However there are a large number of obstructions to overcome with the connectionist computational theory of mind, and although proponents of it want us to give up on rule-based systems of explanation such as the classicists version, these things need to be rectified before it can be considered a sufficient theory of mind.

Although the connectionist version has a viable approach to association processing it seems to fail on such tasks as language and reasoning. The hindrances range in magnitude from minute to monstrous and include quandaries such as problems with the concept of an individual, the problem of compositionality, the problem of quantification, recursive thoughts/propositional problems, trouble with commonsense questions and last but not least the problem of systematicity.

The problem with the concept of an individual is an painless point to understand but is not so easily solved. If we have a set of identical twins or even something a bit more general like two trees of the same species, height, age etc. the connectionist system is blind to the fact that they are actually separate entities. There are a number of proposed responses to this dilemma but none of them are adequate.

The archetypal property of all our languages is compositionality and this is where another predicament with the connectionist CTM occurs. Representations are built out of parts and have their meaning based on the meanings of all the parts and from the way they are combined. Subsequently, due to the fact that our thoughts are built out of concepts, and are not stored as whole entities within our mind, we encounter major problems and all the attempts to rectify them turn out to be substandard halfway measures. After all, babies eating slugs, and slugs eating babies are two very distinct ideas whose meanings are assembled on the fly using syntax, and even though you may never have seen either sentence before you can understand it with ease.

Systematicity on the other hand seems to be one of the more obtrusive problems the connectionist CTM must face, and was identified by Fodor & Pylyshyn (1988). It states that the ability to think/create/understand a sentence of a particular structure is intrinsically connected to the ability to think/create/understand a sentence that has a related structure, in that, there is no human that can understand the meaning of the sentence “Jason loves Ashley” but fails to understand the concept of “Ashley loves Jason”. Regrettably though, connectionist models, even once they have been trained to recognize one sentence of the previous example, still fail to recognize the second. Systematicity must be guaranteed to work in connectionist models for their theory of mind to be a viable option as this is a key component of human intelligence, and, according to Fodor & Pylyshyn, is a given in the classicists approach.

The connectionist computational theory of mind, although on distant inspection appears to be a much healthier option than other models, seems to have just as many drawbacks as them, and once again, cannot be considered a viable option for a complete theory of mind without at least some of these issues being overcome.

The ones we have discussed would be a good starting point but there are also other smaller, but still very relevant ones that must be addressed. Such as, the length of time it takes to train a connectionist model vs the length of time it takes a human to learn similar tasks, the fact that back propagation in connectionist models appears to ‘cheat’ as there is no evidence of our minds working in such ways and the fact that connectionist models fail to recognize our recursive thoughts i.e. a proposition embedded within another proposition. Until such times as these are solved, the connectionist CTM is promising, but still has a lot of work ahead of it.

The two main arguments that Descartes uses to put forward his thoughts are that of clear and distinct ideas, and the indubitability argument. However, in this paper it will be shown that the clear and distinct ideas argument is a very weak attempt at explaining the mind – body dilemma and realistically has no force, and while the indubitability argument is definitely a lot stronger than the one from clear and distinct ideas it still has major flaws that were successfully attacked recently by those such as Gilbert Ryle and Ludwig Wittgenstein, but as early as in Descartes own life time by Princess Elizabeth of Bohemia.

Dualism states that bodies are spatial, are subject to mechanical laws and they are open to public observation, while the mind is non-spatial, not subject to mechanical laws, and is private. Hence, we have privileged access to the contents of our own mind that cannot be mistaken or corrected.

By the fact that Descartes cannot prove that anything is real other than his own mind, he comes to the radical conclusion through meditation that only his mind exists, for a deceitful demon could be deceiving him or he could be hallucinating and any number of other things. Hence the indubitability argument comes to life based around the premise that you can never come to doubt the existence of your own mind, but can reasonably doubt everything else.

However there do appear to be a number of problems with Descartes reasoning and Wittgenstein, with who I agree, suggests that solipsism is absurd. To suggest that the body, and more importantly everything outside ones mind can be doubted, seems to defy all logic. Nobody could truly suggest that the earth does not exist, the fact that the world exists seems to be a given, in that it is a common sense proposition. “The reasonable man,” says Wittgenstein, “does not have certain doubts; that is to say, the questions that we raise and our doubts depend on the fact that some propositions are exempt from doubt.” (Ludwig Wittgenstein, On Certainty) The idea of common sense propositions were first brought to light by Moore against Descartes sceptical method and was later commented on by Wittgenstein but we have no time to discuss these comments here.

