effort may have been wasted.
information he wanted, as a result of B’s communication, even It was argued in [3] that the architecture should not be regarded
though B did not have that information. Both have achieved as genetically determined, but as grown using both innate
their communicative goals, but not by transmitting a specific and acquired meta-competences influenced by the physical
information structure from B to A. and cultural environment, as depicted in Figure 1. So there
So, many kinds of linguistic and non-linguistic commu- may be considerable differences in how adult minds work,
nication involve communication of a partial or schematic as a result of a different developmental trajectories caused
structure, leaving gaps to be filled by the receiver using by different environments. This might affect kinds of self-
information that may or may not be available to the monitoring, kinds of self control, kinds of learning ability,
sender. In that case, many linguistic expressions may be and so on. Another implication is that attempts to quantify the
best thought of as having a higher-order semantics, to influence of genes and environment in terms of percentages is
be applied to more specific non-linguistic information as completely pointless.](https://image.slidesharecdn.com/machinesintheghost-120102113734-phpapp02/85/Machines-in-the-ghost-11-320.jpg)
![presentations are in the Birmingham CogAff and CoSy project
web sites. Jackie Chappell and Gerhard Pratl made useful
comments on an early draft.
Fig. 1. This indicates some of the many and varied relationships between
the genome and behaviours, produced at different stages of development after
layers of competences and meta-competences have been built up. (From [3])
IX. C ONCLUSION : MACHINES IN GHOSTS
Although I was asked to report on developments in AI that
might be relevant to psycho-analysis I have instead focused
on some very general features of requirements for AI models
rather than specific AI models, because I think that despite all
the advances achieved in AI, the models are still very primitive
compared with human minds and have a very long way to
go before they can be directly relevant to working therapists
dealing with real humans.
Nevertheless, I hope it is now clear why a human-like
machine, a human-like intelligent ghost, and indeed a human-
like human must contain very specific kinds of information
processing virtual machines. So our ability to understand
human minds and ways they can go wrong, including acquiring
false, beliefs, inappropriate motives, inadequate strategies,
tendencies to be over-emotional, and worse, must be informed
by deep knowledge about information-processing systems of
the appropriate kinds.
The complexity of the machine, and the fact that self-
monitoring mechanisms within it provide only a very limited
subset of information about what is going on cause many
individuals to start wondering what is really going on in them.
Ignorance about information-processing mechanisms can make
things seem so mysterious that some people invoke a special
kind of stuff that is quite unlike physical matter. From there it
is a short step to immortality, souls, etc., and the conceptual
confusions discussed by Ryle in [35]. But if we don’t go that
far we may still come up with deep new ways of thinking about
phenomena that were previously thought resistant to computer
modelling. This could, among other things, lead to a revolution
in psychoanalysis.
ACKNOWLEDGEMENT
Thanks to Dean Petters, Jackie Chappell, Matthias Scheutz,
Margaret Boden, members of the EU-Funded Cosy Project,
and other collaborators and discussants over the years. Some
of the ideas here were inspired by John McCarthy and Marvin
Minsky, as well as Immanuel Kant, [6]. Related papers and](https://image.slidesharecdn.com/machinesintheghost-120102113734-phpapp02/85/Machines-in-the-ghost-12-320.jpg)
![R EFERENCES [27] N. Chomsky, Aspects of the theory of syntax. Cambridge, Mass: MIT
Press,, 1965.
[1] D. C. Dennett, Brainstorms: Philosophical Essays on Mind and [28] E. Jablonka and M. J. Lamb, Evolution in Four Dimensions: Genetic,
Psychology. Cambridge, MA: MIT Press, 1978. Epigenetic, Behavioral, and Symbolic Variation in the History of Life.
[2] A. Sloman, The Computer Revolution in Philosophy. Has- Cambridge MA: MIT Press, 2005.
socks, Sussex: Harvester Press (and Humanities Press), 1978, [29] A. Sloman, “The primacy of non-communicative language,” in The
http://www.cs.bham.ac.uk/research/cogaff/crp. analysis of Meaning: Informatics 5 Proceedings ASLIB/BCS Conference,
[3] J. Chappell and A. Sloman, “Natural and artificial Oxford, March 1979, M. MacCafferty and K. Gray, Eds. London: Aslib,
meta-configured altricial information-processing systems,” 1979, pp. 1–15, http://www.cs.bham.ac.uk/research/projects/cogaff/81-
International Journal of Unconventional Computing, 2007, 95.html.
http://www.cs.bham.ac.uk/research/projects/cosy/papers/#tr0609. [30] A. Sloman and J. Chappell, “Computational Cognitive Epigenetics
[4] A. Sloman and J. Chappell, “The Altricial-Precocial Spectrum for (Commentary on [28]),” Behavioral and Brain Sciences, 2007.
Robots,” in Proceedings IJCAI’05. Edinburgh: IJCAI, 2005, pp. 1187– [31] J. Fodor, The Language of Thought. Cambridge: Harvard University
1192, http://www.cs.bham.ac.uk/research/cogaff/05.html#200502. Press, 1975.
[5] S. Harnad, “The Symbol Grounding Problem,” Physica D, vol. 42, pp. [32] L. Osborne, “‘A Linguistic Big Bang’,” The New York Times, 24 October
335–346, 1990. 1999, http://www.indiana.edu/ langacq/E105/Nicaragua.html. A five
[6] I. Kant, Critique of Pure Reason. London: Macmillan, 1781, translated minute video including examples of the invented language is available
(1929) by Norman Kemp Smith. at http://www.pbs.org/wgbh/evolution/library/07/2/l 072 04.html.
[7] A. Sloman, “The irrelevance of Turing machines to AI,” in Computa- [33] M. A. Arbib, “From monkey-like action recognition to human language:
tionalism: New Directions, M. Scheutz, Ed. Cambridge, MA: MIT An evolutionary framework for neurolinguistics,” Behavioral and Brain
Press, 2002, pp. 87–127, http://www.cs.bham.ac.uk/research/cogaff/00- Sciences, vol. 28, no. 2, pp. 105–124, 2005.
02.html#77. [34] B. Bridgeman, “Action planning supplements mirror systems in language
[8] A. Damasio, Descartes’ Error, Emotion Reason and the Human Brain. evolution,” Behavioral and Brain Sciences, vol. 28, no. 2, pp. 129–130.,
New York: Grosset/Putnam Books, 1994. 2005.
[9] R. Picard, Affective Computing. Cambridge, Mass, London, England: [35] G. Ryle, The Concept of Mind. London: Hutchinson, 1949.
MIT Press, 1997.
[10] A. Sloman and M. Croucher, “Why robots will have emotions,” in Proc
7th Int. Joint Conference on AI. Vancouver: IJCAI, 1981, pp. 197–202.
[11] A. Sloman, “Towards a grammar of emotions,” New
Universities Quarterly, vol. 36, no. 3, pp. 230–238, 1982,
http://www.cs.bham.ac.uk/research/cogaff/96-99.html#47.
[12] A. Sloman, “Real time multiple motive-expert systems,” in Proceedings
Expert Systems 85, M. Merry, Ed. Cambridge University Press, 1985,
pp. 213–224.
[13] A. Sloman, “Damasio, Descartes, alarms and meta-management,” in
Proceedings International Conference on Systems, Man, and Cybernetics
(SMC98), San Diego. IEEE, 1998, pp. 2652–7.
[14] A. Sloman, R. Chrisley, and M. Scheutz, “The architectural basis
of affective states and processes,” in Who Needs Emotions?:
The Brain Meets the Robot, M. Arbib and J.-M. Fellous, Eds.
Oxford, New York: Oxford University Press, 2005, pp. 203–244,
http://www.cs.bham.ac.uk/research/cogaff/03.html#200305.
[15] A. Sloman, “Do machines, natural or artificial,
really need emotions?” 2004, Invited talk:
http://www.cs.bham.ac.uk/research/projects/cogaff/talks/#cafe04.
[16] A. Sloman, “‘Ontology extension’ in evolution and
in development, in animals and machines,” University Aaron Sloman Aaron Sloman studied for a BSc in mathematics and physics
of Birmingham, UK, 2006, PDF presentation: at Cape Town in 1956, then, after flirting for a while with mathematical
http://www.cs.bham.ac.uk/research/projects/cosy/papers/#pr0604. logic was seduced by philosophy, and completed a D.Phil at Oxford in 1962,
[17] A. Sloman and R. Chrisley, “Virtual machines and consciousness,” defending Kant’s philosophy of mathematics. He later learnt that AI provided
Journal of Consciousness Studies, vol. 10, no. 4-5, pp. 113–172, 2003. the best tools for doing philosophy and since about 1970 has been working on
[18] A. Berthoz, The Brain’s sense of movement, ser. Perspectives in interdisciplinary problems concerned with understanding the space of possible
Cognitive Science. London, UK: Harvard University Press, 2000. designs for behaving systems including humans, other animals and possible
[19] A. Sloman, “How to derive “better” from “is”,” future robots. In recent years this has particularly involved collaborating with
American Phil. Quarterly, vol. 6, pp. 43–52, Jan 1969, a biologist studying cognition in birds and roboticists designing robots able to
http://www.cs.bham.ac.uk/research/cogaff/sloman.better.html. perceive and manipulate 3-D objects in the environment. The various strands
[20] A. Sloman, “The mind as a control system,” in Philosophy and the of his work are summarised in http://www.cs.bham.ac.uk/∼axs/
Cognitive Sciences, C. Hookway and D. Peterson, Eds. Cambridge, my-doings.html He is a Fellow of AAAI, ECCAI and SSAISB, and was
UK: Cambridge University Press, 1993, pp. 69–110. given an honorary DSc by Sussex University in 2006.
