Assembling the heterogeneous elements for (digital) learning

Making the LMS more like the Globe Theatre: distributed cognition, the extended mind and Moodle

The following draws together some work on distributed cognition in the Globe Theatre and some early thoughts on how that might be useful in improving the design of learning management systems such as Moodle (there’s also an argument to be made about university teaching and learning, but I’ll mostly leave that alone). While primarily focusing on helping teachers teach and thus improve the student learning experience, the ideas could equally be applied to learners (and probably have been).

At the core of the following seems to be the question of whether o not the design of LMS, and e-learning tools in general, have been limited due to the following observation made by Tribble (2005)

consistently distorted by a tendency to view cognition as individual rather than social, which has caused us to imagine the workings of complex group structures in mechanistic terms. In other words, we have mistakenly assumed that properties of the system as a whole must be possessed by each individual within it.

Shakespeare's - Globe Theatre

The post continues to build some recent thoughts about this issue and was sparked by listening to this episode from ABC Radio National show – The Philosophers Zone. This led me to the article Distributing Cognition in the Globe by Tribble (2005).

After finishing the post, the idea of adding “plots” to LMS tools might be the most interesting outcome. Will need to do more reading and thinking. Must be other folk looking at this, pointers?

Cognitive demands of the Elizabethan repertory system

Imagine you are a member of an Elizabethan repertory theatre company in the late 16th century. Just how cognitively demanding is your work? Tribble (2005) draws on a number of sources to give the following examples:

  • As a leading player you would have to remember 71 different roles, around 52 of which were new.
  • Your company would be putting on 6 plays a week, there would be infrequent repetition, and would have a new play every fortnight.

This appears to be quite different, and cognitively a lot more demanding, than more modern theatre companies.

So how did they do it? Tribble (2005) summarises some literature that suggests it was achieved through “mechanistic” approaches such as: actors performing the same type of role, learning roles by imitation, and having little or not interest in creative interpretation.

How did the Globe and distributed cognition help?

Tribble’s (2005) argument is that the design of theatre system – embedded in the theatres, the plots, actors’ roles, the plays’ verbal structures, the apprentice system, and the organisational practices of the companies – provided elements of cognitive structure that helped individuals and the company to fulfill their cognitive demands. In particular, she argues that the design of the theatres played a part

The more thinking that can be off-loaded onto the environment, the more mental energy remains available for those tasks that are primarily internal (memory for the spoken lines, for instance).

For example, it is argued that “plots” – summaries/directions of when, who enters the scene – hanging back stage from pegs by stage entrances provided players with reminders of the order of scenes. Especially helpful for actors playing multiple small roles. Another is that rather than all players having the full script, they instead have their lines in detail and various cues to indicate when they are to be performed. Similarly, it’s argued that some of the physical spaces within the theatre (e.g. a balcony) provided cognitive hints as to what type of scene you were meant to perform.

The podcast that brought me to Tribble and her work includes a better description starting at about 13m48s. Tribble’s summary

The productive constraint of the stripped down part reduces the need to filter signal (one’s own part) from noise (everyone else’s); the plot provides a schematic diagram of the shape of the play as a whole to supplement the part; the physical space of the theater and the conventions of movement it supports enable the transition from the two-dimensional maps of plot and part to its three-dimensional embodiment onstage; and the structures and protocols of the theatrical company pass on its practices to new members. Such a theater can best be understood, in other words, through a framework that takes group practices seriously, that assumes that systems can work well, and that sees individual agency as constrained but not contained by these practices.

Tribble (2005) draws inspiration from Hutchins (1995) and describes on of his contentions as

that the lines between “inside” and “outside” are frequently misdrawn or misidentified, “creat[ing] the impression that individual minds operate in isolation and encourag[ing] us to mistake the properties of complex sociocultural systems for the properties of individual minds.”

and that when the design of these systems are done well

“Novice[s]” can thus be “embed[ded]” in the system and perform at a level far above that which they could attain by means of their individual cognitive powers alone.

The design of the LMS and the tyranny of the individual mind

It is my suggestion that the design of the LMS (and probably the assumptions underpinning much of the actions around improving university teaching) is limited by the tyranny of the individual mind. Sturgess and Nouwens (2004) describe one universities rationale for an LMS as being

to enable teaching staff to develop and
manage online courses with little professional support

. This fits with the broader practice of university teaching being essentially a solo act. It’s my course, I’ll teach it my way. Give me control of the course on the LMS and let me do my thing.

