At the recent DevLearn, several of us gathered together in a Junto  to talk about issues we felt were becoming important for our field. After a mobile learning panel I realized that, just as mlearning makes it too easy to think about ‘courses on a phone’, I worry that ‘learning experience design’ (a term I’ve championed) may keep us focused on courses rather than exploring the full range of options including performance support and eCommunity.

So I began thinking about performance experience design as a way to keep us focused on designing solutions to performance needs in the organization.  It’s not just about what’s in our heads, but as we realize that our brains are good at certain things and not others, we need to think about a distributed cognition solution, looking at how resources can be ‘in the world’ as well as in others’ heads.

The next morning in the shower (a great place for thinking :), it occurred to me that what is needed is a design process  before we start designing the solution.  To complement Kahnemann’s Thinking Fast and Slow (an inspiration for my thoughts on designing for how we really think and learn), I thought of designing backward and forward.  Let me try to make that concrete.

What I’m talking about is starting with a vision of what performance would look like in an ideal world, working backward to what can be in the world, and what needs to be in the head.  We want to minimize the latter.  I want to respect our humanity in a way, allowing us to (choose to) do the things we do well, and letting technology take on the things we don’t want to do.

In my mind, the focus should be on what decisions learners should be making at this point, not what rote things we’re expecting them to do.  If it’s rote, we’re liable to be bad at it.  Give us checklists, or automate it!

From there, we can design forward to create those resources, or make them accessible (e.g. if they’re people).  And we can design the ‘in the head’ experience as well, and now’s the time  for learning experience design, with a focus on developing our ability to make those decisions, and where to find the resources when we need them.  The goal is to end up designing a full performance solution where we think about the humans in context, not as merely a thinking box.

It naturally includes design that still reflects my view about activity-centered learning (which I’m increasingly convinced is grounded in cognitive research).  Engaging emotion, distributed across platforms and time, using a richer suite of tools than just content delivery and tests.  And it will require using something like Michael Allen’s Successive Approximation Model perhaps, recognizing the need to iterate.

I wanted to term this performance experience design, and then  as several members workshopped this with me, I thought we should  just call it performance design (at least externally, to stakeholders not in our field, we can call it performance experience design for ourselves).  And we can talk about learning experience design within this, as well as information design, and social networks, and…

It’s really not much more than what HPT would involve, e.g. the prior consideration of what the problem is, but it’s very focused on reducing what’s in the head, including emotion in the learning when it’s developed, using social resources as well as performance support, etc.  I think this has the opportunity to help us focus more broadly in our solution space, make us more relevant to the organization, and scaffold us past many of our typical limitations in approach.  What do you think?

A respected colleague recently suggested Andy Clark’s Being There as a read to characterize the new views of cognition, so I checked it out.  The book covers the new emerging views of cognition, grounded in the connectionist revolution and incorporating a wide variety of neural and robotic studies. The interesting thing to me are the implications for learning and instruction.

The book makes the case that the way we think is not only heavily tied to our contexts, but that we co-construct the world in ways that affect our thinking in profound ways.  Studies across economic behavior, animal cognition, simulation studies, and more are integrated to make the point that they way we think is very different than the models of conscious minds sitting in meat vehicles.  Instead, we’re very driven from below and outside, and our conscious thinking is rare, hard, and language based.  Moreover, the constructs we create to think affect our thinking, making it easier. We automate much not only through learning, but we externalize.  And, our representations and understanding are very much constructed ‘on the fly’ in each new situations, as opposed to existing abstract and robust.

This isn’t easy reading.  Clark is a philosopher of mind, and covers much complex research and deep neuroscience.  The emergent picture, however, is of a mind very different than the cognitivist model.  I’m grateful that while I pursued my PhD in Cog Psych, the research going on in our co-shared lab by Rumelhart and McClelland on connectionist networks sensitized me to this viewpoint, and Hutchins work on Cognition in the Wild was similarly taking place at the same time.   Despite the challenge, there are important reasons to get our minds around this way of thinking.

The notion that providing abstract knowledge will lead to any meaningful outcome has already pretty much been debunked both empirically and theoretically.  What these models seem to suggest is that what can and will work is deep scaffolded practice and guided reflection, based upon a situated cognition. For other reasons, this is the model that Collins and Brown had in Cognitive Apprenticeship, and now we’ve a more solid philosophical basis for it.  (I also think that there are rejoinders to Kirschner, Sweller, & Clark, and Anderson, Reder, & Simon; discussing how language, including writing, is social, and that iterations between abstract models and meaningful practice is guided reflection.)

This model suggests that language is our differentiator, and that much of our higher level cognition is mediated through language.  There’s a reason consciousness feels like a dialog.  Much of our processing is beneath consciousness, and things we monitor and develop through language become compiled away inaccessible to language.

The point, to me, is that the activity-based learning model I’ve proposed has both bottom up grounding in new cognitive models, theoretical framing from anchored instruction and social constructivism, as well as empirical validity from apprenticeships and work-place learning. We need to start aligning our learning design to the cognitive realities.