Topic letter 11

Models in our brain

1. Introduction

Argyris (1990) and Senge (1992) introduced the concept of mental models in the literature of organizational development. These models were seen as having an huge impact on behavior. Changing these models was and is regarded as a key element in organizational development. Piaget (1924) introduced the principle of models in our brain almost a century ago in an attempt to describe the development of the thinking of children in their early life. Craik (1943) later introduced the concept internal model as a way of describing how people understand the world around them and use the model to do mental experiments. This article elaborates further on this subject, using recent knowledge from neurobiology, in order to get a better understanding of what happens in our mind.

2. The perception process

You walk on the street and you meet somebody. Your retina makes a registration of what happens. Hundred million receptors in each eye are activated by the light. A million special cells, called neurons, collect the information of the receptors. Neurons have a long tail and can send an electrical pulse through the tail into the brain. Some neurons report colours, others light intensity. In the brain other neurons send this information to the back of our head, to the visual cortex. There starts a highly organized process of data analysis and transformation. The result of the first step is an exact registration of what the retina has detected. For the second step, the output travels to a new layer, where neurons compare information from groups of neurons in the first step. A neuron in this layer can analyse that the neurons left and right of a certain point report a different colour. A border is detected. This neuron asks other neurons whether they also see an border. A group of “border neurons” together can detect a line. This observation “line” is the first interpretation and building stone of a visual construction. In the next steps, lines are connected to squares. The combination of squares, together with shades, creates a figure. This process is usually referred to as sensory coding, adding codes to perception. At this moment, at least a hundred million neural actions have taken place in the brain. The result is an internal image of what might be happening outside the body. This image is one of the most simple constructions of a model. This model however has no meaning yet. Therefore the visual cortex sends the model to several other parts of the brain. In one part, the model is related to a lot of other models. In an associative process, comparison indicates that this model is a face. In another part of the brain, in the database of faces, the model is compared with stored faces, to see whether this one is familiar. A few neurons who have been instructed to react when this familiar face is seen, become active. Within two tenths of a second, from the moment that the light stimulated the retina, we recognize this person. During this process, the brain used a half billion neural actions to help us realize that we know this person. It is an old friend and we already experience a warm emotion. We feel glad and want to shake hands.

In the mean time, while neurons are sending the information from the retina to the visual cortex, other neurons are listening and also started to send information to other parts of the brain. These areas don´t want to wait until the visual cortex is ready and produces a sharp image. They want to do a first analysis themselves. Neurobiologists estimate that there are at least 10 areas that do an independent research. These parallel processes help the brain to prepare emotional and behavioural responses, just in case a sudden action is needed. There are indications that there are cases in which the emotional reaction is even earlier than the moment that we know what we see. Looking at this from a perspective of survival we can understand why, because if we are walking in the wood, anticipating on a possible snake is more important than checking whether we see a stick on the ground.  

3. Perceptual model

The above example of meeting someone, makes clear that for a “simple” observation of a person, an enormous amount of brain activity is needed. Our brain is really specialised in recognizing faces, but we still need half a billion actions to do so. Our brain has an enormous capacity, but it is not workable and also senseless to keep on using all information. Therefore, the brain reduces the amount of  information while identifying the face. During this compression, the most relevant building blocks are kept. Of course the brain needs some expertise of what is relevant and what not. This knowledge has been gathered during the years and is used as filters during the compression phase. The result of compression has the same value and status as the original information. Each next step, like “who is this person?”, “when have I seen him before?”, “is there still something I like to tell him?”, etc, is done on the basis of compressed information. After compression the original input is gone and the compressed model takes its place. The neural routes are now ready for new input. The facts are forgotten, the interpretation of the facts is the starting point for future actions in the brain. The resultant, the compressed model after sensory coding, is defined here as a perceptual model.

