Gardens of Discussion: What Makes Online Communities Work?

gardenI’ve been contemplating online communities lately — specifically, that I don’t know what comprises the magic that makes one community tick and another one fail, despite the fact that I’ve participated in many different communities online over the last twenty years. I can tell you that the magic of a successful, vibrant community comes from people, but I still don’t know what the catalyst is. Finding this magic is somewhat like looking for the secret of life: we can identify aspects of what makes something alive, but we don’t know where the spark comes from. I’d like to examine a few aspects of online communities.

Sowing Discussion Online

Sometimes, online discussions seem to encourage the incidence of tumbleweeds and not much else.

Sometimes, online discussions seem to encourage the incidence of tumbleweeds and not much else. [Image from Flickr user OakleyOriginals.]

One piece of the question is, why is it that a seminar class might have lively, engaging discussions in the classroom, yet tumbleweeds blowing through the discussion board? It might just be human nature to rise to the occasion in a physically present group while the topic doesn’t really resonate strongly enough to make one type words on a screen at eleven o’clock at night. And at the same time, every once in a while the magic happens and discussion — honest, organic, creative discussion — manifests itself.

I can think of three major typologies of classroom discussion board, at least as far as the traditional definition of one (i.e., students/instructors post topical posts and others respond to both posts and responses):

  1. The Waste Land, which I hinted at above. Tumbleweeds. Nobody for miles, except for perhaps a lonesome professor shouting, “Hellooo! Is there anybody out there?”
  2. The Organic Garden, also mentioned above. Discussions grow and flourish naturally, and volunteer posts sprout forth from the screen.
  3. The Industrial Farm, where regimented numbers of postings and responses appear in chartered rows and in exact numbers. Usually planted via a line in the syllabus that says, “Students will, on a weekly basis, make one discussion posting, respond to three other discussion postings, and respond to two responses.”

All discussion boards start out as a Waste Land and are expected to grow by the rules of either the Organic Garden or the Industrial Farm. I’m fairly certain, though, that every proprietor of an Industrial Farm discussion board hopes that growth will sprout between the rows and become an Organic Garden, since the goal is always to achieve self-directed learning through the free exchange of ideas by the participants. Unfortunately, this is often not the case — instead, students jump through the hoop of the requirements, but don’t go beyond what is required.

I wonder, though, if some intentional vagueness might find a middle ground and foster something more akin to the desired outcome. For example, “class participation” is often a syllabized component of grading in a class but not clearly defined; if it were made clear that this included online as well as classroom discussion, perhaps students would strive to find a Goldilocks zone of online participation just as they do in the classroom setting, putting the focus on a “just right” amount of interaction rather than “just enough.”

Choosing the Proper Container

Another interesting thing to think about regarding online community is, what’s the proper format for the goals you are trying to achieve? There are many social sites on the Internet, and each one of them is characterized by their user interface (UI) and organizational characteristics. Think about the difference between the types of discussion that grows around posts on Facebook, Reddit, WordPress blogs, Youtube videos, Tumblr, Chronicle articles, etc. Each one has a different clientele, sure, but each one also differs in how posts and comments are displayed (Threaded? Flat?), entered by users (in a large text box? A tiny one? Can pictures be added? Can one reply to replies?), and organized (Chronologically?  By vote or moderation?).

I haven’t read any academic studies on the topic, but it’s obvious to me that the structure of a discussion framework has a lot to do with the types of discussion that grows there — ranging from pithy or obtuse one-liners to multi-paragraph missives and screeds (even rants! — always good for generating replies). The number of people in the community matters, too; some frameworks don’t scale terribly well (for example, think of a Facebook post with 500 replies).

An odd example of problems of scale I encountered was when I registered for a MOOC. New students were encouraged to post in the “Introduce yourself!” thread in the discussion board. There were over a thousand replies in this thread. Well, I figured I’d be a good sport, even though there’s no way anyone was reading all of these — but I couldn’t for the life of me figure out how to post. After fiddling around for a few minutes and getting increasingly frustrated (I’m an IT professional, and I can’t find the “New post” button?!?) I had the bright idea to look at a thread with only a few replies.  And there was the “New post” button–right there at the bottom. Back at the “Introductions” thread, there was no “New post” button–just a “More replies…” button.  “Really?  Noooo…” I thought; I clicked “More replies…” and again… and again… loading 30 or 50 replies at a time. When I finally got to the bottom of the thread, there at last was the button I was looking for. Someone neglected to consider scale.

What Does the Garden Club Think?

While I have plenty of experience as a participant in online communities, I’m positive faculty readers of this blog have more experience than I do at attempting to create them. Allow me to toss out a seed and say, please share your knowledge! What kind of experiences have you had with online discussion outside the physical classroom — and did the format and formal class requirements make a difference? The comments section below would be a perfect place to do that. This is, after all, a community garden!

About John Drummond

John Drummond is the Academic Engineering and eLearning Manager at the College of William & Mary. Originally from Mathews County, VA, John graduated from James Madison University with a BA in English in 1996 and an MS in Technical and Scientific Communication in 2002, and has been with W&M IT since 2007. In addition to working in AIS, John has taught occasionally at W&M and previously at Tidewater Community College, and in other roles has been an author, a musician, a Perl programmer, a UNIX systems engineer, and a network manager. He resides in Toano with his wife Andrea and daughter Rebekah.


  1. Hi John,

    What I have found most helpful in promoting engagement is to include one-question polls with the posting that one is making. For e.g. if a course lead is sharing an journal article in one of the class forums, they include a poll about one of the key concepts discussed in the article. The more interesting the poll, the more engagement the post will see. Somehow single questions polls prompts discussion in addition to the poll responses.


  2. Great post, John. This is something I’m really in the middle of navigating right now with a course I’m prepping for next week (if you’re interested here’s the link to a post I’ve written on the experience – We’ve opted for the “industrial farm” approach – mostly out of necessity in a complex course – but the teacher in me feels a little uncomfortable with this. I feel like the “organic” approach is more in line with my teaching philosophy and probably a better experience for students. I guess the trick is in finding a good balance.

  3. Andrew,

    Thank you for the nice comment, and thanks for sharing your experiences! I think I hinted at it above, but practitioners can and should learn from sharing and reflecting on their varied & valuable experiences. Good luck with the evolution of your methods as you navigate the course with the larger cohort. We’ve had some success (and a few learn-by-doing lessons) using Adobe’s Connect product for synchronous distance/disperse learning.

  4. Andrew Dalgliesh says:

    Hi John
    Thanks for this very interesting post. I’m about to work on the second stage of a Virtual PD project for teachers exploring various ICT tools in their classrooms. The whole project is facilitated online and in stage 1 we made some interesting observations.

    l. The rule of thirds for group commitment seemed to apply in our Project community – about 1/3 were regular contributors, 1/3 were occasional contributors but keen lurkers while 1/3 were not engaged much at all.
    2. We had a structured requirement involving participants posting an introduction in one forum, their project briefs in another and commenting on at least one colleague post in both. This got most of the group started but there was still a high early drop off .
    3. We got the best results when we nurtured the conversations – we’d make sure we personally replied to all posts in the initial threads and encourage participants to go reply to each other. We’d matchmake participants with similar project topics, feed them with links relevant to their topics, and if they emailed us with a question or comment we’d lean on them to post to the forum.

    Even though there were clear project “expectations” it took a lot of cross platform / media encouragement to get discussions going and maintain them.

    This year we’re scaling up to a much larger cohort but I think that will make it even more important to focus on the human touch. Last time we used Skype and other synchronous media mainly for l: l but this tIme I’m planning to use a lot more live and social media all the way through.