This post is inspired by a profound essay from Mike Caulfield called Design Patterns and the Coming Revolution in Course Design. It’s the text of his keynote address at the Northwest eLearning conference earlier this week. While Mike’s examples come from the arena of academic course design, instructional designers and learning consultants in the corporate world will recognize the issues and relate to the recommendations. I urge you read Mike’s post before continuing with this one. Get a cup of coffee first; it’s a long one – but worth every word.
Thanks for coming back… here’s why I’m so excited about Mike’s propositions:
The need for a pattern book
I am sometimes bothered by the way we can block-copy design ideas from one context to another, resulting in same-same course designs that lose something in the translation. Since I am educated in adult learning theory and research, I have always recommended that designers get solid grounding in the theory and research around how learning works so that they can make more informed decisions about design. I can see, though, that documenting design patterns and using these as guideposts for course design is a fabulous idea.
Patterns translate key principles from adult learning into clear recommendations. Folks like Julie Dirksen, Karl Kapp, Michael Allen, and Ruth Clark (references below) have done some of this translation work for people in L&D, but I don’t think they quite constitute design patterns as Mike describes them. I’ll have to check out the book that Mike mentions,Technology-Enhanced Learning: Design Patterns and Pattern Languages, edited by Goodyear and Retalis.
We need a design pattern book for our work. I’ve no idea how to create that, but I’ll be noodling on it for quite a while. As I think about it, it feels like a herculean task. There may be too many if-then’s (or when… therefore…’s) that form the foundation of design. And if you do create a pattern book, would it be too big to be really useful? You’d have to categorize the patterns, I think. But I bet it would be more useful than reading and trying to apply a textbook on learning theory. (Don’t tell my adult learning students I said that.)
The design process
Another aspect of Mike’s post that resonated with me was the analogy between course design and music composition. He says, “When you work solo on multi-track, as I do, and you are scoring 5 to 20 instruments you become acutely aware that each decision you make constrains you further. As you progress, there really are limited ways these things can fit together.”
I have come to the point where I describe design as a set of decisions (audience, objectives, content, delivery method, activities, and structure). These decisions can’t be made in a linear fashion because each decision imposes constraints and opportunities for the other decisions. There are many different ways to finalize the decisions that will work, and some that simply won’t work. The music analogy is truly applicable. While you are composing/designing, you constantly adjust your decisions to ensure they continue to blend together as a whole. There are limited ways these things fit together, but there are multiple ways that would produce a pleasing piece.
This insight is timely because next week, I’m working on the design process section of my Learning Environments by Design book. Designing a learning environment isn’t as constrained as designing a course – it’s a much more emergent, ongoing process. I want to be sure I communicate that eve though you might need a vision to launch an environment, it will morph and change over time – perhaps more like an improvised jazz piece.
Thanks to Mike, then, I’ll have lots to think about – as I am designing courses and learning environments – and as I am helping others to develop their design skills. Comments welcome!
Books referenced above: Julie Dirkson’s Design for How People Learn, Karl Kapp’s The Gamification of Learning and Instruction, Michael Allen’s Designing Successful e-Learning, and Ruth Clark’s e-Learning and the Science of Instruction.
Hi Catherine,
I’d go a step beyond a book to a live connected community of cultivated and shared pairings. Julie and I have bounced this idea around for the past couple of years, chatting with a few designers. We’ve gotten quite a bit of pushback, some seeing a community pattern library as a stack of templates or constraining guidelines.
This is part of our problem. A lot of isolation in the field, often connected only by tools, technology, and – in a looser sense – theory. A pattern language could revive all of the great foundations that have underpinned fantastic learning opportunities while pulling designs that don’t reach these heights up a notch or two.
Thanks, Steve. I had a chance to look at the article you left on Mike’s post and your sketch and I’m still noodling them. The article really demonstrated to me just how hard it will be to create a usable pattern library. The problems and circumstances designers encounter are never quite that simple. It’ not a straight if… then… – the thought process is more like, “well, if I do this, then…, but that impacts this, so what if I do that as well, and it would then be better if I also…” I’m thinking that the approach might need to be phrased more as an option, and there might be multiple options for any given situation. For example, if I want to build community in an online course, my design options might include having learner profiles, encouraging students to contribute content, including social media components for those who use social media, etc. As I say… still noodling. Thanks so much for sharing your work on this point. Please keep me in the loop if/when you do more.
Thanks Catherine. It feels like the difficulty will come more from bringing folks together to talk about commonalities and to move towards agreement than to create a collection.
Totally agree about the variety and specialization within the problem space. However, specialized complex problems aren’t incompatible with a well constructed pattern language. Patterns are intended to be represented in mutable forms and at various scales. If we think of a pattern description as a frame of reference attached to a common descriptor, the pattern becomes a bridge to communication and what could be seen as a universal anchor point.
