Natural data processing nodes tend to specialize themselves. Our brain works with different internal structures where neurons are activated under different kind of stimuli. Neuroscientists have created maps of different regions for the different senses. This kind of regions are located in similar for all individuals, however this is not totally true. People that have born with a congenital lack of a sense can use that part of the brain to store information from other stimuli. This has been demonstrated scientifically many years ago, with an experiment where a group of cats was blinded at born. The results showed that the visual cortex in the blinded cats was activated with other kind of stimuli while the sighted group of control preserved the location of the visual cortex. This does not happen in individuals that have lost vision some years later. In this case, visual cortex usually is not activated again.
In a previous article I was trying to show why modern companies try to reduce the number of managing levels considering managers as center of information processing with a simple neural network model. Today, I am going to make a similar description in order to show why the distribution of the information in large organization is not uniform. As I proposed before it is possible to consider that information in a large organization is distributed in brain-like way although the processing centers (manager have human brains) are much more complex than a single neuron, however, information can flow in a similar way because managers’ work is done establishing connections with other processing centers and filtering the information (part is sent to the following level and other pieces of information are not). In other words, some information activates the output to the next level and some information does not in a similar way as a neuron does.
Most business organizations are organized in functional departments. There are some tasks that must be done always and organizations are designed from their starts with that specialization: accountancy, sales, marketing, production, R & D, human resources, and so on. The information flow in the organization in order to accomplish a task would be moving through certain departments depending on the task.
Imagine now that the company enters in a crisis time, and the board of directors eliminates the budget of marketing for instance. People at the marketing department would not receive any new task while the rest of the organization would continue with their own work. This kind of full specialization would produce an inefficient organization when the external environment changes.
We can think now, that it would better a more flexible organization where people can be assigned to different tasks in order to fit better environmental changes. However, this is not totally true. If you look back in this discussion and you remember the experiment with the blinded cats, you could see that the specialization of neurons is not genetic (only). The visual cortex in the blinded cats could be activated by auditory stimuli, because they were not specialized initially. The external stimuli from the environment are driving how the neurons are being specialized.
This fact can happen in business organization without functional areas. Any organization tends to specialization through a similar process. It is very probable that the board of directors assigns similar tasks to the same people in order to take advantage of the previous experience and to improve the efficiency, and if they do not make it, it is probably that people with the assignation ask for support to the people with that kind of experience. Finally, there will be group of people that will receive information about certain tasks always inside the organization while others will not.
This process can be seen through mathematical models with different kind neural nets. In the field of artificial intelligence there are well known the Kohonen self-organizational maps. This kind of artificial neural network is commonly used for classification tasks. Kohonen networks produce a low level representation of the input space and they apply competitive learning algoritms instead of error-correction learning. With this kind of network it is easy to show a “brain-like” configuration of processing centers where finally certain group of networks are activated under similar stimuli.
In a real organization links are established in a natural way between managers in order to improve the efficiency of their tasks, and these links are reinforced when the relationship lets to make their assigned tasks in an easier way. This kind of collaboration is not so odd because it is proposed by the experts of human resources. As in a Kohonen self-organizational map, when a network of relations is created between managers of different groups of people, information about certain tasks would flow through the same group of people always.
The interesting question under this discussion is to analyze what happens in this organization when the environment changes. If the environment eliminates certain kind of tasks, would some part of the organization be without working because the information flow will move through other groups of people in the same way as the visual cortex is not activated in a blind person?
An affirmative answer is very probable, but how can we analyze this fact? There is a way to do this through complexity analysis. Those self-organizational links between are building the structure of the company in the same way as the predefined hierarchy is the structure of the company in a functional organization. A complexity analysis can be used in both situations with the same meaning for the concepts of complexity and resilience of the organization. A complexity map will reveal the configuration of the links between different managing centers both functional and self-organizational. And it would reveal if the managing centers would be establishing new relationships or they would not.