[21] M. Lungarella and O. Sporns, “Mapping information flow in
sensorimotor networks,” PLoS Computational Biolology, vol. 2, no.
10:e144, 2006, dOI: 10.1371/journal.pcbi.0020144.
[22] A. Sloman, “Sensorimotor vs objective contingencies,” University of
Bimringham, School of Computer Science, Tech. Rep. COSY-DP-0603,
2006, http://www.cs.bham.ac.uk/research/projects/cosy/papers/#dp0603.
[23] A. Sloman, “Architecture-based conceptions of mind,” in In the Scope
of Logic, Methodology, and Philosophy of Science (Vol II). Dordrecht:
Kluwer, 2002, pp. 403–427, (Synthese Library Vol. 316).
[24] A. Sloman, “Requirements for a Fully Deliberative
Architecture,” University of Bimringham, School of
Computer Science, Tech. Rep. COSY-DP-0604, 2006,
http://www.cs.bham.ac.uk/research/projects/cosy/papers/#dp0604.
[25] M. Minsky, “Matter Mind and Models,” in Semantic Information
Processing, M. Minsky, Ed. Cambridge, Mass,: MIT Press,, 1968.
[26] A. Trehub, The Cognitive Brain. Cambridge, MA: MIT Press, 1991,
Online http://www.people.umass.edu/trehub/.](https://image.slidesharecdn.com/machinesintheghost-120102113734-phpapp02/85/Machines-in-the-ghost-13-320.jpg)
Machines in the ghost
- 1. Machines in the Ghost Invited Position Paper for ENF 2007 Emulating the Mind 1st international Engineering and Neuro-Psychoanalysis Forum Vienna July 2007 http://www.indin2007.org/enf/ DRAFT: August 30, 2007(MAY CHANGE) Aaron Sloman http://www.cs.bham.ac.uk/∼axs/ School of Computer Science, University of Birmingham, UK Index Terms— architecture, artificial-intelligence, autonomy, is the common assumption that we already know what needs behaviour, design-based, emotion, evolution, ghost in machine, to be explained and modelled, leaving only the problem information-processing, language, machine, mind, robot, philos- of finding good theories and designs for working systems. ophy, psychotherapy, virtual machine. Alas, what needs to be explained is itself a topic still Abstract— This paper summarises ideas I have been working on over the last 35 years or so, about relations between the requiring much research, as acknowledged by the Research study of natural minds and the design of artificial minds, and Roadmap project in the euCognition network (http://www. the requirements for both sorts of minds. The key idea is eucognition.org).1 . that natural minds are information-processing virtual machines A less obvious, but even more serious impediment to produced by evolution. What sort of information-processing progress is the common assumption that our ordinary language machine a human mind is requires much detailed investigation of the many kinds of things minds can do. At present, it is is sufficient for describing everything that needs to be not clear whether producing artificial minds with similar powers explained, leading to over-reliance on common-sense concepts. will require new kinds of computing machinery or merely much When scientists and engineers discuss what needs to be faster and bigger computers than we have now. Some things once explained, or modelled, and when they report experimental thought hard to implement in artificial minds, such as affective observations, or propose explanatory theories, they often use states and processes, including emotions, can be construed as aspects of the control mechanisms of minds. This view of mind concepts (such as ‘conscious’, ’unconscious’, ‘experience’, is largely compatible in principle with psychoanalytic theory, ‘emotion’, ‘learn’, ‘motive’, ‘memory’) that evolved not for though some details are very different. The therapeutic aspect the purposes of science, but for use in informal discourse of psychoanalysis is analogous to run-time debugging of a among people engaged in every day social interaction, like virtual machine. In order to do psychotherapy well we need to this: understand the architecture of the machine well enough to know what sorts of bugs can develop and which ones can be removed, • What does the infant/child/adult/chimp/crow (etc) or have their impact reduced, and how. Otherwise treatment will perceive/understand/learn/intend (etc)? be a hit-and-miss affair. • What is he/she/it conscious of? • What does he/she/it experience/enjoy/desire? • What is he/she/it attending to? I. T HE DESIGN - BASED APPROACH TO STUDYING MINDS • What sort of emotion is he/she/it having now? M Y PRIMARY goal is to understand natural minds of all kinds, not to make smart machines. I am more a philosopher than an engineer – though I have first- These everyday usages may be fine for everyday chats, gossip, consulting rooms, or even law courts and medical reports, but it does not follow that the concepts used (e.g. hand experience of software engineering. Deep scientific and ‘conscious’, ‘experience’, ‘desire’, ‘attend’, ‘emotion’) are any philosophical understanding of natural minds requires us to more adequate as concepts to be used in scientific theories than describe minds with sufficient precision to enable our theories the everyday concepts of ‘mud’, ‘cloudy’, ‘hot’, ‘vegetable’, to be the basis for designs for working artificial minds like ‘reddish’ or ‘smelly’ are useful in theories about physics or ours. Such designs can be compared with psychoanalytic chemistry. and other theories about how minds work. If a theory about how natural minds work is to be taken seriously, it should From vernacular to deep concepts be capable of providing the basis for the design of artifical The history of the physical sciences shows that as we working minds. understand more about the architecture of matter and the Part of the problem is describing what needs to be variety of states, processes, and causal interactions that lie explained, and deciding which concepts to use in formulating hidden from ordinary observation, the more we have to explanatory theories. A very serious impediment to progress construct new systems of concepts and theories that make use This work was partly supported by the EU CoSy project. 1 Available online: ‘What’s a Research Roadmap For? Why do we need one? A. Sloman is at the School of Computer Science, University of Birming- How can we produce one?’ http://www.cs.bham.ac.uk/research/ ham, UK, http://www.cs.bham.ac.uk/˜axs/ projects/cosy/papers/#pr0701
- 2. of them (and thereby help to define the concepts), in order between theoretical concepts and observable phenomena is to obtain deep explanations of what we can observe and new referred to as ‘symbol attachment’ in [3] and ‘symbol more reliable and more precise predictions, regarding a wider tethering’ in [4]. Once this possibility is understood, the need range of phenomena. In particular, in physics, chemistry and for so-called ‘symbol grounding’ (deriving all concepts from biology, new theories have taught us that however useful our experience) presented in [5], evaporates. (The impossibility of ordinary concepts are for ordinary everyday social interactions, ‘concept empiricism’ was demonstrated by Kant [6] over 200 they are often grossly inadequate: (a) for distinguishing years ago.) all the phenomena that need to be distinguished, (b) for The ontology used by an individual or a community (i.e. formulating precise and reliable predictions, and (c) for the set of concepts used to describe things in the world) can describing conditions (often unobservable conditions) that are be extended in two ways, either by definitional abbreviation the basis for reliable predictions. or substantively. A definitional abbreviation merely introduces So it may be unwise to go on using the old concepts a new symbol as a short-hand for what could be expressed of folk psychology when describing laboratory experiments previously, whereas substantive ontology extensions introduce or field observations, when formulating descriptions of what new concepts that cannot be defined in terms of pre-existing needs to be explained, and when formulating supposedly concepts. Such concepts are implicitly defined mainly by their explanatory theories. But choosing alternatives to ordinary role in explanatory theories, as explained above. So substantive language needs great care. Recently attempts have been made ontology extension always requires theory construction, as has to give these concepts scientific status by using modern happened many times in the history of science and culture. technology to identify precise brain mechanisms, brain states, brain processes that correlate with the states and processes described in ordinary language. Compare trying to identify the II. V IRTUAL AND PHYSICAL MACHINES precise location in a particular country where some national We can do for the study of mind and brain what was characteristic or process such as religious bigotry, scientific previously done for the study of physical matter and biological ignorance, or economic inflation is located. processes of evolution and development, if we understand that The ‘design-based’ approach minds and brains are not just matter-manipulating, or energy- Is there any alternative to going on using pre-scientific contexts manipulating machines, but information-processing machines. in our descriptions and theories? Yes, but it is not a simple Minds are information processing virtual machines while alternative. We need to adopt what Dennett [1] calls ‘the brains are physical machines, in which those virtual machines design stance’, which involves constructing theories about how are implemented or some would say realised. minds and brains actually work, which goes far beyond what In computing systems we also have virtual machines, we either experience