Frances Bell’s Blog

moar funny pictures

I have been a sporadic participant in CCK09 (which seems pretty quiet anyway) but I was up late working on a bid on Thursday and came across the Elluminate session via Twitter. I was paying partial attention while I finished off my bid and I found the last 10 minutes quite interesting, see recording.

I asked the following question in the text channel:

“Regarding power and authority - I would like to ask if and how George’s and Stephen’s ideas about connectivism have been influenced by CCK08/09?”

At that time Stephen was talking about a “neuronal’ view of networks with talk about nodes and firing.  A little later he responded to my question to say that what he had just been talking about was one of the main things he had learned from CCK08/09.  I have just gone back and re-listened to what he said , and here is my rough synopsis (but you can go back and check for yourselves in the recording)

Stephen was talking about small changes in networks that can be easily defined and can have impacts on network configuration.  As an aside, it interested me that Stephen didn’t sound slightly cross as he has the last few times I have tried to engage him in dialogue, and I found that quite encouraging. Stephen then went on to say that discussions about ‘big’ things (or the groups/network thing with which I have disagreed in the past)  become laden with contextual factors and he recommended that we talk about the ’small changes’ - the probability functions of one node connecting to another causing it to fire - to make the discussion different (better I guess).

At this point I had a partial dawning of what might be happening in the (non-)dialogue between myself and Stephen.  I am now going to give my interpretation of this (that will almost certainly be flawed) and would welcome comments from others, Stephen included.

This strikes me as an essentialist approach, this attempt to strip the discussion of context, to try to discuss the behaviours of networks that contain as some of their nodes the complex things that are human beings which have complex relationships with other (possibly complex) things and people in their networks.  Furthermore, doing this might make sense if we were using network as a metaphor to understand connections between people and things, and could then discuss how behaviours were ‘not like a network’ but it seems to me that the line is blurred between these networks of people and things are like networks, and they ARE networks.

I find it impossible to decontextualise such a discussion, there is much more to our behaviours and interactions than probabilities of firing (though I can see that automated network analysis, visualisations and even recommender systems can bring evidence to the discussion).  So I have to admit that the attempt to conflate the behaviour of neural networks with more complex ones is one that I resist, and maybe that is limiting my understanding.  I would be happier then, if we stepped back and looked at what are the advantages and disadvantages of treating learning between humans and non-humans as a network of connections that may or may not fire.

What does it reveal?

What does it conceal?

Anyway, George says he will anser my question next week so we can have some more discussion.