It would seem that Descartes starts off on the wrong foot and continues trying to defend a flawed position. One of the more difficult problems to solve for any dualist is that of interaction. Descartes claimed that we have a two-way psychophysical interaction with the physical world and explained the problem of interaction thusly; when we have desires, wants, motivations, a mental state occurs, when such a mental state occurs it creates a certain state in our brain, which in turn moves our bodies, which then causally influence the physical world. The physical world then affects our minds in the same way; we perceive our surroundings using our five senses and send information to our brain which passes to our mind.

One of the more damaging ideas raised on this point was brought out by Princess Elizabeth of Bohemia who kept up a correspondence with Descartes up until the day he died. She questions how an immaterial being that has no extension and is not made of physical matter can affect a purely physical body. She quite justly and correctly makes the observation that for the ‘spirits’ of the mind to pass information to the body they would have to be intelligent, yet these intelligent ‘spirits’ are somehow affected by hormones, drugs and alcohol etc. This raises an issue at the core of the argument; how can the mind, that can allegedly subsist without the body and has nothing at all in common with it, be ruled by it?

Descartes cannot ignore the flaw so again reaches into his metaphysical bag of tricks and attempts to respond by introducing another primitive notion. Having already stated that the primitive notion of the body is extension, and that the primitive notion of the mind is thought, Descartes states that the primitive notion of the unity of the soul and body is their union. On top of this it is interesting to note that even if a non-spatial, non-physical mind with no extension could be located the problem of interaction would still remain. Descartes also has an answer for this; he attempts to convince us that where mechanical laws explain the movements in space as the effects of other movements in space, there must be laws that explain the non-spatial workings of minds as the effects of other non-spatial workings of minds. In other words he is saying that just like there are laws of physics in the physical world, there must be special laws of ‘physics’ that are true for the metaphysical world, however he fails to provide any proof or reasoning for believing these things, hence the problem of interaction is left unanswered.

It is not under debate that Descartes was a brilliant philosopher, however his thoughts seem to be somewhat clouded by his religious beliefs and as such, despite all available knowledge, seems unwilling to let go of the idea that our minds are likely purely physical and cannot survive past the death of the body.

A fine example of his inability to accept this idea is the argument from clear and distinct ideas which is an amazingly weaker argument from the beginning. The background to why he believes in the validity of his clear and distinct ideas can be found in his Third Meditation, in which he finds the idea of a supremely perfect being, and concludes that this thought must have been placed there as an imperfect mind could not possibly have conceived it.

Descartes believes that everything he clearly and distinctly understands is capable of being created by God, and that clear and distinct ideas are true because God exists. This is known as the Cartesian circle but stretches from this argument into the indubitability argument as Descartes uses it to ascertain that the mind and body are separate. “Finally, so far as my parents [from whom it appears I have sprung] are concerned, although all that I have ever been able to believe of them were true, that does not make it follow that it is they who conserve me, nor are they even the authors of my being in any sense, in so far as I am a thinking being; since what they did was merely implant certain dispositions in that matter in which the self, i.e. the mind, which alone I at present identify with myself, is by me deemed to exist. And thus there can be no difficulty in their regard, but we must of necessity conclude from the fact alone that I exist, or that the idea of a Being supremely perfect, that is of God, is in me, that the proof of God’s existence is grounded on the highest evidence.” (Rene Descartes, Meditations On First Philosophy, He comes to the conclusion that because he can clearly and distinctly understand his mind apart from his body that the two are distinct from one another in that he has a clear and distinct idea of himself as a thinking, non-extended thing and that he has a clear and distinct idea of his body as a non-thinking, extended thing. He accordingly concludes that his mind is therefore completely separate from the body, and can exist without it.