of their working in ourselves or observe such as running operating systems, firewalls, email systems, in others. (My own early attempts at doing this 30 years ago spelling correctors, conflict resolution mechanisms, file are presented in [2].) Moreover, our theories must account for optimisation mechanisms, all of which run on, i.e. are the existence of many kinds of minds, with different designs. implemented in, the underlying physical hardware. There are In particular we can’t just start defining precise new no simple mappings between the components of the virtual explanatory concepts in terms of precise measurements and and the physical machines. There are often several layers observations. Thinking that definitions come before theories is of implementation between high level virtual machines and a common mistake. In the more advanced sciences, concepts physical machines. This provides enormous flexibility for re- are used (e.g. ‘electron’, ‘charge’, ‘atomic weight’, ‘valency’, use of hardware, both in different systems using the same ‘oxidation’, ‘gene’, etc.) that cannot be defined except by hardware, and in one system performing different functions their role in the theories that employ them. The structural at different times. It seems that evolution discovered the relations within the theory partially define the concepts. That power of virtual machines before we did and produced brains is because a theory is a formal system, and, as is familiar from implementing hierarchies of mental (and indirectly social) mathematics, and made more precise in the work of Tarski, virtual machines. the structure of a formal system determines which things The various components of a complex virtual machine are possible models of that system. Usually there are many (such as an operating system distributed over a network possible models, but the set of possible models can be reduced of processors) need not run in synchrony. As a result, by enriching the theory, thereby adding more constraints to be timing relations between processes can change from time satisfied by any model. This may still leave different models, to time. For this and other reasons, Turing machines do of which only one, or a subset is intended. Such residual not provide good models for minds, as explained in [7]. ambiguities are reduced (but never completely removed) by Moreover if sophisticated memory management systems are links between the theory and methods of observation and used the mappings between components of virtual machines experiment, and practical applications associated with the and physical parts of the system can also keep changing. theory. However those links do not define the concepts, If such complex changing relationships are useful in systems because old methods of observation and old experimental we have designed and implemented we need to keep an open procedures can be replaced by new ones, while the old mind as to whether evolution, which had several billion years concepts endure. This loose, theory-mediated, connection head start on us, also discovered the power and usefulness of
- 3. such flexibility. If so, some apparently important searches for can be causes’ by claiming that the objects, events and mind-brain correlations may turn out to be a waste of time. processes in virtual machines actually are physical objects, events and processes viewed in an abstract way. This is a Can virtual machines do things? ‘Mind-brain identity theory’, or a ‘virtual-physical machine It is sometimes thought that the virtual machines in a computer identity theory’, sometimes presented as a ‘dual-aspect’ theory. do nothing: they are just figments of the imagination of A detailed rebuttal of this thesis requires showing (a) that it software engineers and computer scientists. On this view, only is based on a bad theory of causation, (b) that the claim of the physicists and electronic engineers really know what exists identity being used here is either vacuous or false because and interacts in the machine. That derives from a widely held the actual relations between contents of virtual machines and theory of causality, which assumes either that only physical contents of physical machines are quite unlike other cases events can really be causes that produce effects, or, more where we talk of identity (e.g. because the virtual to physical subtly that if events are caused physically then they cannot mapping constantly changes even within the same machine) be caused by events in virtual machines. Some researchers (c) that the concepts used in describing virtual machine have inferred that human mental events and processes, such phenomena (e.g. ‘software bug’, ‘failure to notice’, ‘preferring as weighing up alternatives, and taking decisions cannot have X to Y’, etc.) are too different from those of the physical any consequences: they are mere epiphenomena. There is no sciences to be capable of referring to the same things. More space here for a full rebuttal of this view, but the key idea is obviously, those virtual machine concepts are not definable that causation is not like some kind of fluid or physical force using only concepts of the physical sciences. that flows from one thing to another. Rather, the concept of X This is a very compressed defence of the claim that virtual causing Y is a very subtle and complex concept that needs to machine events can have causal powers. There are more be analysed in terms of whether Y would or would not have detailed discussions on the Birmingham Cogaff web site.2 happened in various conditions if X did or did not occur, or if X had or had not occurred, or if X does or does not occur in the Psychotherapy as virtual machine debugging future. The truth or falsity of those ‘counterfactual conditional’ This design-based approach using the concept of a ‘virtual statements depends in complex ways on the truth or falsity of machine’ can, in principle, justify techniques that deal with a various laws of nature, which we attempt to express in our subset3 of human mental problems by manipulating virtual explanatory theories. machines instead of manipulating brains using chemicals, However, in addition to their role in true and false electric shocks, etc. If a virtual machine is suitably designed statements counterfactual conditionals may also play a role then it is possible to identify and in some cases repair, certain in instructions, intentions, or motives. Thus ‘Had he not ‘bugs’, or ‘dysfunctional’ processes, by interacting with the cooperated I would have gone ahead anyway’ may be not so running system. Such debugging is common in teaching. much a retrospective prediction as an expression of resolve. Doing that well requires deep, explanatory theories of Plans, intentions and strategies involve causation as much as how normal mental virtual machines work. Some debugging predictions and explanations do. These too can play a causal techniques for minds have evolved through various kinds of role in virtual machines. social experimentation without such deep theories, and many of them work well. For instance, when a child gets the wrong Real causation in virtual machines results for long division we don’t call in a brain surgeon, Suppose one of your files disappears. A software engineer may but check whether he has perhaps learnt the wrong rule or conclude (after thorough investigation) that one of your actions misunderstood one of the mathematical concepts involved. In activated a bug in a running program, which led another part of such cases the lack of competence can be seen as a bug that the program to remove that file. These are events and processes can be fixed by talking, drawing diagrams, and giving simpler in a virtual machine. This diagnosis could lead the programmer examples. It might be thought that this is unlike fixing bugs to make a change in the operating system or file management on computers because in the latter case the process has to be system that alters its future behaviour. Sometimes this can be stopped, the new program compiled and the program restarted done by altering software rules without restarting the machine whereas humans learn and change without having to cease – rather like telling a person how to do better. functioning. But some programs are run using interpreters Typically, the causal connections discovered and altered by or incremental compilers, which allow changes to be made software engineers do not require any action by electronic to the running program without stopping and restarting the engineers or physicists to alter the physical machine, and most process with a recompiled program. Several AI programming of the people with expert knowledge about the hardware would languages have that kind of flexibility. More sophisticated not even understand how the bug had occurred and how it was software development tools can also plant mechanisms for fixed. interrogating and modifying running systems. Moreover, virtual machine components can constantly Of course, teaching someone how to improve his long change their mapping onto hardware, so the change made 2 See this presentation on virtual machines by the software engineer need not have any specific physical http://www.cs.bham.ac.uk/research/projects/cogaff/ location when the software is running. talks/#inf and this discussion of free-will and causation http://www.cs.bham.ac.uk/research/projects/cogaff/ The mind-brain identity defence misc/four-kinds-freewill.html Some philosophers defend the thesis that ‘only physical events 3 Don’t expect it to remove brain-tumours, for instance!
