In the second part of August I have been involved in two major events here at the University of Tartu. One was the on-site session of our Master’s programme in Educational Technology – two weeks of intense study with our new master students. And the other was our summer school on “Educational change in times of rapid technological innovation”. Eventually, there were roughly 40 people involved – teaching staff members included, and some 20 countries represented. These were two weeks where we had the chance to converse about education and technology, and try out a few gadgets in a rather convivial way. I will write more extensively about these 15 days. As for now I’d just like to point out a few bunch of ideas that came to my mind during this period.
It seems to me that, when we are dealing with the issue of technology in education, there are at least four levels of analysis that we should take into consideration. I list them here:
understanding how different technologies work;
operating them;
using them;
the role of institutional arrangements as another type of tool.
The first level covers everything that concerns what we may call “technology education”. This level contains, in other words, all “how” questions related to the very technologies we are using – our tools. Questions such as “how does a computer work?” or “how fiber optic cables transport information?”, etc.. , go into this category.
The second level of analysis pertains to how we operate our tools. The main difference is that operating a tool already involves interaction. So, at the very basic level, operating a tool means essentially being able to turn on and off a device. But it may also mean that we are familiar with the features of an application. For example, being able to use a webcam during a video call. Or record a video call and save it to one’s computer.
Operating a tool tells us how we should interact with a tool, but it does not tell us how we should use it. I believe that the distinction between using something and operating something is fundamental here. And the reason is that using a device is something more than merely operating it. So, I may be able to know how to post an article to facebook, but I may not know how to use a facebook post in one of my sociology classes where I talk about Max Weber. What is then the difference between operating something and using it? The short answer is that the difference lies in the fact that the use of something is always given in a certain context. Our tools don’t fall from above, but they are already involved in a dense network of meanings – culture, in other words. For example, when the television came about, it seemed to be natural to record a teacher speaking and send the recording to those schools that could not always have a teacher coming over because of their location (see the case of Samoa Islands). More in general, what usually happens is that a piece of technology does not change the way we practice something (e.g. teaching or learning). It is simply recruited to do the same very things that people are used to. We saw this phenomenon in more recent times with Moocs.
So, at a general level of abstraction, we may say that using a tool always happens in a specific context. In other words, it is situated. Therefore, we cannot talk about the use of a tool in abstraction, that is, without making an explicit reference to the context of a practice (what a practice is, I’d leave it to another post). So, for example, the so-called “features” of an application are something that we operate, but their actual meaning only becomes clear in their actual use. Yes, Facebook allows me to post an article or a learning material. But what is the actual meaning of posting an article in the very context of my practice as a teacher? That’s the sort of question that educational technology should care about.
I am now coming to the last point. If we understand that the use of a tool always happens in a specific context, then we should also turn our attention to another type of tool – a tool sui generis – which are our institutional arrangements – the education system, in short. That’s also part of educational technology. How can we talk about innovation in education without discussing merit and demerit of the broader institutional system with all its components (the curriculum, etc.). This is also part of the context I am talking about. Very often the main barrier to innovation in education is..the education system itself, how it is structured, the very goals that we have set, etc… So, what we see is that it might be the case that technology simply allows us to do more efficiently what should not do in the first place.
What has become clearer and clear is the idea that understanding is an activity in which we do one thing: (re-)contextualizing. In other words, understanding means putting something into context. And that is essentially conversational. We need to be in dialogue with something or somebody.
During a talk about educational data (I talked about here), I showed this image asking what it was:
Indeed, people recognized that there was a carpet (on the left). And there was a footprint of a shoe. But in general they were just puzzled.
Next, I showed this picture asking again the same question:
This time the people in audience were able to understand that the first picture was a picture of a gym. And that the blue thingy next to the carpet was a mattress. And yet they were still puzzled. What did I want to know from them?
After that I showed a third picture with a small animation, in which the image of a guy was popping out:
Only then, things started getting a bit clearer. There is a gym. On the floor there are mattresses, which people, when passing by, are not supposed to step on. They should walk along this narrow black carpet. Bingo!
One way of describing this sort of experiment of mine is this: essentially, my audience was puzzled, because they were not able to contextualize the information provided in the first picture. Or, to be more precise: they were not able to select an interpretative frame in the light of which to describe the first picture.
What I did afterwards was essentially to cue them. To cue means essentially to suggest or hint at something. But what is that I suggested? And how did that happen?