It is assumed that the academic (or their proxy who has been employed to create the course) has the skills and knowledge to effectively use the features of the LMS to prepare the course site. When adopting a new LMS, a focus for most institutions is on training sessions to provide these skills and knowledge to staff.

But what additional scaffolding or “distributed cognition” does the design of most LMS provide? In my experience it is generally limited to either some canned “help” document or the contact details of a support group. The “help” resources are usually fairly limited and it’s not unusual for them to be out of date. The support group are generally keen and willing to help (often above and beyond the call of duty) but are often limited to knowing where the academic might find a query, or about how to use the LMS features as they exist.

I’m not sure that most LMS are designed in a way such that

“Novice[s]” can thus be “embed[ded]” in the system and perform at a level far above that which they could attain by means of their individual cognitive powers alone.

. The evidence from past experience with LMS seems to indicate that most e-learning is not all that good, perhaps indicative of how well the LMS support distributed cognition.

Some ideas for BIM

BIM is a Moodle module that I’ve written. It’s probably going to be the first place I experiment with some of these ideas. Some initial thoughts:

  • Improve the help documentation provided with BIM to include the videos that have been produced and also provide links and summaries of some of the literature around the use of blogs in higher education.
    This is a simple first step that arguably should have been done already.
  • Link the help documentation with some the BIM community.
    Rather than just static help documentation, provide mechanisms for asking questions of and seeing answers/resources from a broader BIM community.
  • Design a scaffolded “BIM configuration” process informed by common problems and linked to the help documentation, especially the literature.
    As it stands, the BIM configuration process is your typical IT artifact. It is a better representation of the data requirements of BIM than a process that helps academics create and design a BIM activity that is effective.
  • Figure out a way to develop standard BIM “plots”.
    There are common events in the use of a BIM activity (e.g. not all students have registered a blog, not all the students within the course have been allocated to groups and markers, etc.). These can be picked up from the BIM interface, but the academic has to apply a fair bit of cognition in understanding how to do this.

    A better approach might be to implement some “plots” (in the sense described above) that suggest when and how to handle a particular event. For example, when BIM sees that not all students in a course are allocated to groups, generate a “plot” entry that includes a list of the unallocated students and links to the Moodle group allocation process and the help documentation associated with groups explaining this problem.

Ideas for Moodle

Some of the ideas for BIM could potentially be applied to any complex process within the use of an LMS/Moodle such as the design of a course site or the management of a discussion forum. This links to some extent to the earlier suggestion of combining analytics and discussion forums.

Misc reflections

Beginning to wonder how much of this is related to the design of computer systems being done by programmers more interested in representing the models underpinning the system, rather than providing cognitive support to the folk using the system?

Wondering how much the type of support given to academics using an LMS within a University context in terms of the skills of the support staff and the organisational structures within which they work limits the provision of appropriate cognitive support?

Wondering if an increasingly hierarchical organisational structure can effectively provide the type of distributed cognition that might actually help improve the LMS and university teaching?

Wondering if the hierarchical organisational structure can effectively grok the need for this type of approach? Can it provide the environment in which this type of work can grow?


Hutchins, E. (1995). Cognition in the Wild. Cambridge, MA, MIT Press.

Sturgess, P. and F. Nouwens (2004). “Evaluation of online learning management systems.” Turkish Online Journal of Distance Education 5(3).

Tribble, E. B. (2005). “Distributing cognition in the Globe.” Shakespeare Quarterly 56(2): 135-155.


The nature of user involvement in LMS selection and implementation


The problem with blended learning


  1. Of course, just realised, that the obvious other thing to do is to look for folk who have done this sort of analysis/research a la Tribble and Hutchins of university teaching and/or using the LMS. Or perhaps do it.

  2. kwilco

    David –

    I believe that, in addition to what you are saying about distributed cognition, adding layers of “choice architecture” for users depending on their level of expertise would be most useful. Let’s face it, beginners are faced with too many options, most of which they know nothing about. There is nothing built into systems that make it hard for people to make BAD decisions.


    • G’day Kevin,

      Agree whole-heartedly. Think the interesting questions will arise from how to design the choice architecture effectively, how widespread its applicability would be (i.e. would the same choice architecture be effective in an Australian university as in a US one?) etc.


Leave a Reply

Your email address will not be published. Required fields are marked *

Powered by WordPress & Theme by Anders Norén