- 4. division, and giving counselling to enable a patient to values. Isn’t that what happens to humans? This point was understand how he unintentionally causes family rows to made long ago in section 10.13 of [2]. So motivation in escalate, differ in detail from the process of fixing a typical artificial virtual machines, including self-generated motivation, bug in a computer program. For example the teaching and is not a problem in principle. therapy depend on kinds of self understanding that few The myth that intelligence requires emotions computer programs have at present. But in future it will In recent years, especially following publication of Damasio’s not be uncommon to find virtual machines, that, instead [8] and Picard’s [9], much has been made of the alleged need of being forcibly altered by a user editing rules, instead to ensure that intelligent systems have emotions. I have argued change themselves as a result of being given advice about elsewhere that the arguments are fallacious for example in [10] how to behave. That’s not even science fiction: it is not [11] [12] [13] [14] and [15]. Some of these claims are merely hard to achieve. This comparison should undermine three poorly expressed versions of Hume’s unsurprising observation assumptions: (a) the belief that the only way to understand that without motives an intelligent system will have no reason and fix problems with minds is to work in a totally bottom to do anything: this is just a confusion between emotions and up way, namely understanding and modifying brains, (b) the motives. belief that nothing we know about computers is relevant to There is also a more subtle error. 26 years ago, [10] understanding and repairing minds, and (c) the belief that argued that intelligent machines need mechanisms of kinds machines can only be programmed to do things, not advised, that perform important functions and which in addition can inspired, or instructed or cajoled into doing them. sometimes generate emotional states and processes as side Motivation in virtual machines effects of their operation, if they lack sufficient processing So, states, events, and processes occurring in virtual machines power to work out what needs to be done. This does not imply can have causal influences which alter both other things in that they need emotions,4 . just as the fact that some operating the virtual machine and also the physical behaviour of the systems need mechanisms that are sometimes capable of system (e.g. physical events in memory, hard drives, internal generating ‘thrashing’ behaviour does not imply that operating interfaces, what is displayed on a screen, the sounds coming systems need thrashing behaviour. Desirable mechanisms can from speakers, or various attached motors, etc.). However, we sometimes have undesirable effects. There can be particular still need to understand the variety of kinds of causation in sorts of situations where an ‘alarm system’ detects a putative virtual machines. need to override, modulate, freeze, or abort some other In very simple computer models nothing happens until a process, and where because of shortage of information or user gives a command, which can then trigger a cascade shortage of processing power, a rough and ready rule operates. of processes. In more sophisticated cases the machine is It would be better if the situation could be fully evaluated designed so that it initiates various activities from time and reasoned about (without any emotions), but if there is to time, e.g. checking for email, checking whether disks inadequate capacity to do that in the time available then it need defragmenting, checking whether current scheduling may be better to take the risk of false alarms, especially if parameters need to be revised in order to improve processing the alarm system has been well trained, either by evolution or performance. It can also be designed so that events initiated individual learning, and does not often get things wrong. from outside trigger new internal processes. E.g. a user The need for therapy to undo bad learning attempting to access a file can trigger a sub-process checking However the danger in having such powerful subsystems that whether the user has the right to access that file. All of can change as a result of learning is that they may change in these cases require the designers of the virtual machine to bad ways – as clearly happens in some humans. In some cases, anticipate kinds of things that might need appropriate checking it may be possible to undo bad changes by re-programming, or corrective actions to be performed. e.g. through discussion, advice, teaching, re-training, therapy, But there is no difficulty in building a machine that acquires etc. In other cases help may be available only when it is new competences while it is running, and also new conditions too late. Of course there are also cases where such systems for exercising old competences. If the virtual machine go wrong because of physical damage, disease, malfunction, architecture allows new goal-generators to be acquired, new corruption, etc., which may or may not be reversible, e.g. strategies for evaluating and comparing goals, new values to chemical addictions. be employed in such processes, then after some time, the machine may have goals, preferences, intentions, etc. that were III. T HE VARIETY OF MENTAL VIRTUAL MACHINES not given to it by anyone else, and which, as remarked in There is not just one kind of mind. Insect minds are different [2], can only be described as its own. It would be, to that from minds of birds and monkeys. All are different in many extent, an autonomous machine, even if the processes by which ways from adult human minds. Human minds are different such motives and motive-generators were acquired involved at different stages of development: a newborn infant, a nine- being influenced by things said and done by other people, for month old crawler, a two-year old toddler, a four year old instance teachers, heroic figures, and other role models. Indeed talker, a 50 year old professor of psychiatry. Apart from a machine might be designed specifically to derive motives, values, preferences, etc. partly on the basis of such influences, 4 As argued in this slide presentation: http://www.cs.bham.ac.uk/ tempered by experimentation on the results of trying out such research/projects/cogaff/talks/#cafe04
- 5. differences resulting from development and learning there are real robots, but that what the researchers were trying to get differences that can be caused by genetic deficiencies, and by their systems to do fell very far short of what they implied brain malfunctions caused by disease or injury. What’s more in their predictions and promises would be achieved, because obvious is that the brains are different too. So we need to find a they had not analysed the requirements adequately. general way of talking about different minds, different brains, Moreover, it could turn out that all of the currently and the different kinds of relationships that hold between (a) understood models of computation are inadequate, and entirely the minds, i.e. the virtual machines, and (b) the brains, i.e. the new kinds of virtual machine are needed, including machines physical machines in which they are implemented. that grow their own architecture, as suggested in [3]. Even virtual machines have architectures We are talking about a type of complex system with many IV. W HAT IS INFORMATION ? concurrently active parts that work together more or less There are many questions still to be answered about the harmoniously most of the time but can sometimes come concept of information used here. What is ‘information’? What into conflict. These parts are organised in an information- is an information-user? What is involved in understanding processing architecture that maps onto brain mechanisms in something as expressing a meaning or referring to something? complex, indirect ways that are not well understood. So we Is there a distinction between things that merely manipulate should ask questions like this if we wish to do deep science symbolic structures and things that manipulate them while studying a particular kind of mind: understanding them and while using the manipulation to derive • What sorts of component parts make up the architecture of new information? In how many different ways do organisms this sort of mind? acquire, store, derive, combine, manipulate, transform and use • What are their functions? information? How many of these are, or could be, replicated • For each such component, what difference would it make if in non-biological machines? Is ‘information’ as important a it were modified or removed, or connected in a different way concept for science as ‘matter’ and ‘energy’, or is it just to other components? a term that is bandied about by undisciplined thinkers and • Which parts of the architecture are involved in various popularists? Can it be defined? Is information something that processes that are found to occur in the system as a whole? should be measurable as energy and mass are, or is it more • Which parts are connected to which others, and how do like structure, which needs to be described not measured (e.g. they interact? the structure of this sentence, the structure of a molecule, the • What kinds of information do the different parts acquire structure of an organism, the properties of a toroid)? and use, and how do they obtain the information? We currently have the paradoxical situation that philoso- • How is the information represented? (It could be phers who are good at conceptual analysis are badly informed represented differently in different subsystems). about and often have false prejudices about computing • Can the system extend the varieties of information contents systems, whereas many software engineers have a deep but that it can make use of (extend its ontology), and extend the unarticulated understanding, which they use very well when forms of representation that it uses? designing, implementing, testing, debugging, maintaining, • What is the total architecture in which they function, and virtual machines, but which they cannot articulate well because how is it made up of sub-architectures? they have not been taught to do philosophy. • How are the internal and external behaviours se- Information-theory is not about information! lected/controlled/modulated/coordinated? The word ‘information’ as used (after Shannon) in so-called • Can conflicts between subsystems arise, and if so how can ‘information theory’ does not refer to what is normally they be detected, and how can they be resolved? meant by ‘information’, since Shannon’s information is a • What mechanisms in virtual machines make those processes purely syntactic property of something like a bit-string, or possible, and how are they implemented in brains? other structure that might be transmitted from a sender to • In how many ways can the different virtual machines either a receiver using a mechanism with a fixed repertoire of individually, or through their interactions go wrong, or produce possible messages. Having that sort of information does not, dysfunctional effects? for example, allow something to be true or false, or to • How did this virtual machine evolve, and what does it contradict something else. However, the more general concept have in common with evolutionary precursors and with other of information, like ‘mass’, ‘energy’ and other deep concepts contemporary animal species? used in scientific theories, is not explicitly definable. That is Answering these, and similar, questions requires a long term to say, there is no informative way of writing down an explicit investigation. One of the reasons why optimistic predictions definition of the form ‘X is Y’ if X is such a concept. All you’ll regarding imminent successes of Artificial Intelligence (or end up with is something circular, or potentially circular, when more recently robotics) have repeatedly failed is that the you expand it, e.g. ‘information is meaning’, ‘information is phenomena we are trying to explain and to replicate, are far semantic content’, ‘information is aboutness’, ‘information is more complex than anyone imagines. I.e. the main reason is what is expressed by something that refers’, ‘information is a NOT that the wrong programming languages, or the wrong difference that makes a difference’ (Bateson), and so on. models of computation were used (e.g. symbolic vs neural), But that does not mean either that the word is meaningless or that the test implementations used simulations rather than or that we cannot say anything useful about it. The same is true
