As I sometimes find hard to locate the threads I will place an extract here and ponder about it during the week.
If you wish you can visit Stephen Downes original post.
"I have advanced a position in my own work proposing four major principles of association:
*resemblance - a.k.a. Hebbian associationism
*contiguity or proximity - a.k.a. salience
*feedback or back propagation
*balance, or entropy a.k.a. Boltzmann mechanisms
Conclusion
A lot of what I've tried to argue in this (admittedly long) post is that Connectivism is a non-intentional theory of learning and knowledge.
What this means is that, in connectivism, learning is not about content. It is not about entering a certain representational state with respect to the world.
Such an account makes the representational state - rather than our actual thoughts, perceptions and feelings - the arbiter of what it is to know, what it is to have learned.
But connectivist learning does not require representational states. It does not require on the part of the learner that they commit to a particular account of the external world.
It allows - indeed, encourages - the idea that people may have different, and individual, accounts of the external world.
Which means that what is negotiated is not some set of statements about the nature of that world - not representational states, not meanings - but mechanisms for communication, protocols for interaction (which may indeed be, and probably are, understood differently by each person engaged in communication)."
To add to this confusion I stumbled into Ken's blog, it did not clarify my ideas but I had a good laugh which I welcomed.
Will I someday be able to understand the riddles?
Time will tell.
