[Revised transcript of a talk for the AAUP Annual Meeting in Washington DC, June 2000.]

Online Pedagogy with Discussion Management Software

Andrew Feenberg

The purpose of this paper is to introduce a new type of software that does not yet exist, but which I hope will soon be in general use in online education. Although there is a technical aspect to this presentation, my main concerns are pedagogical. The software is designed to encourage and support a specific type of pedagogy which I have identified in earlier articles as most appropriate in the context of online education (Feenberg 1987, Feenberg 1999a). This is a Socratic pedagogy based on structured discussion led by a teacher in an asynchronous computer conference or discussion forum. The software I describe addresses two pedagogical problems that have arisen around this approach to online education: first, the lack of a base of experience and widely shared guidelines for practice often results in very poor use of discussion forums, and second, there are real technical obstacles to making the best use of forums. Appropriately designed software can suggest an effective way of accomplishing a goal, in this case successful online discussions, and facilitate the technically difficult tasks required for success.

To begin with, I'd like to tell you about my early experiences in online education in the course of which I developed my ideas on pedagogy and the first inklings of what I am now calling "discussion management" software. These early experiences were in fact the first experiences that anybody had in the field. I was already teaching online in 1982, long before the World Wide Web and the current fascination with online education. In those days I was working at the Western Behavioral Sciences Institute (WBSI) in La Jolla, California, where we started the first school that delivered its educational program on line (Feenberg, 1993).

This was a school for busy executives who couldn't get away from their work for more than a few weeks out of the year. We invited them for two one-week seminars in La Jolla each year, where we introduced them to the computer and to their online teachers. We used computer conferencing for the oneline portion of the program. This is a type of software which resembles a newsgroup or a discussion forum such as you may have seen in Blackboard or Web CT. But we were fairly limited by the lack of graphics and sound in the computers of that time. We could not sketch or write on an equivalent of a blackboard or show any pictures, much less film, recordings or slides.

There were even more serious problems with pedagogy. Normally one learns to play dominant roles through occupying the corresponding subordinate roles. You learn how to be a parent by being a child, how to be a teacher by being a student. But if you've never been a student in a classroom of this new type, it's difficult to teach in it. And of course no one had ever learned online before so none of the teachers knew from their earlier experience as students what to do. Although our teachers were for the most part well established professors from major universities who had rarely if ever given a thought to pedagogy before, we talked a lot about it in learning to deal with a new medium.

We had to solve a very basic problem: how to organize a class around written interactions instead of spoken ones. One discovers new constraints and opportunities in shifting from the medium of speech to the medium of writing. For example, it is impossible to "lecture" online; one might as well mail a written text to the students. Why complicate their lives with computers? Under this constraint, the equivalent of classroom discussion must serve to deliver the course. It turns out that strong leadership is necessary to get good results with online discussion. Why? Because in an unfamiliar situation, without direction, students remain isolated in their corners, unable to take initiative. The teacher has to start out by clearly defining the goals and norms of the conference because students don't have a common background to draw on. They don't even know such things as how long a reasonable comment should be or whether they must write formally and spell correctly or simply set out their thoughts. The teacher must also set an agenda and define expections, for example, regarding online participation. To implement that agenda, the teacher must provide topic raisers that are clear and thorough. You can't just toss out a simple question like "What did you think of last week's reading" because no one will answer. It's hard enough to get answers to a question like that in a face-to-face classroom where the students can see your eyes burning into them and demanding a response, but it's impossible online. To get an interesting discussion going, the teacher has to actually write a little essay, a page or two of text, to explain the subject and the angle to take in responding.

We also discovered that once an interesting discussion gets going, it can serve as a valuable learning resource. Lots of interesting ideas come out and if the teacher, or perhaps a student, takes the time and trouble to summarize what's been said, conceptualizing the different contributions and the differences of opinion that have emerged, that can be a tremendous learning experience for the class. We called notes or comments that summarize the different threads of the discussion "weaving comments," and they became central to our pedagogy at WBSI (Feenberg, 1987).

These are the beginnings, the very earliest beginnings of online pedagogy. But our equipment was very simple and lacked many facilities we would have liked to have. As I mentioned earlier, we had no graphic capabilities at all, no techniques of presentation and contextualization, no blackboard, no movies or slides, no handouts, no online background readings. Of course we could mail these things to the students, but we could not put more than a page or two of text online and expect it to be read. Our discussions were handicapped by the difficulty of reviewing the conference archive and constructing weaving comments. I recall many sessions with printouts and magic markers! We longed for what we called a "moderators' interface" that would support the pedagogical efforts of the online instructor, although we had only vague ideas at first and no way of implementing them.

