Interview Stephen Downes by Renée Filius

Renée Filius: [Introduction] So what I see at course evaluations is that teachers say: 'Well, when I teach online I tend to ask brief questions. Or students just give brief answers and leave again. There is not a lot of interaction going on and learning tends to stay at a surface level.' Do you recognize this problem?

Stephen Downes
: Well, I don't really give brief answers. I mean, I think I recognize the phenomenon that may cause the problem. It's really hard to get the minute by minute feedback that you get in a personal environment when you're responding to a question. So, whereas if you're talking to a person  you can judge by their facial reactions and that, whether you can just continue talking, but when you're online, you don't have those cues and at a certain point you lose track of that nonverbal communication you're having with the other person. So you stop talking sooner in order to get feedback. That's what I think. What I observe is the lack of the personal cues, you know, except in a video environment like this, where I can actually see you nodding, for example, and things like that. But in typical online communications, even synchronous communications, are usually audio only at best. You know, in environments like Illuminate or some of the others, unless you're doing person-to-person videoconferencing, which is difficult and rare in a classroom environment. So, yeah, but as you can tell, I don't stop myself, so I think the phenomenon may be as much a propriety of the person as it is a propriety of the environment.

Renée Filius: Yes. So, you just mentioned, when we talk about the difference between online learning and offline learning in a classroom, you mentioned the difference because of the lack of facial expressions or body language. What other differences do you see between online teaching and face-to-face teaching and how they can affect the learning results, the learning outcomes?

Stephen Downes: One of the really big things that I see, is the lack of a shared object. For example, when I'm working with somebody, I very often haul out this sheet of paper and start drawing. Because that's the way I think and that's the way I communicate. And when I'm online with someone I can't just haul out a sheet of paper and start drawing. Or, you know, even now, what you've just missed is, I made a hand gesture, like pulling up the sheet of paper and drawing, right, which you didn't see. Online there are shared whiteboards, but it's a different kind of experience using a shared whiteboard, than using a whiteboard in a room, for example. The whiteboard in the room, it's much larger in size, even a piece of paper is larger in size than what you have to use online. The implements are much easier to use, you can just draw. So , the scale of the visual display really limits the kind of interaction in sharing objects that you can have. That's the big thing I find.

Renée Filius: Yes, I understand. So that is a major drawback of online learning, you could say. Could you also see, like, an advantage of online learning, if you would compare it with offline learning?

Stephen Downes: Oh yeah, there's huge- Some of the advantages, bridging distance is one. We couldn't do this session without online learning, it's just not possible. Even with the time issues and that, still, it just simply would not be possible. So that's a huge, huge advantage. I'm doing this in my living room, I didn't have to leave the house. For a day like today, when we have three feet of snow on the ground, that's a huge advantage as well. So it's location independent. Also, there's more modalities. I mentioned paper, whiteboards and of course speaking and gesturing, but that's the limit to the modalities, well there's a few more, but that's pretty much the limit to the modalities face-to-face. Online I can share my desktop, I can open up an application, I can use screen sharing, you know, there are systems that allow me to take control of your computer and do things for you. I won't do that, because I'd need your permission, but do you know what I mean? There's a whole range of things, you know, we can simulate environments online, forming digital environments, generally that would be very dangerous offline, such as an airplane or a nuclear reactor or brain surgery. You know, we don't want people practicing on real airplanes, or brains, or things like that, or especially nuclear reactors. So, I think those are some of the affordances. Also, I think communication is different online from the very early days. People talk to the- How they communicate differently in an online environment than an offline environment. Face -to-face is really intimidating, especially for someone like me, oddly enough, and many other people as well. It's easier to try new thing, to say new things, to put on different identities, to be more expressive, in the online environment. Of course that leads to a weakness in the online environment: people don't feel so inhibited as they would face-to-face, and they start doing things like flaming and stalking and trash talking, and all this bad stuff that happens online too.

Renée Filius: Yes, so there's two sides of the coin.

Stephen Downes: Yes, two: advantage, disadvantage. But I've seen, you know, very open, very personal communications happen online that often wouldn't be possible offline, just because of cultural differences, location differences, personal differences, whatever. So it's a different kind of communication that becomes possible.