We do a lot of things in this industry that we might call unique. However, I argue that within these unique contexts, many of the things we do are very similar and amount to complex arrangements of simple things that pair with common problems. To define these things in a language that helps us communicate and layer solutions in ways that honor and connect to prior works could provide a huge boost to all of the related industry verticals.
If we asked 20 practitioners to define what e-learning means to them, you might come away with 20 unique definitions. It’s soft-science. Patterns provide anchor frames that link problems to potential solutions. So we could (in my opinion, should) drop the e-learning moniker for something specific and meaningful.
A common language clarifies concepts of application and helps folks find common ground to visualize the when and why behind design choices.
Imagine how unsafe it would be to fly if pilots didn’t have a common language for the mechanics of a flying vehicle. Extend this common language to a language for maneuvers and now we can have conversations in our domains with specifically named root patterns. This doesn’t describe a pattern library or collection but it does help to articulate the value of shared language.
http://en.wikiversity.org/wiki/Airplane_Flying_Handbook/Basic_flight_maneuvers
http://en.wikipedia.org/wiki/Aerobatic_maneuver
It would take forever for a single person to build a pattern library. And that library would be of most value to the individual that assembled it. Pattern libraries aren’t single person endeavors. Built by community, the language begins to form a sense of connection with the domain. Because it IS the domain:) It simply describes the tacit stuff in a way that makes solutions useful an extendable.
I could submit a problem description at a broad level along with a solution pattern that worked in a particular instance, I might call this the “Stubby Alligator”. You could come along and say “my situation was a little different. Part of that pattern worked for me but we needed to do X. So I’m going to vary your pattern and document the elements I changed, why I changed them, and what makes the variant valid in my case.” I’m going to call the pattern the “Baby Crocodile”. I also established a few new child patterns since my arrangement changed and I attached existing patterns “X” and “Y” to this pattern to address specific challenges.
This is a goofy example with absurd names. But it illustrates the intent and mechanics of a pattern collection. It’s never finished and the collection will quickly build to become far more than the sum of its parts. 🙂
Wow. That does help me see where we might go with this.
But how do you keep the pattern library from becoming unwieldy and too difficult to search? The other question that continues to spin in my mind is whether a pattern library is better than a solid knowledge of adult learning theory. I thought a pattern library would be more help and easier to grasp than having and maintaining a deep background in learning theory and research, but now I wonder if it would be better for a practitioner to invest energy in understanding the theoretical base on which we stand. I remember hearing many years ago the advice that if you know adult learning theory, you can design using any tool or technique available (or at least you can make assessments of their affordances), and I think that’s true. I wonder if it would be difficult for a practitioner to make decisions about which patterns to use without that.
The advantage of a pattern library is that the translation from theory to practice as it applies to problems is already done for you to some degree. A pattern library may be better than our idea books (collections of activities for learning) in that it associates solutions with problems and it could explain the reasoning behind the pattern, which our idea books don’t often do.
Still noodling.
I don’t see a pattern library and a deep connection with applicable theory as mutually exclusive. A pattern library is one way to make practical sense of theory and increase the chances that choices are made that align with validated research.
A library can get unwieldy without commitment to maintenance. I think that’s a danger. However, if valuable, useful, and open I think the risk of a pattern library turning into a jungle or a ghost town is reduced.
There are precedents. A few pattern libraries that continue to be useful in software. I think these have a life. At the beginning of the pattern library’s life, it grows and sprouts in a lot of directions during a period of heavy activity. The second phase looks to be where curation happens. In the third phase, popular patterns at the height of the use curve become domain canon. Eventually the library cycles back to the first and second phases where additions are made when new research or trends push industry in one direction or another.
When Ian and I presented the concept to a room of ISD folks a decade ago, the dominant response was “why do we need this, we already have best practices within our organization?” Which, of course, missed the point entirely. Best practices (a term of which I am not fond) are tuned to a homogenized process view within a closed system. Patterns are specifically matched to a context that worked but doesn’t assume *best*. Pattern libraries are also built by community.
Circling back to the theory pattern connection. I think there is a huge opportunity to connect theory and research with practice by using research as a validation factor for patterns.
A few folks would take a pattern as it is, and use that pattern or set of patterns with impunity and without insight. Certainly we’d be no worse off than the glut of mindless application and misunderstanding by even those educated in theory. Others would take a pattern and remake it as their own, returning to the pattern with validation of efficacy, citing their case, and expanding the body of research (to a degree). In this case, I think collectively we’d be far better off than much of the works and supports we see in action today.
Here’s a quick sketch of how use cases could connect with a pattern library service:
Click to access patterns.pdf