It is impractical to accept this based on the given argument as it is quite obviously based on unsound reasoning. The idea that simply because I can form a conception of myself existing as a thinking thing alone while at the same time doubting that anything physical exists seems to be lacking and easily dismissed. As Wittgenstein states, propositions such as those made by Descartes cannot be used to express metaphysical truths, they are simply a special part of our language game. Ultimately Descartes, although having a scientific background, really begins to let his religious beliefs and his belief of the afterlife interfere quite badly with his philosophical findings and theories. Aristotle states that “entities should not be multiplied beyond necessity”, in other words, just because we are unable to explain the mind as a purely physical thing, is no reason to jump to conclusions and begin inventing metaphysical theories and beliefs, against all logic. This is also known as Ockham’s razor, or the principle of parsimony, and is the common sense approach to things.

While science has only just begun to scratch the surface of the brain, it is likely in years to come, with advances in technology, that we will see the mystery of the brain unravel before us, and unfortunately, for all of us out there that believe in an afterlife or reincarnation, our hopes of ‘living’ in heaven forever or coming back to earth in a different form will dissipate with dualisms demise.

Opponents questioned the portability and reliability of such external artifacts, and the fundamental differences between internal and external goings-on. They also raised concerns around the fact that the brain has total choice and control and that cognitive load is increased when one uses external resources.

The Clark and Chalmers hypothesis, and what is commonly referred to as the Parity Principle, states that if in performing a certain task, a part of the external world functions as a process that, were it done in the head, we would have no hesitation in considering as part of the cognitive process, then that external artifact is part of the cognitive process.

All of the hypothetical discussions involved in the initial Clark & Chalmers paper are centred on Inga and Otto. Inga relies solely on biological memory while Otto, who has Alzheimer’s disease, relies heavily on his notebook, that he carries everywhere, for information. Otto is considered coupled with his notebook as the external resource plays a pivotal role in his cognitive processes. Coupling is described as a reciprocal, causal link between an agent and an external resource in that an agent rarely makes any decisions without using said resource. Thus the artifact is not at the end of a long causal chain but has direct impact on the agents’ cognitive processes in such a way that, if decoupled, behavioural competence drops as if a part of the biological brain were removed.

The four requirements of external artifacts for them to be considered part of an agent’s cognitive system are that the artifact must be constantly available when it is needed, and must be available without difficulty. In the case of Otto this means that his notebook must be available when he needs to look something up, like an address or phone number, and that he can easily access it i.e. it is in his pocket and not at home in his bedside table. The other two requirements are that the resource must be automatically endorsed by the agent and any information used from the resource must have been previously endorsed. Again in the case of Otto this means that any information he finds in his notebook he accepts as true and that he has previously confirmed the information. Although this fourth requirement is debatable, as beliefs may be acquired through other means, the first three are essential to the extended mind hypothesis as we shall see later in this paper.

An important aspect of a coupled cognitive system is that in using external resources we commit epistemic actions. Such an action is conducted to help determine the procedure to be executed, such as physically rotating a shape to see if it will fit in a given hole, and help with the decision being made. Pragmatic actions on the other hand are executed as a result of internal cognitive processes having made a decision already and have no influence on said decision.

One of the simplest ways detractors attempt to disprove the extended mind hypothesis is by raising objections to an external resources portability, but such objections have a specious nature. External resources can be considered part of an extended mind if they are there when needed and without difficulty, there is nothing that states that occasional decoupling cannot occur. It is unlikely that when we are asleep or in an unconscious state that our internal resources are considered any less a part of our cognitive system than if we were in a conscious state so it seems foolish to treat external artifacts this way.

Upon accepting the idea that we do extend our mind when coupled with external resources, and that portability is not actually a problem we encounter a much more obstinate dilemma; what is to stop an outside agent from altering or even stealing our external resource? This is quite patently a very real quandary that makes the use of external resources fundamentally different to those of internal ones and raises a number of questions that could derail the extended mind hypothesis.

The foremost problem is that of actual altering or theft of such resources. Obviously in the case of Otto and his notebook the chance of his artifact being sabotaged or stolen is quite slim. It seems unlikely that anybody would go out of their way to attempt to steal it, and it seems even more unlikely that given the opportunity an outside agent would be able to alter the information in it without Otto realizing it had been altered. But in the case of mobile phones, laptop computers and other more technologically advanced external artifacts these problems become more apparent and harder to detect.