One way to put it is to say that I was hinting at the interpretative frame, which would allow them to contextualize the picture. It is important to stress that I was, as I said, hinting, alluding, suggesting. In other words, I was not telling them directly “use this frame”. That would have been giving them the answer. What I was doing was to progressively giving off some pieces of information – clues, which I assumed they would be able to account for by using the right frame. I was in other words helping them put the picture into context. At least, a possible context. Or, to put it another way (pun intended), I was enabling them to select/identify an interpretative frame. It is important to stress this, because very often we tend to think that problems are solved by simply adding “data” or “information”, when in fact data is useless, if we are not able to connect it to an interpretative frame. Such connection is vital.
As I mentioned before, the idea is that the identification, selection or even the creation of an interpretative frame is the product of a conversational situation. That is, a situation in which one is in dialogue with something or somebody else. In the example presented above, what I was doing was precisely to engage my audience in a conversation. In the conversation, it is worth noting, the two partners (my audience and myself) are essentially exchanging something that gets modified every time that it goes from one end to the other. So, I asked a question, I got back a few answers, which then triggered a response from my side, which, in turn, triggered a reaction from the other side, and so on and on. In doing so we were building on each other’s response. This can be described precisely as an activity in which contextualization and re-contextualization naturally occur.
I said that I was cuing the audience of my talk. But they also were cuing me. Meaning that the responses that they were giving me were clues as to what they had understood so far. I wanted to walk them to a particular place. And I needed to know where they were at every passage – are you following me?.
Admittedly, the whole thing was not open-ended. I wanted the example to have a specific meaning. It had to serve a certain purpose. Yet, the whole conversation – like any other conversation – was characterized by a certain degree of uncertainty, which is constitutive of this contextualizing and re-contextualizing I have just mentioned. While I knew what I wanted them to get to, I could only guess from the clues they gave off whether they were following me or not. As I said before, I could only assume what kind of clues would work. The guessing is a product of contextualization. The uncertainty here is vital, because when contextualizing I am going beyond what I know. In other words, it’s generative and therefore it can be otherwise.
I’d like to spend a few more words on a couple of passages that one can find in David Bohm’s book On Dialogue. At the very beginning of the book he clarifies the nature of the matter. In a dialogue,
when one person says something, the other person does not in general respond with exactly the same meaning as that seen by the first person. Rather, the meanings are only similar and not identical.
Then, he concludes,
when the second person replies, the first person sees a difference between what he meant to say and what the other person understood.
Seeing this “difference” is key to conversation. It is precisely what allows us to move forward.
This point is clarified, when Bohm extends the idea of conversing to inanimate objects. In doing so he asks a simple question about the work of artists.
Can it properly be said that the artist is expressing himself, i.e., literally “pushing
outward” something that is already formed inside of him?
To Bohm it cannot be the case that the artist simply implements an idea that is already formed inside of him. And he explains why.
what usually happens is that the first thing the artist does is only similar in certain ways to what he may have in mind. As in a conversation between two people, he sees the similarity and the difference, and from this perception something further emerges in his next action.
Bohm stresses that, as the artist pushes something outward, something else comes about outside of him, which then gets back to him in a different form. The crucial insight that Bohm provides is that the artist perceives a difference between what is yet to be articulated inside of him and what has been already taking shape outside of him, which offers now some kind of scaffolding structure. Interestingly, such scaffolding structure is temporary. It will last until the next move, which, in turn, will pull it down to erect a new one. And so on and on.
A wonderful example is provided by this video documenting the making of One vision, one the greatest hits of a certain band called Queen.
In this segment we have a perfect representation of what conversing means:
One says something, the other objects, which forces the other to put forward an alternative. In some other cases, they finish each other’s line.
This back and forth has something in common with tinkering, which can be described precisely as perceiving the differences that in subsequent iterations become generative of something new (a song or a lyric). It is capable of generating something new, precisely because we tinker with the perceived difference(s) – something that kind of stands out, that even may confuse or irritate us. As I explained in this blog, there is always a chance element, which is nothing but this something that gets back to us over which we literally have no control, that may may even surprise us. If we had it, we would not need to act. We would simply be content with what we have.
The example that Bohm makes is about the kind of work done by artists. But essentially this is a universal feature that holds water for any kind of medium that we are using, any kind of “tool”, be it pen&paper, technologies (including AI) and even language itself, etc. As soon as we start using a tool – any kind of tool, this very dynamic is potentially there. Incidentally, this means that the very act of distributing cognition is a conversational one, because what is established is precisely a conversation between us and the tool(s), which only then acquires a meaning. The conversing and the distributing are coupled together.