In 1986 I published a paper in which I speculated that perhaps videodiscs could be used to solve these problems (Feenberg, 1986). At the time there was some excitement around this nearly forgotten technology. SONY introduced a computer with a built in videodisc player. George Lucas created a research foundation to explore the use of videodiscs in education. I proposed that we create videodiscs incorporating both graphical materials such as films, hypertextual access to related background readings, and communications software linking the user to an online conference. The conference would then be contextualized by the materials contained on the videodisc. So far as I know, no one took up this suggestion and in any case videodiscs were not the success that we had hoped for. Instead, ten years later the World Wide Web accomplished something quite similar and became the means of realizing the potential of online education.

Current online educational software solves the problems of presentation and contextualization we confronted in the early 1980s. In the typical contemporary web-based software, a whole series of presentation techniques are available. Typically, there is a frame along the side of the active window: click on this button and you get the assignment, click on that one and you get the reading list; click on this other one and you get background materials, and so on. This is the equivalent of handing out a syllabus, writing on the blackboard, showing slides and movies.

Discussion forums are still available and are still used in software such as Blackboard and Web CT, but surprisingly they have not evolved significantly since 1982. If you look at a discussion forum today you will find that in many ways it is actually more primitive than the software we were using nearly 20 years ago. We had functionalities on the Electronic Information Exchange System (called "EIES") that included access to a status list on which we could find the number of the last comment read by each participant, we could delete comments and messages before they were read if we changed our mind about whether to submit them, we could easily print out any sequence of comments that interested us, at any time we could collect statistics on participants' online contributions, and, with the terminal interface we devised at WBSI, we could download the entire conference, adding unread items day by day to create a mirror of the online materials on our local computers, reply to comments off-line, and sign on and upload our replies with a single command. Some or all of these functionalities are missing from current software.

There is clearly a new imbalance here. Where we could only write messages to each other, now the emphasis has shifted in the opposite direction, to the presentation of canned materials. Today when teachers talk about "course conversion" they seem to mean putting a lot of material on the web for students to read and study. Usually they are thinking about applying the presentation techniques available in the software.

In some utopian projections there's no need for discussion at all; online education is conceived as an automated system based entirely on presentation techniques which by themselves are supposed to be able to deliver the course content. In this utopia there are no teachers, only "star performers." But I do not believe that this is going to happen, no matter how many venture capitalists hope so. Some training programs will be successfully automated--you can learn to use Microsoft Word this way--but not English literature or physics. Student-teacher and student to student interaction is still the essence of education, and that is not going to change because of the World Wide Web. Just as speech is the backbone of classroom education, and presentations supplement it, so online, writing will continue to be the backbone of the educational experience and there too presentations will serve only as supplements, not replacements for the human contacts essential to learning (Feenberg, 1999b).

But if discussion is still the backbone of online education, why is there so little emphasis placed on it? This is understandable given the way in which people are being introduced to online education today through programs which emphasize presentation techniques. This is the aspect of online education that has been technically mastered so far. We know how to put things in html format and how to create software that facilitates that process. And of course trainers focus on helping teachers use the elaborate presentation techniques in the programs. When the teachers finally get around to delivering the class, they discover that while they have beautiful presentations, they have rather primitive discussion tools. They often don't know how to use these tools very well and find online discussion very time consuming.

Many online courses fail to take full advantage of the potential of the medium, and when researchers study online education, they often draw conclusions from classes that are offered by inexperienced teachers equipped with impoverished tools. The results are not as good as they might be. What can we do about this situation? First we need to identify the problems with existing software, and then figure out how to orient and enhance online discussion. I will call this general domain of problems and solutions "discussion management."

Conventional newsgroup and web-based discussion software was not developed for online education, and it shows. The programs have many defects in that context. Composing and reading at the same time is very difficult. In newsgroups, you compose in a little box that opens below the message to which you want to reply. You can't see other messages without closing the box and losing its contents. What is more, the box offers only a very primitive word processor, and if there is a problem on the phone line or with the host computer, you lose your work. It's also difficult to use the computer to quote from others' texts in the body of your response.

There are problems with keywording. Everyone is expected to keyword their own contribution, and that keyword is then supposed to be helpful to the group in looking back over its past discussions and finding items of interest. But authors' keywording is often unreliable or even incompetent. Who takes the time to outline a note in advance in order to identify every significant issue for keywording? Standardizing keywords in a group is difficult, and sometimes we change our mind in the middle of a comment about what to say without changing the keyword. In newsgroups, you often find that half the keywords have "re:" in front of them so everything is classified twice, one under its own first letter and again under the letter "r." This is hardly the sophisticated and transparent software we need for education.