Renée Filius: Yes, I see. And when it comes to the use of feedback? When I see feedback, I mean, the feedback used by the teacher or the lecturer, but also peer feedback or canned feedback. How would you say- What would you say about the differences between feedback when it comes to online learning, how could we use feedback to promote deep learning?

Stephen Downes: One of the tendencies in the online environment is to make do with very quick  and easy to do feedback, because it's possible. In face-to-face, or even a classroom environment, there are no such things as a like, or a checkmark or a thumbs-up. And things like counting the number of followers is absurd in a person-to-person environment. No, we use that in an online environment, a lot of the time that substitutes for more traditional forms of feedback. Now, that's not a propriety of the technology per se, it's just the way we use technology. I think that the same kind of feedback that's possible mostly in a person-to-person environment is possible online, so whatever the professor does, for example, to stimulate deep learning offline, the professor can do in an online environment. The reason why I say that is, the bulk of that feedback consists of dialogue and conversation, like we're doing now, and as this conversation gives us dramatic evidence for it, we can do that in an online environment. So, if you were to offer a hypothesis or a methodology or something like that, the two of us could work through it, pull load, underline principles, expose assumptions and all of that, right in exactly the same way we did it offline, except my camera is such that you can't see my hand gestures, still. The main significant difference is in physical activities and skills, you know, things like neural surgery, where the feedback doesn't necessarily get transmitted through a computer screen. I've played around with vital feedback simulators so that the design of the equipment that you're on simulates the physical feedback you get, doing neural surgery. But something like that is expensive and it's also very domain specific, but you know, a lot of simulators are made for medical training and things like that. Flight simulation as well, they try to emulate the actual cockpit environment, even to the point of shaking the plane. But again, that's very expensive to do. But still, it's cheaper than a real airplane, quite a bit cheaper than a real airplane, because you don't require fuel. So there is that. So what the point here is, the physical feedback gets harder and harder and is less and less natural to create, than the audio conversation type feedback.

Renée Filius: Yes. Let's go back to the course evaluations. I noticed that a lot of the feedback that instructors give to their students is written feedback, it's not audio, it's not visual.

Stephen Downes: Yeah, and that's why brief hits hard for people to write. It takes time to write, you know I write 250 words in 15 minutes, is that right, that's about right, I don't know. But I can say 250 words in 3 minutes. So the actual time to compose and come up with feedback in an audio environment, audio-visual environment, is a lot faster. It's slower for you, interesting, because you could probably read faster than I can speak. I can read about 600 words a minute, so I'm writing at 250, I don't know how fast I'm speaking, but it's faster, but I'm still able to read a lot faster than I can listen to someone speak. Of course I can do other things while they're speaking, like check my email, watch Netflix. But that's probably why, if it's typed, it's probably going to be short. You know, I get requests to review things all the time and if I want to do anything like a decent review, it's gonna start consuming hours of my time and I don't have hours of time. A professor and a class, you think about it: 250 words an hour, so that's probably- did I say an hour? 250 words in 15 minutes, a 1000 words an hour. A 1000 words is something like two pages of text. If you expect two pages of feedback, which isn't a lot, on an essay, then if you have a 100 people, that's a 100 hours of work. And that's why you get one line. If you look at offline, if you look at the written feedback on an essay, say, it's the same. It's like little remarks here and there, they're not writing a page of text about the essay that was handed in. Although, they will be happy to have you come into their office and they'll speak about it for a bit, as it's easier. And I also know that most people won't do that. If they had to actually give verbal feedback to everyone in the class, they probably have a different view of how great it is. That's my feeling anyways. When I taught, I only ever spoke to a minority of the people in the class, not because I didn't want to, but I can't chase them down and only a minority came to see me.

Renée Filius: And there's another possible cause. I noticed that online teaching, like in the traditional classroom an instructor knows that he has to give a lecture between 13.00 and 15.00, so that's the time that he blocks in his agenda. But online it's different and after ten minutes he thinks: 'Well I have to finish this quickly, because I have to do so much other work and then I will go.' Is that something that you would recognize?