- 6. of ‘energy’. It is sometimes defined in terms of ‘work’, but though U has never encountered S or anything with similar that eventually leads in circles. So how do we (and physicists) information content, for instance when U encounters a new manage to understand the word ‘energy’? sentence, diagram or picture for the first time. Even before The answer was given above in Section I: we understand any user encounters S, it is potentially usable as an information the word ‘energy’ and related concepts, by understanding their bearer. Often, however, the potential cannot be realised without role in a rich, deep, widely applicable theory (or collection U first learning a new language, or notation, or a new theory of theories) in which many things can be said about energy, within which the information has a place, and which provides e.g. that in any bounded portion of the universe there is a substantive ontology extension for U, as discussed in [16]. scalar (one-dimensional), discontinuously variable amount of This may be required for growth in self-knowledge too. it, that its totality is conserved, that it can be transmitted in A user with appropriate mechanisms has potential to various ways, that it can be stored in various forms, that it can derive infinitely many distinct items of information from a be dissipated, that it flows from objects of higher to objects small structure, e.g. infinitely many theorems derivable from of lower temperatures, that it can be used to produce forces Peano’s five axioms for arithmetic. Physically quite small that cause things to move or change their shape, etc. etc. As objects can therefore have infinite information content, in science progresses and we learn more about energy the concept combination with a reasoning mechanism, though limitations becomes deeper and more complex. The same is happening of the implementation (e.g. amount of memory available) may with ‘information.’ constrain what is actually derivable. It follows that physical An implicitly defined notion of ‘information’ structure does not constrain information content, unless a type We understand the word ‘information’ insofar as we use it in of user is specified. a rich, deep, and widely applicable theory (or collection of Information processing in virtual machines theories) in which many things are said about information, Because possible operations on information are much more e.g. that it is not conserved (I can give you information complex and far more varied than operations on matter and without losing any), that instead of having a scalar measure energy, engineers discovered, as evolution had ‘discovered’ of quantity, items of information, may form a partial ordering much earlier, that relatively unfettered information processing of containment (information I2 is contained in I1 if I2 is requires use of a virtual machine rather than a physical derivable from I1), and can have a structure (e.g. there are machine. E.g. digital electronic calculators can perform far replaceable parts of an item of information such that if those more varied tasks than mechanical calculators using cog- parts are replaced the information changes but not necessarily wheels. the structure), that two information items can share some It seems to be a basic law that increasing usable information parts (e.g. ‘Fred hates Mary’ and ‘Mary hates Joe’), that it in a virtual machine by making implications explicit requires can be transmitted by various means from one location or the physical implementation machine to use energy. Similarly object to another, that it can vary both discontinuously (e.g. (as suggested by Jackie Chappell), using greater information adding an adjective or a parenthetical phrase to a sentence, content requires more energy to be used: e.g. in storage, like this) or continuously (e.g. visually obtained information sorting and processing information. So biological species able about a moving physical object), that it can be stored in to acquire and process vast amounts of information must be various forms, that it can influence processes of reasoning near the peak of a food pyramid, and therefore rare. and decision making, that it can be extracted from other information, that it can be combined with other information Causal and correlational theories of meaning are false to form new information, that it can be expressed in different It is often thought that learning to understand S as referring syntactic forms, that it can be more or less precise, that it can to X, requires an empirical discovery that there is a causal express a question, an instruction, a putative matter of fact, relation between S and X, such as that occurrences of X always and in the latter case it can be true or false, known by X, or often cause occurrences of S to come into existence, or unknown by Y, while Z is uncertain about it, etc. etc. such that the occurrence of X is shown empirically to be a reliable predictor of the occurrence of S. That this theory Information is relative to a user, or potential user is false is shown by the fact that you have no reason to Whereas energy and physical structures simply exist, whether believe that occurrences of the word ‘eruption’ are correlated used or not, information in a physical or virtual structure with or reliable predictors of eruptions, and moreover you can S is only information for a type of user. Thus a structure understand the phrase ‘eruption that destroyed the earth 3000 S refers to X or contains information about X for a user years ago’ even though it is impossible for any such correlation of S, U. The very same physical structure can contain or causal link to exist, since the earth was not destroyed then. different information for another user U’, or refer to something Further we can use concepts that refer to abstract entities different for U’, as shown by ambiguous figures, and also whose existence is timeless, such as the number 99, or the written or spoken languages or notations that some people shortest proof that 2 has no rational square root. So empirical understand and others do not. This does not make information correlations and causal influences are impossible in those inherently subjective any more than mountains are subjective cases. because different people are capable of climbing different mountains. (Indexicality is a special case, discussed below.) Information content determined partly by context The information in S can be potentially usable by U even It is sometimes thought that artificial minds would never
- 7. be able to grasp context-sensitive information. For example, parts cannot access. When we ask ‘Did X know that P?’ it an information-bearing structure S can express different is not clear whether the answer must be ‘yes’ in cases where information, X, X’, X”, for the same user U in different some part of X made use of the information that P. E.g. human contexts, e.g. because S includes an explicit indexical element posture control mechanisms use changes in optical flow. Does (e.g. ‘this’, ‘here’, ‘you’, ‘now’, or non-local variables in that mean that when you are walking around you know about a computer program). Indexicality can make information optical flow even though you don’t know that you know it? incommunicable in the sense that the precise information Your immune system and your digestive system and various content of one user cannot be transferred to another. (Frege, metabolic processes use information and take decisions of for example, showed that one user’s use of the word “I” has many kinds. Does that mean that you have the information, that a sense that another person is incapable of expressing.) A you know about the information, that you use the information? corollary is that information acquired by U at one time may Some people might say yes others no, and some may say not be fully interpretable by U at another time, because the that it depends on whether you know that you are using the context has changed, e.g. childhood ‘memories’. In [17] it was information. argued that such indexicality accounts for the ‘ineffability’ of Likewise there are different parts of our brains that evolved qualia. However this does not usually prevent the intended at different times that use different kinds of information function of communication from being achieved. The goal of (even information obtained via the same route, e.g. the communication is not to replicate the sender’s mental state, or retina or ear-drum, or haptic feedback). Some of them are information content, in the receiver, but to give the receiver evolutionarily old parts, shared with other species (e.g. posture information that is adequate for some purpose. control mechanisms), some newer and possibly some unique Many structures in perceptual systems change what to humans, (e.g. human face recognition mechanisms, and information they represent as the context changes. Even if mechanisms that can learn to read music and other notations). what is on your retina is unchanged after you rapidly turn A deep feature of at least the human architecture, and your head 90 degrees in a room, the visual information will perhaps some others also, is that sensors and effectors be taken to be about a different wall – with the same wallpaper providing interfaces to the environment can be shared between as the first wall. Many examples can be found in [18]. different subsystems, as described in [20]. E.g. your vision Sometimes U takes S to express different meanings in mechanisms can be shared between: posture control, visual different contexts because S includes a component whose servoing of manipulation actions, mechanisms involved in semantic role is to express a higher order function which reading instructions, and affective mechanisms that appreciate generates semantic content from the context, e.g. ‘He ran aesthetic qualities of what is seen. Your walking mechanisms after the smallest pony’. Which pony is the smallest pony can be shared between a subsystem concerned with moving can change as new ponies arrive or depart. More subtly what to the door and also a subsystem concerned with social counts as a tall, big, heavy, or thin something or other can vary communication, e.g. flirting by walking suggestively. We can according to the range of heights, sizes, weights, thicknesses describe such architectures as using ‘multi-window’ perception of examples in the current environment. and ‘multi-window’ action, whereas current artificial systems There are many more examples in natural language that lead mostly use only ‘peephole’ perception and ‘peephole’ action, to incorrect diagnosis of words as vague or ambiguous, when where input and output streams from each sensor or two each they actually express precise higher order functions, applied effector are go along channels of restricted functionality. to sometimes implicit arguments, e.g. ‘big’, ‘tall’, ‘efficient’, Sometimes the sharing is concurrent and sometimes ‘heap’, or ‘better’ (discussed in [19]).5 sequential. Conflict resolution mechanisms may be required Information content shared between users when concurrent sharing is impossible. Some of the Despite the above, it is sometimes possible for X to mean the mechanisms that detect and resolve conflicts may be same thing to different users U and U’, and it is also possible inaccessible to self-monitoring (discussed later), so that an for two users who never use the same information bearers individual may be unaware of important decisions being taken. (e.g. they talk different languages) to acquire and use the same Much philosophical, psychological, and social theorising information. This is why relativistic theories of truth are false: misguidedly treats humans as unitary information users, although I can believe that my house has burned down while including Dennett’s intentional stance and what Newell refers my neighbour does not, one of us must be mistaken: it cannot to as ‘the Knowledge level’. be true for me that my house has burned down but not true for my neighbour. Truth is not ‘for’ anyone. Meaning depends Just the beginning of an analysis of ‘information’ on the user. Truth does not. The analysis of the concept of ‘information’ presented here amounts to no more than a small fragment of the V. I NFORMATION - USING SUBSYSTEMS full theory of types of states, events, processes, functions, An information-user can have parts that are information mechanisms and architectures that are possible in (virtual users. A part can have and use some information that other and physical) information-processing machines. I doubt that anyone has produced a clear, complete and definitive 5 This idea is developed in the context of Grice’s theory of communication, list of facts about information that constitute an implicit with implications for the evolution of language, here: http://www.cs. bham.ac.uk/research/projects/cosy/papers/#dp0605 Spa- definition of how we (the current scientific community well- tial prepositions as higher order functions. educated in mathematics, logic, psychology, neuroscience,