A perceptual model is an internal model of the actual outside world

A perceptual model is an internal model of our present environment, made by compression of perceived facts from our sensory channels. Perceptual models are the starting point for other activities in our brain. In case of a missing relevant detail in this perceptual model, the brain can ask the compressor to adjust subsequent images by changing the compression filters. Perceptual models therefore are not static. They can develop themselves due to interaction with other sorts of models in our brain. This creates the possibility to learn how to see, to hear, to smell, to taste and to touch. If we realise that we can do this, it is also inevitable to realise that everybody has his own learning history and his personal way of perceiving reality. When two people watch the same face, they will see something different. Despite all this, the way we perceive facts is our reality. We believe in it so strong, that we most of the time forget that it is only our reality and not the reality. Unfortunately, nobody can perceive the reality, because everybody has to compress in order to perceive. There is innumerable research that shows us that we are fooling ourselves a little while perceiving the outside world. On the other side, we should not exaggerate this point. The way we perceive reality probably has a strong overlap. If we were not able to make a good estimation of the traffic on a busy square, we would never reach the other side of the road. In a split second we can estimate how many vehicles there are and with which speed in which direction they are moving. The real differences between people come later in the process.

Each sensory process has it´s own specific area in out brain. The visual information is by far the most important and is processed in the visual cortex in the back of our brain. Information connected to hearing and touching is processed in the temporal lobe, just behind our ears. Al these parts belong to the neo-cortex. The smell has a different position. The information goes first to the limbic system, the basis of our emotions, and then continues to the neo-cortex. A smell therefore immediately can evoke emotions, memories and even behaviour (Vroon 1994). The processing of the information of each of the senses is done autonomously and separated. This has been discovered by exposing people to conflicting sets of information. For this experiment a small film is made of a mouth saying “ga ga” while at the same time the sound is manipulated so that you hear “ba ba”. The process center has to digest two conflicting perceptual models. Our brain is really sensitive for this kind of conflicts, because it is an indication that the world is not perceived correctly. In terms of survival this can be dangerous. In order to solve this conflict, the brain can either delete parts of the information or transform it. In this case, both the perceptual models of hearing and seeing are transformed into a new model. We hear “da da”. Closing our eyes is also a way of reducing the conflict. Then we immediately hear the original “ba ba”. If we open our eyes again, we again perceive the constructed “da da”. This phenomena is named after the discoverer, McGurk (1976) and is called the Mcgurk-effect (you can watch it on You-Tube). The process of transforming the sound can´t be stopped, even if we know that we are being tricked. This experiment shows that we hardly have any conscious influence on our perceptual processes. Besides that, everybody make the same mistake here. This shows that the perception is programmed, inclusive the errors in the system.

4. Emotional models

The relative uniformity of perception reduces as soon as the brain starts giving meaning to the perceptual model. This meaning is given parallel in many places in the brain, which can be clustered in two quite different areas. The one is the neo-cortex (literally “new lamb”), on the top and outside of our brain, and the other is the limbic system, deeper inside our head. If we look  at it from an evolutionary perspective, the limbic system is the oldest. All mammals have these kind of brains. In this part, we define how we feel related to the other model. The emotional model is defined here as a representation of how we feel affected by other (perceptual or mental) models. Usually we don´t recognise emotional models as part of our consciousness, we see it as part of the model. So we say to ourselves “this fact is not reliable” or “ this person makes me happy”. We tend to see our emotional model as an attribute of an object or person. In fact every emotional model, every reaction we perceive is a part of ourselves. In our way of perceiving facts, we add something to these facts, we dress them with our own attributions. In the literature on perception this is called the attribution process. Facts themselves don´t have any emotional meaning, we give them meaning. Due to this attribution process, we can feel angriness while perceiving an event or can fall in love with a person.

An emotional model shows how we feel related to another model

Emotional models are closely involved in planning, in deciding what we want, in experiencing what we like and dislike, in sensing how we want to develop us and even in deciding with whom we like to reproduce us. A special attribute of emotional models is that they are living in an area of our brain in which there is no logic nor language. This causes that we hardly know why we prefer one above the other and that we even can have contradictory wishes. Why did I choose for this partner or why do I like wholemeal bread. If asked, we will give an answer, but in general it isn´t more than a justification afterwards. If we want to talk about an emotion, we first add a mental model to an emotional model. The attribution process is very personal and connected to our genetic codes, our individual learning history and our experience. Emotional models are the most personal models we have.