However this will not prove the extended mind hypothesis to be false, it is just more likely to limit the types of resources we can consider a part of our extended cognitive system. Security of such external resources seems to be the major concern here, in that if we can make an external resource as secure as our own biological ones then surely these can count as part of our cognitive system. Software security on technology such as password protection is obviously the first step but physical security needs to be considered also. With such items getting smaller by the day theft is becoming easier, as such an intelligent thief can easily acquire possession of your external resource and have it sold and shipped to the other side of the world using an online auction facility such as eBay within a matter of days.

Nonetheless, since the dawn of human intelligence we have encountered different types of threats in our social domain, and there is evidence to support the idea that evolution has made cheat-detection or the deployment of social guards central to our success as a species. Due to the deployment of such social guards, it is stated that our cognitive load is increased in such a way that it makes our external resources development and functional poise radically different to that of our internal states or information. The conclusion drawn from this is that by increasing cognitive load, our attention to the external must increase subsequently endangering automatic endorsability.

Not to worry! This idea of Sterelny’s seems to help the hypothesis as much as it detracts from it in that he states evolution has made the deployment of social guards central. One may tend to concede the point of increased cognitive load if the mind worked as a modularized system but it is more likely that it works as a highly distributed connectionist net (Footnote #1). Such a network is comprised of thousands of nodes across a number of separate layers, and because of these nets distributed nature, social guards deployed in this type of economy are not only naturally deployed, but spread across the entire net.

Hence the social guard part of our cognitive economy partakes in any processing that involves social exchange thus cognitive load is not increased and Parity prevails. It seems that social guards are not the only thing that evolution has made a central part of our cognitive processing and that denying such external artifacts would seem to deny a lot of our evolutionary history, not just the feasibility of the extended mind. Our brains have evolved in such a way as to take advantage of such resources, including the pliable, external environment, for example our visual processing system takes advantage of the structure of natural scenes as does it use bodily motion and locomotion to take cognitive shortcuts.

We can begin to see how cognitive niche construction plays a major role in our cognitive economy. Niche construction is defined as the activities, choices and metabolic processes of an organism through which they define, choose, modify and create their own niches. This refers not only to their choice of mate, habitats and resources but the construction of things in their local environment such as burrows, hives, nests, damns etc.

We as humans similarly construct cognitive niches, for example, expert bartenders not only line up differently shaped glassware but do so in temporal sequence to the likely drink orders coming in, hence the problem of remembering what drink to prepare next is transformed from an internal memory process to simply associating a drink with a certain glass shape. It is now clear how other items such as Otto’s notebook can be considered a cognitive niche, and that such niche construction appears to be linked quite causally to our evolutionary history, without which our competence as cognitive agents would drop substantially.

It is also possible to consider language as a niche constructed by human agents. We use internal language in a range of ways, most commonly to motivate ourselves or help revise complicated processes until such time as we complete the given task. Language, whether internal or external, is a form of scaffolding used to help us either remember things (such is the case with road signs) or in an example proposed by Daniel Dennett help a chess player remember not to get their Queen out early (by internally revising what not to do whenever said player reaches for the Queen early on in the game, Footnote #2).

A problem raised by Adam & Aizawa is the idea that because of the fundamental differences between internal and external goings-on in the way that they causally drive action that there can be no unified science of the extended mind. To suggest that external and internal resources must act in the same way for them to fall within the one scientific umbrella is too presume too much, they simply need to behave and follow the same set of laws. Evolution has made us capable of using the plethora of external resources available; this does not mean they have to be fundamentally the same. It could quite easily be entertained that somewhere in the future we come to the conclusion that intelligent behaviour or cognitive processing is simply the point at which a number of illuminating paradigms coalesce into one.

Along a similar line of thought, who is to say that the way in which our internal cognitive processes operate will be able to form a unified science? It is likely as we continue to learn more about the inner workings of the biological brain that somewhere down the line we discover that there are fundamental differences between abilities such as the way we do mathematics in our head versus the way our brains interpret signals of pain from our extremities. It is beginning to appear that unification is not an overly important consideration when it comes to the extended mind hypothesis.

To take a slightly different line of thought again about the fundamental differences between internal and external resources it has been claimed that Otto’s notebook represents an outdated theory of how biological memory works and therefore any text-based storage having role-parity with that of biological memory must be false. But there is, however slight, a definitional oversight to this claim; Clark & Chalmers were never attempting to claim that the notebook alone was a cognitive system simply that it plays a crucial role within Otto’s cognitive economy and hence does not need to have role-parity.