- 8. biology, computer science, linguistics, social science, artificial brains, add only more of the same. However, in humans and intelligence, physics, cosmology, ...) currently understand the many other animals there additional kinds of information and word ‘information’. information processing. E.g. there’s a great deal still to be said about the Brains are needed for more than movement molecular information processing involved in development of Most organisms manage without brains. Why not all? an individual from a fertilised egg or seed, and in the huge One function is resolving conflicts between different parts variety of metabolic processes including intrusion detection, responsible for decisions that could be incompatible, e.g. damage repair, transport of materials and energy, and control decisions to move one way to get to food or to move another by hormones and neurotransmitters. It may be that we shall way to avoid a sensed predator. This requires coordination, one day find that far more of the brain’s information processing possibly based on dynamically changing priorities. A different is chemical than anyone dreams now is possible. kind of requirement, discussed later, is doing more complex A more complete theory would provide a more complete processing of information, e.g. in order to acquire a better implicit definition of the concept ‘information’ required understanding of the environment, or to acquire something for understanding natural and artificial systems (including like a terrain map, or to plan extended sequences of actions. far more sophisticated future artificial systems). A hundred In organisms with many complex parts performing different years from now the theory may be very much more deep functions it may also be necessary to coordinate internal and complex, especially as information processing machines changes, for example all the changes involved in reaching produced by evolution still seem to be orders of magnitude puberty in humans, or during pregnancy, where many internal more complex than any that we so far understand.6 changes and external behaviours need to be coordinated. Lewis Wolpert wrote, in an Observer book review, March VI. B IOLOGICAL INFORMATION PROCESSING 24, 2002: ‘First the only function of the brain from an All living things, including plants and single-celled evolutionary point of view is to control movement and so organisms, process information insofar as they use sensors to interaction with the environment. That is why plants do not detect states of themselves or the environment and use that have brains.’7 This often quoted, but grossly misleading, claim information either immediately or after further processing to reflects a widespread current focus on research that models select from a behavioural repertoire. The behaviour may be brains as sensory-motor control mechanisms, e.g. [21]. Even externally visible physical behaviour or internal processes. if the original function of brains was to control movement, While using information an organism normally also uses up much of what human brains do has nothing to do with control stored energy (usually chemical energy), so that it also needs of movement – for example explaining what is observed, to use information to acquire more energy. predicting future events, answering a question, and doing There are huge variations both between the kinds of mathematics or philosophy. information contents used by different organisms and between So we see that information in organisms may be implicit different ways in which information is acquired, stored, and transient, implicit and enduring, explicit and capable of manipulated and used by organisms. The vast majority of multiple uses, used only locally, used in controlling spatially organisms use only two kinds of information: (a) genetic distributed information processing, and may vary in kind of information acquired during evolution, used in replication, abstraction, in the ontology used, in the form of representation physical development and maintenance of the individual, and used and in the kinds of manipulation that are available. (b) transiently available information used by online control Varieties of explicit information systems. I call the latter ‘implicit’ information. There may Only special conditions bring about evolution of mechanisms also be some less transient implicit information produced by that create, store and use explicit, that is enduring, re-usable, gradually adjusted adaptive mechanisms. information structures that can be used in different ways Since the vast majority of species are micro-organisms (e.g. forming generalisations, making predictions, building the vast majority of information-using organisms can use terrain maps, forming motives, making plans, remembering only implicitly represented information; that is to say they what happened when and where, communicating to other use only information that is available during the transient individuals, etc.) As explained in [22], this requires new states of activation produced by information being acquired architectures involving mechanisms that are not so directly and used, and the information is represented only transiently involved in the sensorimotor relationships and their control.8 in activation states of sensors, motors and intervening There may have been intermediate stages of evolution mechanisms, and also in parameters or weights modified by in which non-transient information was stored only in the adaptive feedback mechanisms. environment, e.g. in chemical trails and land-marks used as The short term transient implicit information in patterns of sources of information about the presence of food or predators, activation and the longer term implicit information in gradually or about the routes followed by con-specifics (pheromone changing adaptive mechanisms together suffice for most living trails), or about location relative to a nest or an enduring source things – most of which do not have brains! In some animals 7 http://observer.guardian.co.uk/science/story/0, 6 For a draft attempt to answer the question ‘What is infor- ,673268,00.html mation?’ see http://www.cs.bham.ac.uk/research/projects/ 8 See ‘Sensorimotor vs objective contingencies’ http://www.cs. cogaff/misc/whats-information.html bham.ac.uk/research/projects/cosy/papers/#dp0603
- 9. of food. (Some insects can use land-marks so this capability individual B believes is the case, but the functional roles probably evolved a long time ago.) This is possible only for of those two forms of representation will be different.9 In animals in an environment with stable structures, unlike some particular, the former will be ‘referentially transparent’ and the marine environments. latter ‘referentially opaque’, and the conditions for truth of the Mechanisms that evolved to use external enduring two beliefs are quite different – something young children do information may have been precursors to mechanisms not learn in the first few years. using internally stored explicit information. There probably The ability to think and reason about the mental contents were many different evolutionary transitions, adding extra of another has much in common with the ability to think and functionality. In previous papers (e.g. [23]) colleagues and reason about one’s own mental contents as far as mechanisms I have emphasised three main categories of competence and formalisms are concerned, though obviously different requiring different sorts of architectural components, namely sensor mechanisms are involved, i.e. external and internal. reactive, deliberative and meta-management capabilities, but However the evolutionary benefits are very different. It is there are many intermediate cases and different sorts of not clear which evolved first. Probably both types of meta- combinations of cases that need to be understood – not only semantic competence co-evolved, each helping to enrich the for understanding how things work when everything functions other. normally, but also in order to understand the many ways things Incidentally, nothing said here implies that any organism can go wrong, including both the consequences of physical has full self-knowledge. On the contrary, self-knowledge malfunctions and also the consequences of dysfunctional will inevitably be associated with a bottleneck with limited processing of information in virtual machines. capacity – a point that has relevance for psycho-analysis. Compare [25]. Deliberative mechanisms A very small subset of organisms (and some machines) have Perhaps a more widely-shared ability to formulate internal a ‘deliberative’ information-processing capability insofar as questions or goals was an evolutionary precursor to meta- they can construct a set of information structures of varying semantic competence, since formulating a yes-no question complexity (e.g. plans, predictions, theories), then compare (‘Will it rain today?’) requires an ability to represent their merits and de-merits in relation to some goal, and propositions that are capable of being true or false without produce new information structures describing those tradeoffs, a commitment to their truth value. The same is true of having then select one of the alternatives and make use of it. That desires and intentions. So the ability to represent ‘X thinks ability to represent and reason about hypotheticals was one I will eat him’ and to reason about the consequences of X’s of the first aspects of human processing modelled in AI thinking that (e.g. X will run away) may have arisen as a subsystems, but it was probably one of the last to evolve, modification of the ability to represent one’s own goals (‘I and only a very tiny subset of organisms have it in its richest will eat X’) that have not been achieved, plans that have not form, described in [24]. yet been carried out, tentative predictions that have not yet It is now fashionable to contrast that early ‘symbolic’ AI been tested, and questions that have not yet been answered. work with biologically inspired AI work. But that is just silly, Old and new in the same architecture since all of the work was biologically inspired insofar as it was It is not always remembered that besides having such an attempt to model processes that occur in human beings. A sophisticated (and possibly unique) capabilities for explicit human making a plan or proving a theorem is just as much a manipulation of information, humans share many information biological organism as a human running, jumping or catching processing capabilities with other species that lack the a ball, even though more species share the capabilities required distinctively human capabilities. For instance many animals for the latter activities. Biology is much richer than some have excellent vision, excellent motor control, abilities to researchers seem to realise. cooperate on certain tasks, hunting capabilities, the ability Meta-semantic competences to avoid predators, nest-building capabilities, as well as Another relatively rare biological information-processing the information-processing involved in control of bodily capability involves the ability to refer to, reason about, or functions. The notion that somehow human cognition, or care about things that themselves contain, or use information. human conscious processes can be studied and replicated For instance when you discover what someone thinks, and without taking any account of how these more general animal when you worry about someone’s motives you are using mechanisms work and how they interact with the distinctively such meta-semantic competence. Humans can also apply this human mechanisms, may lead both to a failure to understand meta-semantic competence to themselves (though probably humans and other animals and also to designs for robots that not at birth: the architecture needs time to develop). Having don’t perform as required. this ability requires a more complex architecture than merely Some of the ideas developed with colleagues in Birmingham being able to refer to things without semantic content (e.g. about the different components in such a multi-functional physical objects and processes). That’s partly because having architecture are reported in the paper by Brigitte Lorenz and meta-semantic content involves representing things that are Etienne Barnard presented at this conference and will not be treated as true in only a hypothetical, encapsulated, way. The 9 J. Barnden’s ATT-META project has developed a way of making that very same form of representation may be used for what an distinction which is also related to the ability to think metaphorically. http: individual A believes is the case and for what A thinks another //www.cs.bham.ac.uk/∼jab/ATT-Meta/