5. Mental models

Next to the trail from perceptual model to emotional model, there is a second trail in the direction of the neo-cortex. This is the domain of the logic. Here we try to understand the perceptual models, trying to understand what we have perceived. Parts of the neo-cortex are specialised in explaining the what, where and how of other models, including the emotional models. Mental models give an explanation of how we perceive or experience things. A mental model can be defined as a model that explains what and why perceptual and emotional models are the way they are. A mental model is a result of a process of creating meaning. During this process the models are related to other models which are stored in networks in our brain. Each group can be seen as a database of a special item. In this way, we can process the different features of a face and see whether we can distinguish the gender, the age, the colour of the skin, the nationality and even the question if we have seen this person before. Distinguishing qualities is an associative process based on finding the same content or similar features. Mental models enrich perceptual models with an identity. A face becomes an old neighbour.

A mental model gives meaning to other models and explains them

Similar to this process we can see that mental models create a justification or an explanation for emotional models of which we become aware. The old neighbour (mental model) is associated with a nice person (emotional model) and if asked why that person is nice, we can add several other mental models explaining why we experience him as nice.
Mental models can also be related to other mental models. In the associative network we find that the old neighbor was a musician, always working in the weekends. It is weekend, he is wearing a guitar case, and I see a music hall in front of me. These are all mental models connected to perceptual models, which we define her as first level mental models. I can imagine that my neighbor is still making music and that he going to perform soon in the music hall. In this reasoning, several mental models together create an image of what is happening at the moment. This is a new mental model on a higher level.

Level 5 mental model	Distinguishes a value behind a series of level 4 mental models
Level 4 mental model	Creates a conviction or a theory about a series of level 3 mental models
Level 3 mental model	Creates a generalization of a series of level 2 mental models
Level 2 mental model	Gives an interpretation to a series of level 1 mental models
Level 1 mental model	Gives meaning to a perceptual model
Perceptual model	Internal representation of the external world

The first level is the direct observation, the second level is an interpretation or construction of a set of first level mental models. The step from first level to second level mental models involves a process of abstraction. The facts are left behind and more abstract ideas are created. This process does not stop at the second level. A third level mental model can be created out of several interpretations which lead to a generalization. A fourth level of mental models can match generalizations and define a conviction. A fifth level of mental models defines the commonality between convictions and defines a value.

This process can go on and on as long as we have enough free work space in our mind. While we rise in the levels of the mental models, we need to combine and compare more and more underlying models. This reasoning is done in the frontal lobe, just above the eyes. The more complex the task is that we do, the larger is the active area in the brain. Neurobiologists can actually measure this, when they scan people with the fMRI while doing complex tasks. Mental intelligence, usually expressed in terms of IQ, is directly related to the available work space in this area. The higher the level of a mental model we can reach, the more broad is our view, the more intelligent tasks we can do. To some extent this can be trained.

It will be clear that if we look at the differences between people, the lower the level of the mental model is, the more related that these models are to perceptual models and the more common the models will be between people. On the other side, the higher the level is, the more personal connections and associations are involved in constructing a mental model, the more differences we can see between people. If we look e.g. at politicians, we can distinguish completely opposite views when they want to solve a problem within the community. The same facts can lead to convictions and solutions that can be seen as typically left wing or right wing on the political spectrum. Both parties are convinced that there view on the reality is the correct one, and they actually can’t understand why the other party is not sharing their view.

6. The relationships between the models

When we examine our brain, we can see that the structure in the different areas varies a lot. In the neo cortex, the logical part of the brain, the structure is very organized. If we follow the path of the neural activity, we can see that the neural activity sometimes spreads like a circle in the water, often in a fixed and predefined order over several layers of a function. On the other side, if we look in the emotional brain, the limbic system, it looks like a sticky mess of chaotic unorganized neurons. Between all areas, there are highways of strong connections, which give as a clue which part of the brain is talking to who. All brain functions have a headquarter somewhere and all these headquarters are connected in an associative network. It seems as if all the functions are involved in a multidisciplinary gossip club, where everybody is talking with everybody, although there are preferences, and where information can go parallel from one area to another via several routes.