The idea that Otto’s notebook only supports derived content while inner symbols in a cognitive agent are considered intrinsic is another objection to the extended mind. They state: “The representational capacity of orthography is in this way derived from the representational capacities of cognitive agents.” and subsequently jump to the conclusion that the extended mind hypothesis must be false as the two types of content are fundamentally different from one another. But the problem here seems to be not the underlying facts of derived versus intrinsic content but why do we need to assume that an agents cognitive processing must be comprised solely of intrinsically content-bearing representations. Yes we can accept that these two types of content are different but to exclude external resources solely on the fact that they can contain only derived content seems outrageous, the relevance to the extended mind hypothesis is not an items’ content but the role it plays within the cognitive process.

The issue of choice and overall control was brought to light by Keith Butler who states that because the brain is the ultimate decision maker in a way that the external resources are not, the external resources cannot truly be considered part of an agent’s cognitive system, although they may occasionally participate in cognitive processing. This is an interesting thought but one that ultimately fails. To eliminate that which is external from the cognitive system based solely on this assumption leads us down a path towards eliminating a lot of the internal as well. With the current information we have regarding the mind it would be impossible to pinpoint the seed of consciousness where decisive control and choice actually occurs. One can only speculate as to how much of the biological brain is needed, and where in the brain these decisions occur. A single neuron obviously isn’t enough, but what of fifty neurons, or one thousand neurons? It doesn’t make sense to eliminate the external on the basis that it does not have the final say.

External resources versus internal ones leads us to another predicament, that because we have to perceive external information rather than introspect, this provides sufficient disanalogy as to exclude it as part of the cognitive. Closely shadowing this along the lines of perception, is the dilemma that mistakes could be made when an agent is attempting to access the information or external resource, but neither of these arguments have any force behind them.

Firstly, to exclude the external based on the way in which we relate to or perceive it depends entirely on our definition of the cognitive system. Clark & Chalmers definition treats the external artifacts as if they were internal, hence this objection fails on principle alone. However the problem with an error in accessing the information is genuine, for example, Otto could misread his notebook. Even when considering a case such as this it is obvious that our internal information can encounter the same sort of problems. If an agent is intoxicated or is highly emotional such as those under the influence of drugs/alcohol or people who are highly depressed/angry, then the chances of misreading information even from our own minds becomes possible. We also have the problem of false beliefs from simply forgetting things; an agent may believe they left their keys on the coffee table in the lounge room when they actually left them in their bedside table. These events show simply that an error has occurred during the retrieval of the data, not that the error is an error in perception.

In closing, it doesn’t seem that those who protest against the extended mind hypothesis have been exceedingly auspicious in proving their case beyond a reasonable doubt. Their demurrals, although some better than others, still fail. Minor problems and definitional oversights aside, it is unlikely that the use of external artifacts in a coupled cognitive system as laid out by the extended mind hypothesis can be shown to be incorrect. By evolutionary means alone the use of external resources seems to play a pivotal role in our cognitive economy.

Also the way we as human cognitive agents, especially in expert fields, use niche construction to help simplify cognitively demanding and complex situations, let alone how we all use language for similar means seem to permanently couple such external resources with our internal goings-on. There is likely to be a large amount of debate still to come on this topic, and a lot of issues that have not been covered in this paper but however one looks at the situation, it appears the extended mind hypothesis holds steady in the face of adversary, and it appears unlikely that it will be rocked at its foundations any time in the near future.

1. This is assuming that we are not symbolic engines and that we do operate more like a highly distributed connectionist net than a modularized economy. But in the event we do operate in such a way it is likely that social guards can still avoid cognitive bloat by off-loading into the environment.
2. There are a number of issues that we could discuss here but not all are relevant.

The main problems that the classical CTM encounters are Searle’s Chinese Room argument, the problem of homunculus, how there seems to be syntax without any semantics, the frame problem of relevance and updating and the cost of representational atomism.

Probably one of the more famous arguments against the classical CTM is that of the Chinese Room argument conducted by John R. Searle. The overall point of this argument is that we have the syntax in place but it is lacking semantics. A computer can use rules to interpret symbols but this does not mean it understands what is going on. To put it another way, semantics is not intrinsic to syntax, just because a computer can understand a set of predefined rules or algorithms doesn’t mean it actually understands what its doing.