This that I am describing has very important consequences for learning (and education). Learning shares something crucial with the case that Bohm describes. To develop understanding – be that on nature, our machines, languages, society, culture or the human being, the learner is precisely like the artist who pushes outwards something in the hope that something will happen as a result – in the case of learning, I believe, it’s understanding. What is crucial then – what enables learning – is that the learner senses or perceives differences and similarities between what is out there and what is growing, say, “within himself” this understanding. That can only happen, when the learner participates in a conversational situation, which can involve a teacher, peers, teacher and peers, or/and all the above plus the tools. Sometimes the conversation can actually take place between the learner and a tool,which may come in the form of a book, a blog post, a search on the net, a discussion on social media, a podcast, etc.. The list is potentially as endless as our ability to establish conversational situations out of our will to learn. Indeed, in conversational situations the teacher may be at the same time the partner in a conversation. Or the person who actually helps the learner sense or perceive differences in the exchange. I would qualify that as a pedagogical act.
It is worth stressing that, if differences cannot be perceived and therefore used as a springboard for furthering the back-and-forth that I described above, then learning becomes some kind of indoctrination (for lack of a better term). This does not mean, though, that conversing implies giving up on what one thinks. This misconception has pressed us to think that, lecturers should not lecture any more. That would be imposing their own view. Actually, it’s the other way around: we can truly converse, if we share our opinions – no matter how incomplete or groundless they are, precisely because in a conversational situation they just provide the temporary scaffold to manipulate this sort of “prima materia”. For example, what the lecturer says can be used by the student(s) as a springboard rather than as a piece of content to be poured into his mind.
So, more in general, when conversational situations are established, learning can become convivial, which I define in opposition and opposite to learning as indoctrination.
I wrote in a previous post, that I was working on a paper with Tony Tonni – educational technologist at Eller Music College here in Tartu. The text is supposed to provide a case study, in which we illustrate how a few ideas coming from cybernetics (notably, from Stafford Beer’s work) can be used to understand the role of a educational technologist in an educational institution at different levels of abstraction – that is, practically and theoretically.
Later on my colleague Irene-Angelica Chounta joined in. The paper is now completed and has been submitted to a journal in the field of educational technology. The draft can be found here.
Proprioception can be viewed as some kind of self-reference mechanism that is built into the whole system to give the system itself a sense of self-awareness. Our body does have proprioception. Thanks to that, we know when we are moving a part of our body. Oliver Sacks in an essay in his well-known The man who mistook his wife for a hat describes a woman who had lost the awareness of her body. So, she could move her hand, but she didn’t know she was moving it. Our brain, conversely, lacks it. So, for example, a person suffering from Alzheimer does not have any “feeling” of that, simply because the brain can be aware that there is something wrong in other organs (e.g., the liver), but cannot be aware of its own problems. In several publications (notably, On dialogue and Thought as a System) David Bohm extends the absence of proprioception to thought. Thought too cannot be aware of its own shortcomings.
Michael Polanyi introduced in the 50s the idea that any act of cognition has a tacit component, which, while being a condition for that act, can never be made explicit. He makes a crucial distinction between focal awareness, which is what we are aware of, and subsidiary awareness, which is what allows us to perform various things without, though, being aware of the underlying mechanism. That remains tacit or unconscious – one’s tacit knowledge. For example, if you start paying attention to what you are saying during a public speech, you may easily get into troubles. While there is a system that takes care that your speech is going well, when you try to do that consciously, the whole system collapses. According to Polanyi, that is due to a switch of our subsidiary attention. If we want to gain conscious control over ourselves, the system “self” crashes, precisely because of the absence of proprioception: the controlling is the controlled. And vice-versa.
We may say that the whole system has a hole right in the middle.
It is worth noting that the absence of proprioception does not mean that any system is doomed to fail. A system may be working fairly well. The thing is that the whole notion of control needs to be drastically revised. Control is seen as something that requires to be outside of the system being control, when in fact there is no such a place. That would be the deusex machina.