Stephen Downes: Yeah, of course I do everything that way. But no, there is a point to it, because I do my most focused thinking about a topic when I have to be there and presenting, especially in person, in a certain place and time. I actually prepare for that and the stuff I do just online, I prepare for a lot less. I just find I do it that way, even if I have a scheduled online time I prepare for it less. I think there is more pressure to doing it in person. If you're unprepared, you have to stand there and look them in the eyes and be unprepared. Nobody likes that feeling. We do it once or twice and you're over it, you don't do it again, you just make sure you're prepared.

Renée Filius
: No. And would there be any way of making this better for the instructors in order for them to give better feedback or to provide the students with better feedback on any way?

Stephen Downes: Well, raise the stakes? I don't know, I have to think about that. Because what we're talking about here, is the professor's own disinclination to be prepared and to provide the feedback, as opposed to anything structural. They could do the work, they could block the time, they could be prepared, they could be as on top of it as they are in person. But because they're not so much at personal risk, real or perceived, they have a tendency not to be. And the answer might be, even something simply physical, like bigger screens. If you're looking at a real sized version of the person, you might be more likely to think of them as a person, rather than as a computer artefact. You know, that's a speculation, but it's a possibility. We have a Cisco telepresence system in the office and basically it's a life-size high quality video representation of the person and they just sit across the table from you. And you know, you're pretty out of the wall, you know you're not gonna get away with checking your email while you're on that. By contrast, I could be checking my email with you right now, you'd have no idea. In fact, why don't I do that, you'll see how it looks. Here I am, I'm opening my email and I'm seeing something about a movie. You couldn't tell by looking at me, right, that I'm reading my email. I'm an expert at that.

Renée Filius: I can see the change in your glasses.

Stephen Downes: Yes, exactly. So, but in a virtual environment, I can get away with that. When we're talking like this, if I haven't prepared - and I've done this before, right - I have you on one side and I have my notes and stuff on the other side and I look stuff up as we're talking. Now that's both an advantage and a distraction. It's a distraction in the sense that it allows me to be less prepared, it allows me to be less focused. On the other hand, I have access to stuff I wouldn't if was standing in front of the classroom. In front of the classroom in the sense that you have no props, depending on the style of the person, but here in the computer world I can take much more like a disc jockey kind of approach to it. Haul by resources as I need them. And I do do that sometimes. Sometimes I do that even during presentations just for fun, but doing that live is riskier, it feels totally different.

Renée Filius: Would you expect any types of futures technologies or inventions that would help us in the future?

Stephen Downes
: Well, first I need to be convinced that it's a problem. Is the online experience so bad when compared to the in-person experience, that we need to make this extra effort to make the online experience better? That is to say: more like the personal experience. You know, we could, for example, build a room, just a little square room where a whole side of the room is a computer screen, but you don't have a mouse or a keyboard, or anything else in the room, just that screen. And you go into that room and the other person sees you, all of you, and you see the other person, all of the other person, and they're in their own room, and you have a really focused interaction, because there is nothing else to do. That would address the problem, that would fix it. Guaranteed. The question is: Is the problem so great that this is worth doing? Now that's a different issue. It might be, you know, it might be.

Renée Filius: Yes, but you could think of just small adjustments to the feedback process that enables the teacher to give feedback easier or that provides students an easy way to ask for feedback.

Stephen Downes: Well, I'm not sure it's a question that it being too hard to do though. I think it's a question of people not being inclined to do it. Do you see the difference?

Renée Filius: What exactly do you mean?

Stephen Downes: Well, the ease or difficulty of doing something is only one explanation of why something is done or not done. 'Why didn't I give Fred feedback? It was too hard. I don't like Fred. I didn't feel like it. I wasn't sure what to say. I didn't have time. I was washing my hair.' You see what I mean? There's a whole range of possible explanations and the difficulty of doing it is just one of those. Making it easier to do makes the feedback more likely and better, only if the difficulty was the cause of the problem in the first place. But how hard is it to give feedback online? Well, there's the typing thing, I get that, but you can also have a video conference like this, and those are very easy to do. Even easier than the one we have, because, you know, both people will show up on time.

Renée Filius: Those people don't get lost in a snowstorm.