Another more recent argument, again by Searle, that by his own confession he should have seen ten years ago is that syntax seems to be imposed on the system by its observers. The fact that we can make a machine out of just about anything and have it perform computation means that the processes are not intrinsic to the system but are completely abstract.

This spawns the problem of homunculus in that if syntax is not intrinsic to the system, and computation is defined syntactically, then nothing can intrinsically be a digital computer based exclusively on its corporeal properties. The computational theory of mind, always in some way commits the homunculus fallacy wherein they treat the mind as a little man inside the brain, and this little man uses the brain to execute his computations. However proponents of the CTM make a reasonable attempt at eradicating such a problem by supposing recursive decomposition. This states that there are multiple levels of homunculus within any system such that at the lowest level there is a ‘stupid’ homunculus doing simple yes-no, 1-0 type computations. Regrettably though the problem still remains in that the syntax is not intrinsic to the system, and we must therefore have a homunculus that stands outside the system to provide the system with its operating syntax.

There are a number of other problems that the classical computational theory of mind encounters which we have not had enough time to examine in depth here. The cost of representational atomism, together with the frame problem of relevance and updating, but also the fact that syntax has no causal powers other than those of the implementing medium all need to be explained and overcome. But that isn’t to say that the brain isn’t a digital computer, on the contrary, some other form of the computational theory of mind might contain a better explanation of how things work.

Stay tuned tomorrow for a short paper on the Connectionist Computational Theory of Mind…

The main dilemma facing the mind-body problem is that of interaction, and how a nonmaterial entity could possibly interact causally with a physical object. There are other dilemmas too; like the overwhelming evidence that the mind is the activity of the brain and how drugs and alcohol could affect and infringe on a non-physical thing. Subsequently the argument behind this particular theory of mind is fragile at best and we must contemplate alternatives that have stronger premises and make more sense.

Around the middle of last century a mathematician named Alan Turing helped break the Nazi enigma code in WWII, and more to the point, designed a hypothetical machine that used symbols and a set of predefined procedures (i.e. an algorithm) that when followed would produce the solution to a given problem in a finite amount of time. This machine was eventually called a Turing machine, and a thesis known as The Church-Turing thesis emerged alongside it.

A Turing machine consists of only a few elements; a read-write head that can read and write symbols, a tape that is divided into unique cells of a predetermined size, a pointer, and a set of mechanical reflexes. Without going into specifics, it was discovered that one could create a universal Turing machine that was able to impersonate the actions of any other Turing machine simply by feeding a description of said Turing machine into it. In this way, Alonzo Church, was led to conclude that any well defined set of instructions or steps that was guaranteed to deliver a solution could be implemented on a Turing machine.

But what does all this have to do with our theory of how the mind works? A Turing machine, although a hypothetical device, showed us how our mind may work and led Pinker to propose the idea of the language of thought, mentalese, and how this is an integrated and integral part of the computational theory of mind. He systematically shows and explains how the popular misconception that we think in our native tongue is wrong, and how a richer language is at work in our minds.

A good analogy for those who know how a computer and its software work would be to consider mentalese as the programming language. Within this language we have representations that are sets of symbols and correspond to aspects of the world, while the homunculus could be thought of as sub-programs or modules within our software. These modules are only required to respond to a small number of symbols, and subsequently the system becomes a lot simpler. As Pinker says “The intelligence of the system emerges from the activities of the not-so-intelligent mechanical demons inside it.”

The computational theory of mind thus has two key elements, our mental representations and the processes or programs (sometimes called demons) that access them. As Pinker states, any complete theory of psychology needs to take these into consideration and explain their use. If a task is considered more difficult than another is it likely that it uses a lot more symbols or simply requires a longer chain of demons to complete it. The way in which we generalize from experience, or react to different situations that may be similar to others is indicative of our use of representations.

You be the judge… Over the coming months though we hope to examine more about life inside of prison and come to some scientifically based conclusions about the jail and justice systems. We are also hoping to examine the positive and negative effects of life on the inside in an attempt to identify changes that should be made in regards to inmate treatment and rehabilitation.