One interesting thing we can look at in search for a different conception of “control” (or steering, to use the cybernetic vocabulary) is conversation. Conversation is the way in which we control a process without, though, deploying any kind of control other than the one emerging out of the ping-pong two or more parties are involved in, which is the trademark of a conversation. This is an example of what I am talking about:
In David Bohm’s own words:
Can it properly be said that the artist is expressing himself, i.e., literally “pushing outward” something that is already formed inside of him? Such a description is not in fact generally accurate or adequate. Rather, what usually happens is that the first thing the artist does is only similar in certain ways to what he may have in mind. As in a conversation between two people, he sees the similarity and the difference, and from this perception something further emerges in his next action. Thus, something new is continually created that is common to the artist and the material on which he is working. (On Dialogue, p. 3)
One way to put it is that systems lacking proprioception are essentially open systems, as they are continuously involved in a conversation with their surroundings. Conversation might be seen here as a synonym of learning, which can actually be seen as the proprioceptive structure for cognition.
The problem here is that we hold the view that learning is synonym to “being instructed”. Which is in a way the attempt to create proprioception for cognition by deploying the traditional notion of control. The teacher has full control over the learner, who is supposed to blindly execute the instructions. Here the teacher becomes the proprioceptive structure for the learner.
However, another way of looking at this is to think that the proprioceptive work – so to say – is actually done in a dialogical way, by means of conversation, for example, between learner and teacher. The control here is not exerted by a centralized entity, but in the interaction between the parties, which would provide the kind of feedback mechanism that is fundamental to transform if not create oneself. This is exactly what Bohm pointed out in the quote I have just referred to:
As in a conversation between two people, he sees the similarity and the difference, and from this perception something further emerges in his next action. Thus, something new is continually created.
Some weeks ago I went to the IKEA store in Riga, the capital of Latvia. I had my wife sitting next to me and my faithful mini tablet with the GPS on to guide us to our destination. On that day we were actually coming from Riga city center, where we had previously had a quick lunch with friends.
When I am using the GPS app, I usually try to visualize the route right before the start. I don’t try to visualize the details. On that day, a brief look at the route on my tablet allowed me to understand that to get to our destination, I would simply drive along Tartu road and then turn right after some 10k. Easy.
When we were about to reach the destination, I saw the big IKEA buildings some 100 metres away from us. There was a big roundabout and I decided that it was time to stop following the GPS. Indeed, I could have continued relying on the GPS. But I did not.
What was it that made me avoid that?
Essentially, as soon as the IKEA buildings camein sight, I could extract the information that I needed from the immediate environment, that is, when to turn, where to turn, etc.. In a rather interesting sense, what I was doing was to resort to the territory as the map of the territory itself. The IKEA buildings were there. And, in spite of the fact that there was nothing like a straight road to there, I could decide whether to turn right or left, or go straight, because I could fairly assess each option at every junction.
More in general, we may say that we turn to the GPS (pun intended), when we are not able to extract all the information we need from our immediate environment. I say “all the information we need”, because actually we can never completely obliterate our immediate environment. Why is it so? It is so because the GPS is sort of a map, which is supposed to help us precisely when the IKEA buildings – to use my example, is not in sight.
However, the GPS is not the territory. This precisely means that the GPS does not contain all other information that we are in need of in order to actually navigate all the way through to our destination (e.g., other cars, pedestrians crossing the road, traffic lights, the road itself, etc.). There are actually tragicomic stories of people who actually drove their car into the water, precisely because they simply stopped paying attention to what was going on in their immediate environment.
I think that the GPS story is sort of allegorical of many other things that are happening around Artificial Intelligence, and that is not just related to self-driving cars. Let’s see why.
Our cognition tends to act ecologically. Which is another way to say what I noted above: we tend to use the territory as a map of the territory itself. And we try to do that as much as we can. Why so? One of the reasons is that ta panta rhei kai ouden menei. Everything flows, and nothing stays. Even when we could in theory have a complete representation mapping everything of the territory, that would just be the analogous of a photo taken at a certain point in time: obsolete as soon as it is taken.
That our cognition is ecological means also that it has to massively rely on perception, which is defined as the capacity to extract information directly from our immediate environment – what is going on around us. Perception, as James Gibson noted, does not necessarily contain “pictures”, but also actions that we perceive as “allowed” – what he termed “affordances”. Perception, in turn, feeds into judgment (or discretion). That is, the ability to decide what to do in the here and now.
Now, Artificial Intelligence is a very powerful tool, which is to some extent analogous to the GPS, as I pointed out. AI can help us decide. It can supplement our ecological cognition by providing maps or representations of the world in which we are called to act. However, as we saw with the GPS, there is an essential tension between the kind of supporting maps, which AI can provide and the actual territory in which we operate as decision-makers. Being aware of that tension is the very first step towards sensible use.
