The ontologies of complexity and learning about complex systems

Abstract

This paper discusses a study of students learning core conceptual perspectives from recent scientific research on complexity using a hypermedia learning environment in which different types of scaffolding were provided. Three comparison groups used a hypermedia system with agent-based models and scaffolds for problem-based learning activities that varied in terms of the types of text based scaffolds that were provided related to a set of complex systems concepts. Although significant declarative knowledge gains were found for the main experimental treatment in which the students received the most scaffolding, there were no significant differences amongst the three groups in terms of the more cognitively demanding performance on problem solving tasks. However, it was found across all groups that the students who enriched their ontologies about how complex systems function performed at a significantly higher level on transfer problem solving tasks in the posttest. It is proposed that the combination of interactive representational scaffolds associated with NetLogo agent-based models in complex systems cases and problem solving scaffolding allowed participants to abstract ontological dimensions about how systems of this type function that, in turn, was associated with the higher performance on the problem solving transfer tasks. Theoretical and design implications for learning about complex systems are discussed.

Appendix: selected texts of mini-lessons for complex systems ideas and concepts

This Appendix provides texts from the Complex Systems Knowledge Mediator hypermedia program for the complex systems concepts (referred to as Formal Ontology Scaffolds) of Self Organization and Emergent Properties and Levels and the complex systems ideas (referred to as Implicit Ontology Scaffolds) of System Control and Agent Actions. The learner would access these by clicking on the Mini-Lessons link on the left side of the CSKM screen (see Fig. 1).

Complex systems concepts

Self-organization

The concept of self-organization is really about hidden order that is all around us. It is a fundamental construct in the sciences of complexity, but it is often misunderstood. This is because many people assume that a system behaves in an orderly or organized manner because “something” or “someone” causes it to be orderly. However, most phenomena in life, such as atoms and molecules, cells in the body, or companies in an economy, are actually complex systems that are made up of many, many independent but interacting parts or agents. These parts interact with each other and with the environment in various ways—including random or chance interactions. Over time, the interactions of the parts may be looked at as patterns or order in the overall system. Take, for example, an ordinary day in New York, or any city in the world. People buy breakfast, drink coffee, put gas in their cars, and generally go about their business. How did all of the different kinds of food “show up?” Who knew how much coffee or gasoline to buy for everyone in the city? And so on… Is there a central planning group that figured all of this out? Obviously not. Magic? Unlikely. A city is in fact a complex system and the various kinds of decentralized interactions (e.g., feedback loops) of the parts (i.e., people, things) lead to the emergence of patterns and order without any centrally controlling mechanism. The dynamics of self-organization are effective at allowing people to get a wide range of food and services. Also, self-organizing processes are quite powerful and often long lasting. Consider, for example, how a city survives even catastrophes such as major fires or earthquakes that cause serious loss of life and major structural damage. Many scientists have been documenting how self-organization is found at all levels of nature, from the quantum and sub-atomic levels through our social and ecological levels to the very large-scale structures of galaxy clusters in the universe. Tip: Note that not all systems self-organize. Does your car, which is a mechanical system, “self-organize” or fix itself? Of course not, you need a mechanic to fix a system like a car. However, the flow of cars in the transportation system of a city does self-organize as there is no master computer that controls the speed and direction of all the cars, trucks, and motorscooters on all of the streets and highways in a city.

Emergent properties and levels

The formation of a flock from the interactions of the individual birds illustrates the concept of emergence or emergent properties. That is, emergence is when a new property is exhibited by a complex system at higher level of the system that is different than what you might expect from the agents or parts of the system. The concept of emergence is in many ways what is at the center of the most important research in the study of complex systems. How is it that molecules might interact in ways in which cell processes emerge? How is it the cells and organs of our body interact in ways from which our human personalities emerge? How is it that the individuals in a company interact in ways in which “organizational personalities” emerge? How is it that companies and people interact such that the characteristics of the global economy emerge? Tip: The new properties that might emerge from the interactions of the agents or parts in a complex system typically are observed “up” a level. For example, birds looking at each other while flying may be regarded as one level of their complex system, while the emergence of the flock formation is one level higher in that system.

Complex systems ideas

Control ideas: centralized vs. decentralized

What controls a clock? How is a classroom of kids controlled? Who controls an army? For each of these questions, most people would probably answer that something or someone is in control. In the case of the clock, it would be the interaction of the gears, while for the classroom, a teacher would be in control, and for the army, a general. We call these examples of central control. Now think about these questions. What controls a rain forest? What controls gas molecules in the air? What controls the flocking formations of birds? (You might want to investigate this model of birds flocking [link]. Is there a “leader” bird that you can identify in each flock? If not, then what “controls” the formation of the flocks?) In these systems, you might notice that there is no readily apparent “controller.” In the rain forest, the animals and vegetation have multiple different ways of interacting with each other and that the overall result of these interactions is the rain forest. Gas molecules, have certain properties that determine how they interact with each other in a room. And a careful examination of the flocking behavior of birds would reveal that there is no “leader” bird. These types of systems, which can be very stable and consist in terms of how they function, are examples of decentralized control. This idea of Decentralized Control is an important concept about complex systems. Note, though, that it is quite different than the more common idea of central control. Tip: When you think about different types of systems, try to decide if the system is controlled in a decentralized manner or a centralized manner. If there is decentralized control in the system, then it is probably an example of a complex system.

Actions: random vs. predicatable

Perhaps the most common example of the effect of an action is hitting a pool ball. Since the time of Newton, if the force, direction, and shape of an object such as a pool ball was known, then it could be calculated in a predictable manner from the actions of the ball, both individually and in interaction with other balls. But a pool table is not a complex system. However, one of the great lessons of modern physics, specifically from quantum mechanics and the Heisenberg uncertainty principle, is “The more precisely the position of a subatomic particle is determined, the less precisely the momentum is known in this instant, and vice versa”. This insight shows how nature fundamentally operates within parameters of randomness and chance. We also find chance factors in our day-to-day lives, with the weather, traffic jams, or when we meet someone who becomes a good friend. What is important from a complex systems perspective is that even though there are random factors in phenomena at all levels of nature, we still see emergent patterns and self-organization—other concepts we study in these lessons. The study of why these patterns occur is at the heart of cutting edge research in the cross-disciplinary fields of complexity. Tip: When you think about actions that occur in a particular system, try to decide if the actions are predictable or random. If the actions have random elements, then it is probably an example of a complex system.