In the above scheme, all the arrows are possible routes, and in practice all these routes are used in any order. On the left we see the sensory input and the first processing in a few steps. Next we see the compression. Compressing is done by using filters, which can be influenced by emotional and mental models. If there is an actual topic in our mind (“can I trust this situation?”), the compression will take care that all information related to that topic, will be kept as much as possible. After the compression gets the form of a perceptual model. This model is the starting point for further action. Usually the information is sent at the same time both to the mental and to the emotional models. Here we see that the neurobiology has a different view than the popular Rational Effectiveness Training (RET), developed by Ellis and well described by him and Baldon (Baldon, 1998). The RET states that the route is perception -> cognition -> emotion -> behaviour. This is a possible route and therefore the theory is partially useful. But there are also other routes available, like perception -> emotion -> behaviour or a route like perception -> emotion -> cognition -> behaviour, or even more complex routes like perception -> emotion -> cognition -> emotion -> behaviour, etc. There is a lively mutual interaction and influence between mental and emotional models and there probably are many loops. The same old-timer can create excitement by the idea that you own it, a feeling of rest if you see it as an investment or worry if you realize how much petrol that it uses during a ride in the countryside. There is also a direct line from the perceptual models to the automatisms in the small brain, which can create automatic and mostly unconscious behaviour. If you are driving a car you can many miles without even noticing it. This topic is not further elaborated in this article.

7. Consciousness

The brain is very active and has billions of activities each second. The majority of these actions will never be noticed by our consciousness. Our consciousness is simply too small to learn to know all these models. In the scheme we can see that perceptual activity, including the compression, is fully done at an unconscious level. We don´t know that we perceive the world, we just think that the world is there. Apart from the automatisms, we are able to become aware of some of the models. The majority of the used models however will probably always stay unknown to us. There are simple tests, like the Implicit Association Test, who demonstrate us that we even have mental models that we don´t want to have because they are less socially accepted (Dijksterhuis, 2007). Lamme (2003) states that when there is enough interaction between the models, and when more areas of the brain are activated, the chances that we become aware of our models are increasing. If we really are willing to invest in learning to know them, we can put the spotlights on some of the models. The best way to do that is interaction, in a dialogue, with others. We should realize however, that the conscious and self-conscious living person is a myth. We can know quite something about ourselves, but we hardly do so.

8. Epilogue

This is a first article about the differentiation of models in our brain. Maybe this article raises more questions than answers. When we look around us, we can see that people change behaviour, which implicates that they also can change the models which are the creators of this behaviour. In a second article in this series, I will elaborate further on the processes that help models to change and develop themselves.

9. Literature

• Argyris, C. (1990). Overcoming organizational defenses: Facilitating organizational learning. Needham, MA: Allyn & Bacon. ISBN: 978-0-205-12338-4
• Baldon, A. & Ellis, A. (1998) Een ander kijk op problemen. Zaltbommel: Thema Uitgeverij. ISBN: 90-705-1223-8
• Craik, K. (1943). The Nature of explanation. London: Cambridge university press. 1967
• Johnson-Laird, P.N. Mental models in thought.  (2005) In Holyoak, K. and Sternberg, R.J. (Eds.) The Cambridge Handbook of Thinking and Reasoning. London: Cambridge University Press.  Pp. 179-212.
  
• Dijksterhuis, A. (2007). Het slimme onbewuste. Bert Bakker. ISBN 978-90-351-2968-9
  
• Lamme, V. (2003) Why visual attention and awareness are different. Trends in Cognitive Sciences 7:12–18.
  
• McGurk, H. & MacDonald, J. (1976). "Hearing lips and seeing voices", Nature 264, 746-748
  
• Miras, M. (2007). Ben ik dat? Nieuw Amsterdam. ISBN: 978-90-468-0216-8
  
• Nelissen, M. (2008). De brein machine. Tielt: Lannoo. ISBN: 978-90-209-7622-9
  
• Piaget, J. (1924). Le language et la pensée chez l´enfant. Neuchatel 1956
  
• Senge, P.M. (1992). De vijfde discipline. Schiedam: Scriptum Books
  
• Servan-Schreiber, D. (2003). Uw brein als medicijn. Utrecht: Kosmos Uitgevers
  
• Sitskoorn, M. (2006) Het maakbare brein. Amsterdam: Bert Bakker, ISBN: 978-90-351-3064-4
  
• Vroon, P., van Amerongen, A. & de Vries, H. (1994). Verborgen verleider, psychologie van de reuk. Paperback. Baarn: Ambo. ISBN: 90-263-1326-8

Juni Daalmans
Daalmans Organizational Development
September 2008