what's in a name ; networks vs groups

I like the idea of connectivism for a theory of learning.
I can accept knowledge is not an out-there thing, that it is held in nodes, constructed in connections.
That the connecting and means of connection have influence and are worthy of focus.
BUT I continue to have a problem with the narrow conception and responses to ongoing dialogues regarding groups and networks.
An integration of actor-network theory (ANT) would provide responsiveness to the continued critiques defended in arguments of well if it doesnt do this it was a group not a network, to which there is no answer.
Does the one do cooperation vs this is collaboration, aligning vs intersecting? I'm not convinced, the real world is so much messier than that. The attributions are not stable.
A deeper understanding of how groups/networks function is needed.
Its the function that matters, not the name. The attributes of a group vs a network seem currently focused on their instrumental nature. It's to narrow a conception to say groups do this, networks do that, as this fails to account for shifts.
The rhizomatic understanding is useful as it can map the shifts, but is less useful in explaining their nature.
ANT provides for a deep understanding of shifts, whether fluid or fractional.
ANT also allows for this to be explored, retrospectively or in process.
Moving the debate away from group vs network as entities of fixed attribution could further add to the development of connectivism as a learning theory by providing insights into how to work the net as well as demonstrating connectivism in action; of learning as something worked in the nets.
Suggested authors for further reading:
Bruno Latour
John Law