Human cognition is the study on how the brain of a human being performs its function of thinking,where the practitioners try to design models in trying to explain on how the human beings thinking works. They develop models on explaining the way human beings think and they normally evaluate this models on the basis of checking their accuracy on explaining human thinking through some kind of comparison. The models designed are subject to change with time in trying to explain human thinking,on the reason of making them more accurate in explaining how human beings think,therefore a cognitive model is an approximation of the real thinking process of a human being.
These Human cognition models have been used in developing sophisticated machines in the name of computers which are designed on trying to imitate the human thinking. These machines are made on imposing artificial intelligence on them, in enabling them to work as human beings intelligence wise,where the human beings brain can be equated to information processing system in the computer. Computers therefore takes a better position in evaluating the human cognitive model, because this is more practical on knowing their ability on explaining the human thinking. If by any case a model fails to be compatible with the computer applications then that model ignored or further refined if almost compatible with the working of a computer in the process of comparing it to the process of human thinking(Harpaz 2007).
On using computers as a means of evaluating human cognition models,there are some assumptions that have been overlooked and which seem to be quite importance in determining the accuracy of these human cognitive models, which need to be factored in when using them as a means of evaluating these human cognition, so as to have an efficient form of evaluating these human cognition models and they are discussed as given below;
Computer memory is quite different from the memory of the human brain because a computer only loads and stores on informations in and from the its memory compared to the human brain memory which needs making an observation or hearing before an information if registered in the brain memory. These makes it impractical in comparing the human brain memory with the memory of the computer, because the loaded information into the computer is systematic and predetermined before registered compared to the human brain memory that receives information in impulses. Storing of information in a computer memory is quite mechanical in the sense that it will reproduce exactly of what was store without altering anything . whereas in the human mind there are possibilities of altering the stored information. The human mind is likely to make an interpretation correction unlike a computer that needs the machine operator to make the correction on giving a command, that is it can not make a correction unless further programmed to. A computer memory is also different in the sense that it’s logic and arithmetic operations on certain values are always contained in the register compared to the human brain memory which have got no these registers .
The use of computer in evaluating the cognitive models is so crucial,in the sense that it gives a chance of correcting the overlooked assumptions in designing the cognitive model in the process of explaining the human thing, because a model is likely not to work if some very important assumptions have been either omitted or presented wrongly, and this removes the possibility of of under specifying a model in working to limited scope of assumptions. According to the nature of these human cognition of it being complex, the models that are designed are also necessarily complex on trying to capture the real know how on how human thinking works, which exposes the models to risk underspecification, and the effort of trying to remove these under specifications is necessary so as to obtain an accurate human cognitive model that exactly represents the human thinking.
There runs a risk of not getting an accurate results that exactly explain the how human brain real thinks if the system that is used to evaluate the model is build on the same principals as of the model that is being put on test. The results are likely to be positive on the sense that their principals are compatible and thus not real explaining the the major purpose of the model, but rather just going round some assumptions. This situation might lead to accepting some hypothesis as a law,whereas it does not representative enough over the real situation which was being explained by the model ( http://human-brain.org/human-brain-index.html ). Therefore, evaluating of human cognitive models using computers might not be convincing because a model might be designed on the same principals like that of the evaluating system,which makes them compatible that may lead to accepting an hypothesis as being true , whereas in the real sense its not the case on the ground. The principals on which these models are designed should be scrutinized and compared to the principals on which the system that is evaluating the models is based. Therefore, misconceptions during modeling should be dealt with if the better and accurate results are to be expected. The identification of these failures that may arise from the issue of shared principals and making corrections on them are difficult because the models are based on the same principals like that of the computer,thus it’s easy to manipulate either of the two in getting a given result that might exactly resemble like that of the human brain thinking, but based on wrong principles or assumptions. This limitation can only be ignored if the number of variables that are used in the model is small and characterized by a small range which is not the case with the models of human cognition that involve a large number of variables, thus not real practical in ignoring this limitation if an accurate result is expected. The limitation can also be ignored, that it may have a negligible impact if the size of the model is small in itself, but in the case of human cognition the models are always big due to the nature and the complexities that are contained in the phenomena they are trying to explain ,thus impossible to ignore this limitation on considering the kind of complexities involved in human thinking.