- 10. developed here. From this standpoint it is almost always a to run, if an interpreter is available. Moreover if it were mistake to ask questions like ‘How do humans do X?’ Instead necessary, then substantive ontology extension during learning we should ask ‘How do different subsystems in humans do and development would be impossible. X?’ (e.g. X = control actions, interpret visual input, learn, store Nobody is yet in a position to say what the prelinguistic information, react to interruptions, generate motives, resolve languages do and do not share with human communicative conflicts, etc.) And we should expect different answers not languages. In particular, it may be the case that the only for different subsystems, as in Trehub’s [26], but also main qualitative competences required for human language for humans at different stages of development, in different use already exist in these pre-linguistic competences in cultures, with and without brain damage, etc. young children and other animals, and that the subsequent evolutionary developments related to human language were mainly concerned (a) with developing means of VII. P RE - LINGUISTIC COMPETENCES generating adequately articulated external behaviours to It is often assumed that there is a massive discontinuity communicate their structures, (b) improving perception of between human linguistic competence and other competences such behaviours, and (c) extending internal mechanisms (e.g. (e.g. as argued by Chomsky in [27]), though there are many short term memories) that are equally useful for sophisticated who hotly dispute this (e.g. Jablonka and Lamb [28]). But internal information processing (such as planning) and for people on both sides of the dispute make assumptions about communication with others. human language that may need to be challenged if we are If so, information (unconsciously) acquired, used and stored to understand what human minds are, and if we wish to for later use in early childhood may have much richer produce working human-like artificial minds. In particular it structures and deeper semantic content than has previously is often assumed that the essential function of language is been thought possible. Whether it includes the kind of content communication. But, as argued in [29], and more recently required to support psychoanalytic theories remains open. in [30] many of the features that make such communication Non-auditory communication possible (e.g. the ability to use varieties of information with It is easier to achieve a large collection of perceptually rich and varied structures and with compositional semantics, discriminable signals by using independently movable fingers, and non-communicative abilities to check whether some state hands, mouth, eyes, head than to do it all by modulating of affairs matches a description, to notice a gap in information, an acoustic signal – especially as that would interfere with expressed in a question) are also requirements for forms breathing. So, if some animals already had hands for which of information that are involved in perception, planning, they were learning to produce (and perceive, during controlled expressing questions, formulating goals, predicting, explaining execution) large numbers of distinct manipulative competences and reasoning. concerned with obtaining and eating food, grooming, climbing, From that viewpoint, communication between individuals making nests, fighting, threatening, etc., then perhaps the in a public medium was a secondary function of language first steps to communicative competence used structured which evolved only after a more basic kind of competence movements, as in that involved producing and perceiving evolved – the ability to use an internal, non-communicative, structured sign language, rather than vocal language. language in mental states and processes such as perceiving, There is much suggestive evidence, including the fact that thinking, supposing, intending, desiring, planning, predicting, humans find it almost impossible not to move hands, eyes, remembering, generalising, wanting information, etc. and in facial muscles, etc. when talking, even when talking on the executing intentions or plans. phone to someone out of sight. More compelling is the ease It is clear that human children who cannot yet talk, and with which deaf babies are able to learn a sign language, and many animals that do not use an external language can the reported fact that some children with Down syndrome seem perceive, learn, think, anticipate, have goals, threaten, be to learn to communicate more easily if sign language is used. puzzled, play games, carry out intentions and learn and use Most compelling is the case of the Nicaraguan deaf children facts about causation. I know of no model of how any of that who invented their own highly sophisticated sign language, can be done without rich information processing capabilities leaving their teacher far behind. See [32].10 of kinds that require the use of internal languages with An implication of this is that the human ability to develop compositional semantics. But I know of no detailed model linguistic competence does not depend on learning a language of what those ‘prelinguistic’ languages are, how they evolved, that is already in use by others. Language learning then how they develop, how they work, etc. appears to be a process of collaborative, creative problem- Fodor, in [31], postulated a ‘language of thought’ (LOT) solving, which may be constrained by the prior existence of a which was supposed to be innate, available from birth and social language, but need not be. capable of expressing everything that ever needs to be said In [33] Arbib proposed that action recognition and imitation by any human being, but he left most questions about how was a precursor to the evolution of language, but the arguments it worked unanswered and was not concerned with non- human animals. Moreover he supposed that external languages 10 http://www.indiana.edu/∼langacq/E105/Nicaragua. are translated into the LOT, whereas there is no reason html ‘A Linguistic Big Bang’, by Lawrence Osborne The New York Times October 24, 1999. A five minute video including examples of the invented to believe such translation is necessary, just as compiling language is available here http://www.pbs.org/wgbh/evolution/ to machine code is not necessary for computer programs library/07/2/l 072 04.html
- 11. are somewhat different, and do not share our hypothesis in context demands, in collaborative problem solving rather [29] that the existence of a rich non-communicative language than in a process of information transfer. (This is related to preceded the evolution of communicative language, though Grice’s maxims of communication, and to current concerns the commentary by Bridgeman [34] makes a similar point, with ‘situatedness’ in language understanding, perception emphasising pre-linguistic planning capabilities. and action. http://www.cs.bham.ac.uk/research/ Once something like this form of structured communication projects/cosy/papers/#dp0605) had developed, enabling information that was previously If the contents of linguistic communication between only represented in internal languages to be shared between individuals make such use of the context then it may be individuals via a public language, the enormous advantages equally true for thought processes within an individual. In that of having such a communicative competence, both in solving case, much of the information content of your mind is outside problems that required collaboration and in accelerating you. This may be one of the ideas people who emphasise cultural transmission, might have led to an unusually rapid the importance of ‘embodiment’ are getting at. However, process of selection for changes that enhanced the competence everything said here could apply to minds in individuals in various ways, e.g. developing brain regions specialised at with simulated bodies located in simulated environments. The storage of ever larger vocabularies and more complex rules for important thing is what sorts of structures, processes and construction and interpretation of action sequences. causal interactions occur within the individual’s information Another development might have been evolution of brain processing architecture and how they are related to its mechanisms to support construction and comparison of environment. Whether that environment is physical or another more complex symbolic structures, e.g. more deeply nested virtual machine environment makes little or no difference to structures, and structures with more sub-structures. the kind of mind discussed here. The nature of linguistic communication From sign language to sound language If linguistic communication evolved from collaborative non- Use of sign language has many problems. As has often been linguistic activities, controlled by sophisticated internal noted, it can work only when the sender’s hands are free, languages, this must change our view of human language. and when the receiver can see the sender, thereby ruling out It is often thought that linguistic communication involves a communication in many situations where visual attention must process whereby an information structure in one individual be used for another task, or where there are obstructions gets encoded in some behaviour which is perceived and to vision, or no light is available. So perhaps our ancestors decoded by another individual who then constructs the same started replacing hand signals with sounds where possible. information internally. However, in collaborative, creative, This could then lead in the usual way to a succession of problem solving the shared physical and task contexts provide evolutionary changes in which the vocal mechanisms, auditory much of the information, and need not be communicated. All mechanisms, and the neural control systems evolved along that is needed is whatever suffices to produce desired results, with physiological changes to the vocal tract, etc. But the without copying an information structure from one brain to ability to learn sign languages remains intact, even if not used another. by most people. If a mother hands a child a piece of fruit, the child may If all this is correct then it may have deep implications thereby be triggered to form the goal of eating it, without the for clinical developmental psychology as well as for mother having had the goal of eating it. Moreover, the mother understanding precisely what the effects of various kinds of does not need to anticipate the precise movements of bringing brain damage are. the food to the mouth nor the precise chewing and swallowing movements produced by the child. Those details can be left to the child. More generally, context allows communication to VIII. A RCHITECTURES THAT GROW THEMSELVES be schematic rather than concrete and detailed. This is obvious when A asks ‘Where are my keys?’ and Much research in AI and robotics is concerned with what B replies ‘Look behind you?’. B may not know exactly sort of architecture an intelligent system needs, and various where the keys are except that they are on the table behind rival architectures are proposed. If the architecture needs to A. A turns round and sees the keys, getting the precise grow itself, some of this research effort may have been wasted. information he wanted, as a result of B’s communication, even It was argued in [3] that the architecture should not be regarded though B did not have that information. Both have achieved as genetically determined, but as grown using both innate their communicative goals, but not by transmitting a specific and acquired meta-competences influenced by the physical information structure from B to A. and cultural environment, as depicted in Figure 1. So there So, many kinds of linguistic and non-linguistic commu- may be considerable differences in how adult minds work, nication involve communication of a partial or schematic as a result of a different developmental trajectories caused structure, leaving gaps to be filled by the receiver using by different environments. This might affect kinds of self- information that may or may not be available to the monitoring, kinds of self control, kinds of learning ability, sender. In that case, many linguistic expressions may be and so on. Another implication is that attempts to quantify the best thought of as having a higher-order semantics, to influence of genes and environment in terms of percentages is be applied to more specific non-linguistic information as completely pointless.
