Have been reading about the science of ‘Complexity’ recently, and find that many of the aspects described apply equally well to learning. Without doubt, learning is a process that is riddled with complexity – which might the reason we don’t understand it very well.
How might ‘complexity’ as a scientific concept affect learning? Especially if you consider learning as a complex adaptive network that is driven by human desire for knowledge (learning) and sociability. I had some thoughts about that.
(1) Learning happens in networks, call it informal or social, but it’s the network and its nodes that actually make learning happen. For learning to occur in an audience, it can’t really be a one on one event, but rather an interaction with the ‘agents’ that form the network. In the case of human networks, the agents (humans) not only act as part of the network but also produce artifacts that can be shared across the network. These could be considered as a type of resource generated by the network. The objects could be simple like a rule or best practice that has been documented, to more complex objects like an excel sheet that helps users calculate specific values.
(2) As is innate in most networks, there will be some sort of competition among the agents for limited resources. In larger networks we might also find that there is local collaboration among network agents – this resembles the typical team plays in an organization. Local teams compete against each other for resources, but remain a part of the larger network. This assumes a network boundary that is fixed and encompasses all employees as agents. However, as anyone associated with learning will say, the boundary can’t really be defined as a static element, it changes – grows and shrinks based on the agents coming into the network and falling out; it is difficult to draw boundaries around the learning networks that humans create for themselves. The competition for resources results in an uneven distribution of said resources. This might explain why knowledge remains in organizational silos, it may be that the network agents use knowledge objects to gain a competitive advantage.
(3) Networks are beautiful structures for enabling feedback. Every agent action ripples through the network and may even be amplified by other agents. This is the sort of behavior that plays an important role in learning. Learning cannot happen without feedback. Looked at in that perspective, feedback within the network is quick and remedial. Also, as objects/behaviors propagate through the varied paths in the network, feedback can come from agent nodes far away (non-local) almost like wave after wave of feedback. This enables agents to react and modify their behavior appropriately far beyond the immediate reactive feedback they receive.
(4) Object propagation through the network is constantly evolving. Objects need to adapt their strategies in response to their history, and the feedback they have received from the agent nodes of their network. In learning, this can simply be equated with ‘objects/containers’ changing form to propagate best. In another light, it means that the nature of the content used in learning is always evolving and that change is based on feedback received from users. If the content isn’t changing based on the feedback, maybe we are doing something wrong. Perhaps the future will see adaptive content.
(5) A learning network can only be successful if it retains the ability to interact with its environment. Without some sort of environmental awareness, the network is unable to respond and it may ultimately result in degradation. With human learning networks, it is difficult to draw a comparison. To me, it seems like the environment is the workplace, and the network automatically responds to those changes. For example, if the environment is a factory shop floor, and the process or machines used in process change, the network must recognize this change and reconfigure itself to be able to provide learning about those changed elements. An appropriate response might be the creation of new objects, or may mean co-opting other agents into the network.
(6) One of the amazing things about complex networks is the principle of ‘self organization’ i.e., the ability of the network to exist without the need for a central controller. The ideal human learning network would be self organized too; and if you look closely, it is. We create personal learning networks that co-opt various agents and objects, there is no one individual or resource that we look to in times of need. Quite naturally, we send out messages across the network, receive feedback, analyze it, and utilize it to meet the need at the time. A centralized authority will only lessen the efficiency of such a network, if not take it away altogether. We must question our centralized L&D or training apparatus.
(7) Emergence of significant patterns of behavior within the network is another fascinating aspect of complexity. Every network, its agents/nodes and the objects contained within display a mix of ordered, disordered and chaotic behavior. A look at organizational learning will probably reveal exactly the same chaos, but individuals do learn and progress in the network and its environment. How does this happen? Can agents recognize a pattern in the underlying chaos?
There is more to come, in the my next blog post.