Stephen Downes: Yeah. But you know, it's funny, one thing I have observed, and this is a good example of the counter example I'm giving, in the online environment, it has been observed that professors get many more requests for feedback and students show higher expectations of feedback and more immediate feedback. So a professor that in a class would speak to maybe ten people in a week, will get twenty emails in a week, or more, they get many more emails, cause emails are easier than going down to the professor's office. Well, maybe that's not the cause, but you know what I mean, right? So, because it's easier to give feedback in an online environment there's more expectation of it and in the end, it becomes harder. So, the way to make giving feedback easier might actually be to make it harder for them to do. Then they'd have fewer requests for it, then they'd be more inclined to treat those requests with more seriousness and give it more weight. So, now of course, that's the opposite of the effect that you want, although it's not the opposite of the effect that you want. It's a horribly confused situation now. What you're doing is, you're giving better feedback, but to fewer people, and that takes you right back to real world environment, where professors give better feedback, but to fewer people.

Renée Filius: Yes. If I may summarize you, you say when it comes down to writing feedback, it is more time-consuming for lecturers to write feedback, but it takes less time to read the feedback. And when it comes to giving feedback in general in online education, it is easier for students to ask for feedback, but it is more difficult for lecturers to provide feedback.

Stephen Downes: Exactly. Now what you could do, and I think some people play with this, is, you get the email in and you just say something back and send the recording back. That's easier than typing and you can be faster. But now you're still looking at a situation where professors are spending all of their time reading emails and answering them, or receiving inquiries and responding to them. I really think the volume of the requests is one of the key parameters here.

Renée Filius: But perhaps we could change that by making a better design of the course, of the education?

Stephen Downes: Well, this is part of the thinking of MOOCs. And here's what the thinking was: existing learning is very labour intensive for the professors. It's very labour intensive in the class, and as a result in a person-to-person class you can only have a certain number of people. You know, I've taught in some very large classes, but I know that I'm still only actually interacting with twenty or thirty of them. Even in the 150 student class, I'm still only actually interacting with twenty of thirty and the rest are what we would call online lurkers, which is okay. But the more learning becomes about the interaction, as it does in higher grade levels , the smaller the classes must become, by necessity. Because you can't get a graduate level education, say, simply by lurking. It doesn't work that way. You have to be completely engaged in the process, that's why it's so hard. It's not hard because the material is hard, it's hard because you have to dedicate yourself to it. But online, these constraints appear to disappear. Everyone thinks they have a personal relationship with the professor, no matter how big the course is. This is a problem. It's been problem because online courses very often have many students. It's just like offline, but you get desires who say: 'It takes no more effort to offer a course to eighty people than to twenty. So let's offer this online course to eighty people.' And the poor professor is drowned in email and the discussion posts. When we opened our course, so now we have an open online course. We're not dealing with eighty, we're dealing with two thousand people. In that case, I'm still only talking to twenty or thirty people personally, so the only way, absolutely the only way in such an environment,  is to remove the requirement that everybody talks to the professor in order to get this interaction and feedback. And so, to create what we would call a network based course, where the interaction of students among each other is an important and vital part of the learning experience. So, they get the feedback, not from the professor, but from other students. In an ideal world, and this actually happened, they get feedback from more experienced students.  This is what happened the second time we ran the same course. The second time we ran the same course, many of the people who took it the first time, came back for the second time and they picked up and led a lot of the interaction. They almost led the course. So, one of the things that we have tried to do when designing these MOOCs is to design it in such a way that it accommodates people at different levels in their professional development. So you can go into the course if you're new to the field and you'll get the basics and the few people who are slightly more knowledgeable will interact with you. Those slightly more knowledgeable people are interacting both with new people and with more experienced people, and they're learning more in depth, and so on. You know, it's kind of like the one-room classroom of old, right? Where part of the responsibility of being in grade five was teaching the grade two students. It's interesting, John Stuart Mill comments on that as all the biographies say. It's a great way to learn, teaching is a great way to learn. It's not necessarily, being taught is not necessarily itself a great way to learn. So the real learning happens in the teaching, which happens later on, which is probably why these people came back to the class.

Renée Filius: Yes, that was what I wanted to ask: Why would those people come back, the more experienced students? Is that because they know that by teaching, it's a great way to learn?