- 12. presentations are in the Birmingham CogAff and CoSy project web sites. Jackie Chappell and Gerhard Pratl made useful comments on an early draft. Fig. 1. This indicates some of the many and varied relationships between the genome and behaviours, produced at different stages of development after layers of competences and meta-competences have been built up. (From [3]) IX. C ONCLUSION : MACHINES IN GHOSTS Although I was asked to report on developments in AI that might be relevant to psycho-analysis I have instead focused on some very general features of requirements for AI models rather than specific AI models, because I think that despite all the advances achieved in AI, the models are still very primitive compared with human minds and have a very long way to go before they can be directly relevant to working therapists dealing with real humans. Nevertheless, I hope it is now clear why a human-like machine, a human-like intelligent ghost, and indeed a human- like human must contain very specific kinds of information processing virtual machines. So our ability to understand human minds and ways they can go wrong, including acquiring false, beliefs, inappropriate motives, inadequate strategies, tendencies to be over-emotional, and worse, must be informed by deep knowledge about information-processing systems of the appropriate kinds. The complexity of the machine, and the fact that self- monitoring mechanisms within it provide only a very limited subset of information about what is going on cause many individuals to start wondering what is really going on in them. Ignorance about information-processing mechanisms can make things seem so mysterious that some people invoke a special kind of stuff that is quite unlike physical matter. From there it is a short step to immortality, souls, etc., and the conceptual confusions discussed by Ryle in [35]. But if we don’t go that far we may still come up with deep new ways of thinking about phenomena that were previously thought resistant to computer modelling. This could, among other things, lead to a revolution in psychoanalysis. ACKNOWLEDGEMENT Thanks to Dean Petters, Jackie Chappell, Matthias Scheutz, Margaret Boden, members of the EU-Funded Cosy Project, and other collaborators and discussants over the years. Some of the ideas here were inspired by John McCarthy and Marvin Minsky, as well as Immanuel Kant, [6]. Related papers and
- 13. R EFERENCES [27] N. Chomsky, Aspects of the theory of syntax. Cambridge, Mass: MIT Press,, 1965. [1] D. C. Dennett, Brainstorms: Philosophical Essays on Mind and [28] E. Jablonka and M. J. Lamb, Evolution in Four Dimensions: Genetic, Psychology. Cambridge, MA: MIT Press, 1978. Epigenetic, Behavioral, and Symbolic Variation in the History of Life. [2] A. Sloman, The Computer Revolution in Philosophy. Has- Cambridge MA: MIT Press, 2005. socks, Sussex: Harvester Press (and Humanities Press), 1978, [29] A. Sloman, “The primacy of non-communicative language,” in The http://www.cs.bham.ac.uk/research/cogaff/crp. analysis of Meaning: Informatics 5 Proceedings ASLIB/BCS Conference, [3] J. Chappell and A. Sloman, “Natural and artificial Oxford, March 1979, M. MacCafferty and K. Gray, Eds. London: Aslib, meta-configured altricial information-processing systems,” 1979, pp. 1–15, http://www.cs.bham.ac.uk/research/projects/cogaff/81- International Journal of Unconventional Computing, 2007, 95.html. http://www.cs.bham.ac.uk/research/projects/cosy/papers/#tr0609. [30] A. Sloman and J. Chappell, “Computational Cognitive Epigenetics [4] A. Sloman and J. Chappell, “The Altricial-Precocial Spectrum for (Commentary on [28]),” Behavioral and Brain Sciences, 2007. Robots,” in Proceedings IJCAI’05. Edinburgh: IJCAI, 2005, pp. 1187– [31] J. Fodor, The Language of Thought. Cambridge: Harvard University 1192, http://www.cs.bham.ac.uk/research/cogaff/05.html#200502. Press, 1975. [5] S. Harnad, “The Symbol Grounding Problem,” Physica D, vol. 42, pp. [32] L. Osborne, “‘A Linguistic Big Bang’,” The New York Times, 24 October 335–346, 1990. 1999, http://www.indiana.edu/ langacq/E105/Nicaragua.html. A five [6] I. Kant, Critique of Pure Reason. London: Macmillan, 1781, translated minute video including examples of the invented language is available (1929) by Norman Kemp Smith. at http://www.pbs.org/wgbh/evolution/library/07/2/l 072 04.html. [7] A. Sloman, “The irrelevance of Turing machines to AI,” in Computa- [33] M. A. Arbib, “From monkey-like action recognition to human language: tionalism: New Directions, M. Scheutz, Ed. Cambridge, MA: MIT An evolutionary framework for neurolinguistics,” Behavioral and Brain Press, 2002, pp. 87–127, http://www.cs.bham.ac.uk/research/cogaff/00- Sciences, vol. 28, no. 2, pp. 105–124, 2005. 02.html#77. [34] B. Bridgeman, “Action planning supplements mirror systems in language [8] A. Damasio, Descartes’ Error, Emotion Reason and the Human Brain. evolution,” Behavioral and Brain Sciences, vol. 28, no. 2, pp. 129–130., New York: Grosset/Putnam Books, 1994. 2005. [9] R. Picard, Affective Computing. Cambridge, Mass, London, England: [35] G. Ryle, The Concept of Mind. London: Hutchinson, 1949. MIT Press, 1997. [10] A. Sloman and M. Croucher, “Why robots will have emotions,” in Proc 7th Int. Joint Conference on AI. Vancouver: IJCAI, 1981, pp. 197–202. [11] A. Sloman, “Towards a grammar of emotions,” New Universities Quarterly, vol. 36, no. 3, pp. 230–238, 1982, http://www.cs.bham.ac.uk/research/cogaff/96-99.html#47. [12] A. Sloman, “Real time multiple motive-expert systems,” in Proceedings Expert Systems 85, M. Merry, Ed. Cambridge University Press, 1985, pp. 213–224. [13] A. Sloman, “Damasio, Descartes, alarms and meta-management,” in Proceedings International Conference on Systems, Man, and Cybernetics (SMC98), San Diego. IEEE, 1998, pp. 2652–7. [14] A. Sloman, R. Chrisley, and M. Scheutz, “The architectural basis of affective states and processes,” in Who Needs Emotions?: The Brain Meets the Robot, M. Arbib and J.-M. Fellous, Eds. Oxford, New York: Oxford University Press, 2005, pp. 203–244, http://www.cs.bham.ac.uk/research/cogaff/03.html#200305. [15] A. Sloman, “Do machines, natural or artificial, really need emotions?” 2004, Invited talk: http://www.cs.bham.ac.uk/research/projects/cogaff/talks/#cafe04. [16] A. Sloman, “‘Ontology extension’ in evolution and in development, in animals and machines,” University Aaron Sloman Aaron Sloman studied for a BSc in mathematics and physics of Birmingham, UK, 2006, PDF presentation: at Cape Town in 1956, then, after flirting for a while with mathematical http://www.cs.bham.ac.uk/research/projects/cosy/papers/#pr0604. logic was seduced by philosophy, and completed a D.Phil at Oxford in 1962, [17] A. Sloman and R. Chrisley, “Virtual machines and consciousness,” defending Kant’s philosophy of mathematics. He later learnt that AI provided Journal of Consciousness Studies, vol. 10, no. 4-5, pp. 113–172, 2003. the best tools for doing philosophy and since about 1970 has been working on [18] A. Berthoz, The Brain’s sense of movement, ser. Perspectives in interdisciplinary problems concerned with understanding the space of possible Cognitive Science. London, UK: Harvard University Press, 2000. designs for behaving systems including humans, other animals and possible [19] A. Sloman, “How to derive “better” from “is”,” future robots. In recent years this has particularly involved collaborating with American Phil. Quarterly, vol. 6, pp. 43–52, Jan 1969, a biologist studying cognition in birds and roboticists designing robots able to http://www.cs.bham.ac.uk/research/cogaff/sloman.better.html. perceive and manipulate 3-D objects in the environment. The various strands [20] A. Sloman, “The mind as a control system,” in Philosophy and the of his work are summarised in http://www.cs.bham.ac.uk/∼axs/ Cognitive Sciences, C. Hookway and D. Peterson, Eds. Cambridge, my-doings.html He is a Fellow of AAAI, ECCAI and SSAISB, and was UK: Cambridge University Press, 1993, pp. 69–110. given an honorary DSc by Sussex University in 2006. [21] M. Lungarella and O. Sporns, “Mapping information flow in sensorimotor networks,” PLoS Computational Biolology, vol. 2, no. 10:e144, 2006, dOI: 10.1371/journal.pcbi.0020144. [22] A. Sloman, “Sensorimotor vs objective contingencies,” University of Bimringham, School of Computer Science, Tech. Rep. COSY-DP-0603, 2006, http://www.cs.bham.ac.uk/research/projects/cosy/papers/#dp0603. [23] A. Sloman, “Architecture-based conceptions of mind,” in In the Scope of Logic, Methodology, and Philosophy of Science (Vol II). Dordrecht: Kluwer, 2002, pp. 403–427, (Synthese Library Vol. 316). [24] A. Sloman, “Requirements for a Fully Deliberative Architecture,” University of Bimringham, School of Computer Science, Tech. Rep. COSY-DP-0604, 2006, http://www.cs.bham.ac.uk/research/projects/cosy/papers/#dp0604. [25] M. Minsky, “Matter Mind and Models,” in Semantic Information Processing, M. Minsky, Ed. Cambridge, Mass,: MIT Press,, 1968. [26] A. Trehub, The Cognitive Brain. Cambridge, MA: MIT Press, 1991, Online http://www.people.umass.edu/trehub/.