[Options] [The Daily Archives] [Feeds] [Posts] [Places]
In this presentation I describe how new technologies are being designed in order to adapt to a rapidly changing and complex world. In this picture, learning is not a matter of knowledge acquisition and memorization, but rather, personal development in such a way that we become able to navigate and work with this knowledge and content in the same manner, with the same capacities, as the expert. This is a type of learning that is based on the creation of tools that support a mesh network; this network, in turn, supports the conversation and interactions that foster the development of an expert character. Slides and audio here Stephen Downes, Stephen's Web, November 25, 2009. [Link] [Tags: none] [Previous][Next]
Hi Stephen
A very interesting and thought provoking presentation. I have a few points where our views differ. One is that you seem to think that the structure of networks is flat, that there are no hierarchies, and that people can build their mesh networks in an environment where all are equal. I disagree with that as it is mathematically not possible to have non-hierarchical networks (see 'Linked: How Everything Is Connected to Everything Else and What It Means' by Barabasi). There are also power relations on networks, and in that respect they don't differ much from groups. It seems naive to believe that there are no political or commercial pressures and sociological influences on networks as anywhere else in society.
You seem to believe that learning is an individual process. There is a huge body of research evidence to support the view that learning is a process that has also social dimensions. This is where in my views groups are more interesting for learning than networks (as has also been established by others), as it is not only the artifacts, information and resources that create a stimulus for thinking, but it is especially the interaction with people, and I mean real human beings with emotions, rather than their messages that engage people in their learning; inside formal education, but also outside education in informal learning situations, for instance within families or tribes.
[Comment]
[Permalink]
[Previous][Next]
> I disagree with that as it is mathematically not possible to have non-hierarchical networks (see 'Linked: How Everything Is Connected to Everything Else and What It Means' by Barabasi).
I don't agree. Barabasi's diagrams presuppose (a) that the networks are scale-free, ie., that they operate the same no matter how large they get, and (b) each node links only once, and (c) preferential attachment.
Adjusting any of these can change a power-law network into a mesh network.
(a) applying a scale-based 'governor' on the creation of new links prevents the creation of single super-powerful nodes. This is what we see in nature, where physical constraints limit the number of connections per nodes. In economics, structural constraints, such as taxes, also limit the growth of supersize nodes.
(b) promoting divesity in linking can also create mesh networks. I didn't discuss the diagrams in detail, but if you follow the reference on the slide page you'll find a wortk by Daniel Lemire, Seb Paquet and myself demonstrating mathematically that increasing diversity of links limits the growth of supersize nodes.
(c) while random graphs will produce some clustering, the large power law effects are driven by preferential attachment, such that links tend to form in places where links already exist. Measures which reduce preferential attachment (eg., anonymizing or randomizing link selection) reduce the growth of supersize nodes.
Therefore, it is far from the case that it is mathematically impossible to have non-hierarchal networks.
> There are also power relations on networks, and in that respect they don't differ much from groups.
I agree that group-link structures form on networks. In a sense, a group is a type of network, albeit one that I would lable extreme and dysfunctional.
And I agree, that power-law networks and groups are functionally and structurally very similar. They represent a type of network structure that has become ossified, so as to exclude non-members. The autonomy (and therefore, communicative capacity) has largely been subsumed under the power node. Such networks do not typically generate new knowledge, but rather, repeat whatever message is emanated frm the power node.
> It seems naive to believe that there are no political or commercial pressures and sociological influences on networks as anywhere else in society.
I am not saying there are no such pressures. I am saying that such pressures are in the main largely negative and destructive.
> You seem to believe that learning is an individual process. There is a huge body of research evidence to support the view that learning is a process that has also social dimensions.
It is interesting to see my own characterization of learning - as a conversation, as an establishment of networks of connections with individuals and environments - depicted as an "individual process".
We need to draw this line with care. While people often draw a distinction between 'individual' and 'group', my own distinction is between 'group' and 'network'. These are two ways of interacting with community, rather than a distinction between interacting or not interacting with community.
I have tried to characterize these two ways as 'groups' and 'networks' but the distinction can also be drawn along a dimension of personal autonomy. While interacting with a community, one may either be subsumed or not subsumed under that community. It is this distinction that is at the heart of my argument.
I believe it is possible - and essential - to be able to interact, to have conversations, to participate, in community, without thereby aligning yourself with the beliefs, values, objectives and interests of the community. Conformity - except on very basic, syntactical levels - is not a requirement for community. Indeed, as I have suggested before, the greater the requirement for conformity, the more dysfunctional the community.
When we look at the huge body of evidence, we need to look at it with these distinctions in mind. It is overly simplicstic to say 'learning is social'. We need to ask, "what are the conditions of this sociality?" What are the elements - other peers, objects, artifacts? What is the structure of that sociality? How autonomously may a member behave? What degree of conformity is required for minimal functionality?
> it is not only the artifacts, information and resources that create a stimulus for thinking, but it is especially the interaction with people, and I mean real human beings with emotions, rather than their messages that engage people in their learning
This is a distinction I do not think you can sustain (though there is no lack of effort in the field of education to maintain such a distinction).
The reason I say this I, you have no access to humqand *except* through their artifacts. If you look at my 'speaking in lolcats' paper, you will see that I regard this as a rich and complex set of artifacts. Nonetheless, it remains the case that any access we have to "real human beings with emotions" is an *inference*, a way of seeing the world, a perspective we acquire rather than a perception we experience.
There are argukents that the fostering of this perception is necessary in order to develop one's own identity, to engage in community, to acquire positive values such as empathy, and to learn about a culture or a discipline. Terry Anderson has characterized this as 'presence', and the idea is that fostering a sense of presence improves learning.
There are arguments that may be made for this - and these underlie his own accounts of groups, collectives and networks. But from my perspective, there is nothing which precludes the possibility of developing a sense of social presence within a community that operates as a network, as opposed to a community that operates as a group. Indeed, I would argue that social presence - and especially the feeling of one's *own* presence - is *stronger* in a network than in a group.
Moreover, I have never argued that we should do entirely without groups. I have argued (for example, here: http://www.downes.ca/post/38554 ) that the emotional sense of belonging that we get from groups is very important. And it is certainly the case that such emotional attachments can motivate our desire to perform and to learn.
But this need for emotional attachment is abused. There are countless cases through history where powerful figures have manipulated group processes, toying with people's need to belong and imposed their own presence over that of any individual member, in order to pursue aims and objectives that were almost always destructive of the individuals within the group. When the community becomes simply a mechanism to amplify the presence of one person, it loses its cognitive capacity.
If you are attempting to impose something on people - such as a curriculum or company policy or doctrine or creed - then the mechanisms of the group will be your first choice, because you want to create presence in the group, and encourage individuals to subsume themselves to the group will (as instantiated by that presence). But where the development of cognitive capacity is more important - where the purpose is the development of the greatest capacities of both individual and community, *without* attempting to predetermine what the outcome of such a process may be, then you want to encourage the properties of a successful network: autonomy, diversity, oppeness, and interactivity. [Comment]
[Permalink]
[Previous][Next]