Stephen Downes: I don't know if they came back intending to teach. I think they came back because they were inherently interested in the subject and thought that they would get a deeper learning experience the second time through. Which they did, but one of the ways they did this was by talking about all of these concepts with the new people who started the second year.

Renée Filius: But wouldn't it make more sense if they would go to a new and different course on a more advanced level?

Stephen Downes: Well, I wouldn't think so, honestly, because a new course- It depends on how you view the subject. If you view the subject as linear, first you do A, then B, then C, and then you go up to P, and that's the end of the first course. And then the next course starts Q, R, S, T, U, so then, yeah, it makes more sense to just continue. But if the course is A, B, C, D, F the first year and then A, B, C, D, F, but in more detail the second year, then there's- You could just go back to the original class. And I think it's more like the second type, than the first. I think it's more a holistic, you get deeper and deeper, rather than you're doing more things and going further down the linear- following your story as it were. Mathematics works that way, it's presented as: 'First you do this, then you do this, then you do this.' But, mathematics, it's all the same thing always. You're doing the same thing when you do simple addition as when you do set theory, it's the same thing.  It's just set theory is addition, but understood at a much deeper level. So I think that- And then it's the simple thing: you learn better when you teach. So get you get this deeper set theoretical understanding, if you will, when you're trying to teach kids how to add, because you know, they make mistakes you've never dreamed of. That's what I discovered when I taught. My students made mistakes: 'How could you- What were you thinking?' You know, that sort of thing. But the thing is they  were looking at this very simple subject from this very weird perspective and to deeply understand the subject, you need to understand how you can see it from that perspective and what's wrong with that. So this is how you go from simple math, which you memorized, to set theory where you understand.

Renée Filius: But then, do I conclude correctly that it's that much as getting or receiving peer feedback that leads to deep learning, it's providing peer feedback that leads to deep learning?

Stephen Downes: Yes, that's a really good way of putting it. And I think that's true. Providing is much better than getting.

Renée Filius: That's very interesting. If I may ask- Can you think of other examples of feedback interventions that may lead to deeper learning that we haven't mentioned yet?

Stephen Downes: Other examples of feedback interventions. Well, there's music critics. No, I'm just kidding. Yeah, but that leads me to- Synchronous coproduction, no, I just made up that term, of the top of my head. But the example that I'm thinking of, as an instantiation of that, is a jazz band. And of course jazz band came from music critic, that's how my brain works, it goes from subject to subject and you don't know where it's gonna land. Think about how it works in a jazz band. You got, say, four or five players, each with a different instrument. What they're trying to do is put on a show for a crowd, and get paid at the end of the night, cause they're jazz players and they're poor. So they start playing, but they're not playing to a predefined tune, that's the old learning object hardly matters sort of way. They're improvising, but in a jazz band, you don't just go do your own thing, because the sound would be terrible and everybody would hate it. So in a jazz band, you work some common themes, you know, a common beat and a common key, etcetera. I don't know a lot about music, so I'm not sure exactly what they do. But then they begin to play off each other. So, one person, they're doing a certain melody and they'll vary the melody. And then the other person sees the variation in the melody and harmonizes with it. See what I mean?

Renée Filius: Yes.

Stephen Downes: Synchronous coproduction.  Another example is an article co-written in Google Docs. So if you're all working on the very same article and you're each writing bits, in Google Docs you can actually see the other person's cursor and the words pop up as they type. So again, you have a case of synchronous coproduction. So, one person writes something, the other person sees that and changes another paragraph or whatever. So you're not actually correcting or giving feedback directly to a person's work, but rather you're watching, responding to what other people do as you work together, engaged in a single project. That is a whole neat concept there. It's probably been written about, and I'm sure - we haven't invented it - but I'm sure it's a good way of doing feedback. I know it's a good way of doing feedback. It just shows up over and over again. You know, in a football team as the play develops, and all the players are interacting with each other and with the opposition and they're attempting to coproduce a goal. Literally in this case. You know, there's an example. Improvising comedy is very popular here in Canada and that's the same sort of thing as well, where they coproduce something funny. Well, something that's intended to be funny, it isn't always funny, cause it can't be easy sometimes, getting this from your improv.
[later added via Twitter: co-creation is a means of mediating between different visions, each adding and learning, the final form emerging, not pre-designed]

Renée Filius: But the thing is, you work on it together and you look at each other and you improvise and you watch each other and by doing so you choose your next steps.