Finally, teachers are at a particular disadvantage in online discussion since they must spend a great deal of time typing in text and are offered no software support for reusing materials in future iterations of the course. Of course this is technically doable already. The teacher can copy each note that he or she wants to reuse to the clipboard, then copy it from the clipboard to a word processor and save it. The process must be reversed the following semester. But this is not very practical and the structure of the available software doesn't encourage it. It's just a possibility among thousands of others offered by the computer.

I would like now to show you some slides which illustrate solutions to these problems. (After viewing the slides, to return to this text, use your browser's "back" function, not the arrow icons on the slides.) The first slide shows a way of improving composition. The interface contains two scrolling windows, one for reading and one for writing. Users can look back and forth at the materials on which they want to comment as they write. They should also be able to select passages from the reading window and drag and drop them into the composing window. That would make it easy to comment on a quotation from the messages to which they are responding. Finally, the user should be able to send messages from the composition window. In other words, writing in a discussion forum should be a lot more like using a good e-mail client such as Eudora where one has special facilities for quoting and commenting.

The next slide addresses problems with keywording. I call the technique illustrated here "active reading." In active reading, keywording is not the responsibility of the author of the text but of the reader. As they read, readers mark what interests them, for example, exciting ideas or material relevant to the test. The reader can keyword the text for future review much more effectively than the authors.

But if this shift in the keywording role is to work, it must be convenient. Just as it's easy to underline sentences in a book with a pencil or write notes in the margin, it ought to be equally easy to do the equivalent operation online. I suggest that we dock a window containing keywords alongside the main browser window. The keywords can be supplied by the teacher and supplemented by the student. To mark a text, the user double clicks in the reading window on a word at the beginning of it and clicks on a keyword. To show that the text has been marked, a yellow square pops up, a little like a post-it note next to the selected word. The user can open that note and see the keyword it contains, and write further comments on the text in it.

This technique is used to create hypertext stacks of related materials. When the time comes to review materials that have been keyworded in this way, for example, to prepare a weaving comment, the user double clicks on a keyword to start a review session. The review works its way back through the conference to show all passages marked by the chosen keyword, starting with the most recent. More complicated or less used operations, such as printing keyworded passages, saving them as a file, reviewing them in different orders, are available on a right click menu.

The last slide addresses the problem of re-using text. A second docking window has been added on the right hand side for this purpose. It contains a kind of virtual file cabinet in which texts can be stored for reuse. I suggest that it be possible to open it as both a file box containing file cards and as a tree of files displayed in Explorer format. The Explorer view is illustrated. The user can drag and drop items--whole comments or passages--from the reading and the composition windows into this file cabinet. There the items can be labeled to make it easy to find them for later use. For example, the teacher might decide to use the same topic raisers in the next session of the online course. Each week as he or she writes them, they are sent not only to the conference but also dropped into the file cabinet under "Topics" and labeled Week 1, Week 2, Week 3, etc. When the course is offered again, the teacher can drag and drop them to the composition window, saving a tremendous amount of time.

This points to one of the unexploited advantages of online education: at least some of what the teacher says can be re-used. In contrast, lectures require the teachers' presence and it takes the same amount of time no matter how often they are repeated. But online teachers can save some time through re-use, which makes up for the extra time they have learned they have to spend on the computer to have good discussions.

Let me conclude now. Software is not just a tool. It also shapes behavior. It is true that you don't have to do what the software makes easy for you and hints that you ought to do, but most people do follow the lead of their software. If the software makes it easy to write weaving comments and to re-use text, these activities are likely to occur far more often. If there are no facilities for weaving or re-using text, then it's much less likely to happen. Thus we can use software to quickly guide teachers toward an appropriate online pedagogy. Programs such as Blackboard and Web CT have already done that as far as presentation and contextualization are concerned. The software helps teachers see right away that they have to offer a syllabus on line and it gives them a good way to do that. Now we need to do the same for the discussion forum. We must provide teachers with software that guides them toward an effective online pedagogy. That's what I've tried to outline here for you today.


Feenberg, Andrew (1986). "Network Design," IEEE Transactions on Professional Communication, vol. PC-29, no. 1.

Feenberg, Andrew (1987). "Computer Conferencing and the Humanities," Instructional Science, vol. 16.

Feenberg, Andrew (1993). "Building a Global Network: The WBSI Experience," in L. Harasim, ed., Global Networks: Computerizing the International Community. Cambridge, MA: MIT Press.

Feenberg, Andrew (1999a)."Reflections on the Distance Learning Controversy," The Canadian Journal of Distance Education, vol. 24.

Feenberg, Andrew (1999b)."Wither Educational Technology?" Peer Review, Summer 1999.

© 2000, All Rights Reserved