The human brain is relatively small and as such models that are explaining the human thinking ate expected to be relatively small equivalent to the brain itself. This is contrary to the expectations, because the models are usually big in size as compared to the human brain., thus neglecting a very important assumption on designing these models. This points out a limitation in modeling on trying to explain human cognition, as the model doesn’t reflect the real size of the brain,thus providing a possibility of either adding extra components or less. Size itself has an impact on how a system works,thus ignoring size variations is likely to distort the conclusion for any test of model carried on a computer. If this constraint is not factored in then there is a possibility of replicating behaviors in the model, on the condition that if the model is big in size then their occurs a possibility of the model replicating some observations which makes it wrong, and thus unable to predict accurately. The nature of the human brain and that of the model system intelligence normally take an infinite time range, that makes it necessary to include more constraints on the model space , if one was to consider the brain alone. The infinite nature both the brain and the computer intelligence system makes it impossible in drawing stud lines in doing major tests on the model being evaluated on its accuracy in explaining human cognition.
The computer memory’s characteristics are not the same as the characteristics of the brain neurons in the sense that in the computer memory, there is an arbitrary specified location for one to access an information, and the information can be removed from the location without affecting the location that was specified,where also the no classifications on definite locations and this is because the computer normally stores information in its memory using the real addresses. Therefore, any device that has the ability of being used a a computer memory then it should be storing values in the form of real addresses, in that on it given given the real address it can produce the value, and this is quite independent on the the nature of material being used a the computer memory in as much as it has the above properties. Given the nature of the human brain, it makes it impossible to apply the real addresses in the brain despite the limited information that is available in explaining the nature of the neurons. From this comparison, their happen to be a great difference between the computer memory and the human brain memory in the sense that the human brain neurons does not accommodate the use of real addresses, thus a weakness in using the computers as a means of evaluating the models of human cognition.
From an angle of trying to understand on what might be working as real addresses of the computer memory, their can happen to be a combinations of neurons,neuronal activity patterns,synapses and possibly diffusible signals, where diffusible signals are not applicable in the case of computer memory. The computer memory normally operates on a given time scale of thinking,which is contrary to the human thinking as the neurons don’t move on any time scale on thinking in enabling the neuronal patterns. The neuronal patterns don’t normally move toward a specified direction, because they are lowly connected, and as such the information transformation movement is not predetermined, and which is not like with the computer which has a predetermined path of information transformation. These arguments weakens the practice of using computers as a way of evaluating the models of human cognition, on th basis of equating the computer memory to the human memory because the human memory is incompatible with the use use of real addresses, and this falsifies the argument of researchers that all the programs that run on computers can run on the human brain. Therefore,models that are run on computers should be interpreted according to the principals of a computer, as they normally depend on the computer memory, which may make them wrong.
Information that has been coded in the brain is normally coded within the neurons and its attribute are mixed with other attributes in that it is not possible to transfer the information may be to another brain or a different location in the brain. This is quite different from the computer in which the information that is stored normally has a real address of obtaining it and the information can transferable to any other location within the computer memory. This difference between the computer and the brain way of coding draws a line makes it unrealistic in using the computers as a way of evaluating the models of human cognition because information in a brain can not be transferable ,whereas the information in a computer can be transferable into a different location within the computer(http://human-brain.org/cognition-computer-models.html ).
The way in which the brain functions is complex in that it normally involves billions of neurons even in a simple activity, in contrast to the models of cognition which operate within a limited scope,because of its simplicity. This is a kind of a deductive reasoning in which an observation of a few cases,is certain of giving a conclusion over the whole process, that has limitations in its self. It has a limitation on the ground that, it can happen that the few cases observed are not representative of the whole process thus making a generalizations that is likely to be false on considering the whole process, as it may only hold within the few cases considered. This raises doubt on whether a simple model can be used to present a complex system. The validation of simplicity can only be achieved on holding the assumptions that the complex system just operates in a simple manner and in which, it’s constituted of a small number of simple ways of doing the operations.
The brain is quite difference from the computer in the sense that the brain is self adaptive, whereas a computer need to be supplies with an external information for it to behave in some course. This makes it unrealistic on using computers as a basis for evaluating models of human cognition because a brain is normally dynamic and adaptive within itself , therefore responses may vary from time to time within the brain compared with the computers which are standard until given an external guidance, on how to adapt to new way of operating.
Conclusion. Models of Human cognition are models that have been designed to explain the human thinking process, where the accuracy of these models is being tested by running these models on computers. It’s taken for granted by the scientists involved in this exercise that if a models runs nicely in a computer then it accurately explains the thinking of human beings, on which is limited to many reasons that makes it difficult in making right collusions about human thinking. The limitations of using computers as a way of evaluating models of human cognition include the issue of similar principals in both designing the models and the computers themselves among other reasons.
Harpaz Y.,2007, Human Cognition in the Human Brains, Retrieved on 14th May 2008 From the World Wide web; http://human-brain.org/human-brain-index.html and