Stephen Downes: Yes, exactly. So the DS106 course which is a MOOC run by Jim Groom, they don't coproduce so much, but they do have a lot of fun creating things and then looking at what each other has created. They don't go back and change their own creations as a result of this, but what you see somebody else do, influences what your gonna do next. So, you see somebody do a film noir photo and then when you do your video you think: 'Yeah, film noir, that'd be cool.' It's a little less overt than that, but you do see that interaction back and forth.

Renée Filius: Yes. And then I have another question. If I may ask, if I would ask you to formulate three statements or golden rules for providing feedback focused on deep learning, what would you say?

Stephen Downes: There are no rules. There are no rules. There are no rules. I don't do rules.

Renée Filius: And what about statements?

Stephen Downes: Similarly. By statement you mean axiom or principle, as opposed to: 'This text is black. Link text is often blue.' That's probably not what you mean. So you're looking for a generalisation of some type and I have a lot of difficulty with rules, principles, generalisations, because I think they're abstractions. I think that they can be useful in certain contexts. They're certainly useful for observing and reflecting on what you've done. You can identify a pattern in your own behaviour, but as prescriptions they're notoriously unreliable. All kinds of sadness and misery has been caused by somebody who is just following the rules, or just doing what you're supposed to do.

Renée Filius: I can see your point. Well, the reason that I ask you was just by trying to formulate a summary of what we just said. And one of those rules would be: Peer feedback is very valuable when it comes to deep learning.

Stephen Downes: Yeah, but, no, that's the other side of abstractions. They can be so abstract that they're not useful. Some feedback is better than none. Oh yeah, true, but not helpful. I'll give you my methodological principles. They're not really principles, don't treat them as such. They're not generalisations or categorizations either. Together I call them the semantic condition. You may have seen that in some of the stuff that I've written. There's four words: autonomy, diversity, openness and interactivity. And autonomy is the idea that it's better when entities - individuals or people or whatever - make their own decisions about their own objectives or goals, than having them determined for them. Diversity is, it's better to have many different things, than many of the same thing. Openness is, as the word suggests, it's better to be open to experiences, better to be open to sharing. The hard one is interactivity, because interactivity is the idea that knowledge is created by the interaction we have with each other, as opposed to something that is created and then transferred one person to the other.  Knowledge, in other words, is an emerging phenomenon and not an inherent phenomenon. So, I decided to call those the semantic principle and they're methodological principles that enable a network type of structure to be dynamic, that is to adapt and change and therefor learn. So, the more you embrace principles like that, in a network or in a system, the more that network or system is able to learn. Conversely, if you impose uniformity, if you impose obedience, etcetera, if you follow principles, the system's unable to learn, it's unable to adapt, it's unable to accept new input to change itself, to change its objectives, its goals, etcetera. Does that make sense?

Renée Filius: Yes, I see your point, yes.

Stephen Downes: That's the best I can do to answer that question.

Renée Filius: Yes, well, thank you. Okay, I have one last question and that is: Are there any other questions that you expected me to ask you and that I didn't ask?

Stephen Downes: Let's see, you covered the weather, so that was important. We haven't talked about the time difference. Sorry, I'm just kidding. I don't think so. You didn't ask me for a definition of deep learning and you didn't define it, that's interesting.

Renée Filius: Well, I did send you a definition of both by email before this conversation.

Stephen Downes: Oh, okay. There's the question that you didn't ask: 'Did you read the email?'

Renée Filius: That's a good one.

Stephen Downes: Because I almost never read the preparatory material for an interview. It's partly laziness and partly, well mostly, because I like to be surprised. That's what makes interviews fun. I don't expect the question and then on the spot I need to think of an answer, that's how we came up with, what was that? Synchronous-

Renée Filius: Synchronous coproduction? Co-creation?

Stephen Downes: Synchronous co-creation is it exactly. Never would have come up with that had I looked at the materials ahead of time and taking notes or whatever. Never would have come up with that.

Renée Filius: No, no, that's great. Well, thank you very much.

Comments

Popular Posts