Gagne, Games, and Learning

Reading Brett over on Rhymes with Purple got me thinking about the Gagne Assumption. The assumption is two part. First, there are different kinds of learning (e.g., learning facts is a different thing from learning to classify). Second, different conditions are most suited to bring about these different types of learning. If one buys into this assumption, which I do wholeheartedly, then a prescient question regarding games becomes – what type (or types) of learning are best promoted by game-like instructional conditions?

I haven’t read nearly as much of Gibbons‘ work as I should have (and neither have you, by the way). But I would hazard a guess that Andy would say that when students are learning about complex systems, games are a great condition under which to learn. Andy talks about students poking and prodding parts of the model to see how the model responds. For example, I’ve watched my nine year old playing the role of Mayor inside Sim City 4, seen him zone huge areas for high-density housing when there was no infrastructure to speak of. The model responded in a particular way. I’ve seen him ignore expert advice to put some fire departments in place. The model responded in another way.

This action-feedback loop is what underlies all learning. And now, for a brief digression in which I reveal more about myself than you wanted to know.

<digress>
Here’s Wiley’s theory of learning in a nutshell. An agent (you or I, or a cat or mouse for that matter) sends messages out, and receives messages in return. The rest is a combination of pattern-matching and purposiveness on the part of the agent.

As a teacher, or from the outside, there is very little that we can do to change the pattern-matching and purposiveness parts of the activities agents engage in (other than the usual meta-cognitive coaching and reward / punishment stuff you already know about). This means that the big opportunity for instructional designers or teachers lies in crafting the messages that go back to the agent.

Now, this all sounds very behavioral. And, in fact, it can work that way when the agent is only pattern-matching and not engaging in purposive activities. However, to the extent the agent begins to act purposively, these same mechanisms will underly the most constructivist types of learning. The agent will engage in some activity which (in an abstract way) sends a message out into the environment (like writing and running a program in Netlogo), the agent will receive a message (some feedback) back from the environment (perhaps the program will throw errors or behave in an unexpected way), and the agent will use this new information to form another probe out into the environment. I do this all the time when trying to install new software from source or tweak the CSS on a webpage. I send messages out to Google, it sends messages back to me, and if the contents of the message meet my purposes I stop. If not, I rinse and repeat. In it’s reliance on conversation as a mechanism, I suppose it’s quite a Pask-ian view of the world. (FWIW, I think this very simple view of the world meshes very well with what we know about the biological mechanism underlying learning.)

So, in a pointless attempt to be clear, I think there are four things worth considering when we think about teaching and learning: (1) the types of messages the agent sends out, (2) the types of messages that come back, (3) the pattern-matching mechanism, and (4) the role of purposiveness.
</digress>

So what does all this have to do with games? I think we have to think carefully about the types of learning games are really suited to facilitate. I think we have to think carefully about the types of messages we enable learners to send into the virtual environment, the types of messages the environment sends back, how the learner is going to make meaning out of patterns in these messages, and how the learner expresses their purposiveness in the game environment. If there’s any response to this post, maybe I’ll do a quick sketch of Sim City or another game from this perspective.