Rhizomes and networks

This week, for #change11, Dave Cormier is facilitating a discussion on rhizomatic learning. I’ve been aware of Dave’s thinking on this topic since his article in Innovate (.pdf). I was the guest editor of this journal issue and Dave’s article is the one that generated the most interest and still continues to resonate with many people. It certainly resonates with me.

One aspect of Cormier’s work on rhizomatic learning that he has not fully resolved is how rhizomes are related to networks. Periodically, he expresses rhizomes as an alternative to the network model. For example, in his discussion on edtechweekly, he suggests that networks are too structured and lack the organic attributes of rhizomes. On other occasions, he has raised concerns about the knowledge or epistemological underpinnings of networked learning (connectivism). I sense a lingering discomfort with networks in Cormier’s view of knowledge and learning. He hasn’t tackled this discomfort directly. Perhaps he is trying to be polite. Or perhaps he’s still thinking through the nature of that discomfort. Personally, I’d like to see him explore in more detail where the networked learning model fails in contrast with rhizomes.

Rhizomes are appealing for several reasons. First, they share the decentralized attributes of networks. There is no centre. Second, rhizomes are organic, they’re living and adaptive. A rhizomatic structure today will be different from a structure that will exist in a few months. Third, each part of a rhizomatic structure is capable of producing a new plant, propagating and replicating itself without central control. Fourth, in contrast with an architected structure such as a road system, rhizomes are not artificially bounded. They continually grow, extend, and develop. It’s easy to see how the decentralized, organic, adaptive, self-replicating, and unbounded characteristics of rhizomes are appealing as metaphors for learning and knowledge.

I’m not sure whether Cormier views rhizomes as a metaphor for learning or whether he is bolder in his claims. Does he think that learning is *like* a rhizome? Or is he making a bolder statement in saying that learning *is* a rhizome? In the blog post I referenced above, he states that rhizomes are “a way of thinking about learning”. If this assertion is central in his work, then I largely agree with Cormier.

However, when rhizomes are considered in contrast with networks, I find the rhizome model begins to lose some appeal. The greatest weakness I see with the rhizome model is that rhizomes do nothing new. They only make more of what already exists. And they can only make more of themselves. This is the opposite of diversity – it is extreme monoversity (it’s not a word, I checked, but it works for me). While rhizomes are diverse in shape and structure – growing, adapting, sprouting, replicating – they are not diverse in substance – i.e. rhizomes do not morph into new organic entities.

In his edtechweekly podcast, Cormier criticizes networks as being “duplicatable”. If someone has a successful network, she is tempted to say “this is how you create your own network”. Suddenly, the network becomes mechanistic. Cormier doesn’t like that. Neither do I. However, networks need not be designed in order to duplicate structure. Networks are organic – consider food webs, ecosystems, and the architecture of the brain. I don’t accept the argument that rhizomes are organic and that networks are not.

Cormier doesn’t seem to appreciate the descriptive attributes of networks – i.e. that we can describe certain network attributes by formula – power laws, propagation of influence, in-degrees and out-degrees. Personally, I think the descriptive and analytic capacities that have been applied to networks are tremendously positive. From the outbreak of cholera to the structure of terrorist cells to the development of the internet, being able to understand, often mathematically, the structure of networks enables us to react to and even build on that structure. It is because we can measure and understand how messages move through networks, the role of hubs, the need for amplification of messages, etc., that we are able to build the internet. The “definiteness” of networks is what gives them their relevance and power in our society today.

Together with Stephen Downes, Cormier and I have had long running discussions about objectivism, subjectivism, knowledge, learning, and a raft of other related concepts. I’m of that rare breed that still believes structure can be good and that some level of objectivity can exist in some situations. Inevitably, when this conversation begins, I find myself arguing with Downes and Cormier. It has never been resolved. And I don’t think it ever will be.

My network view of knowledge is simple: entities (broadly defined as well, anything: people, a chemical substance, information, etc) have attributes. When entities are connected to other entities, different attributes will be activated based on the structure of those connections and the nature of other entities that are being connected. This fluidity of attribute activation appears to be subjective, but in reality, is the contextual activation of the attributes of entities based on how they are related to other entities. Knowledge then is literally the connections that occur between entities.

I don’t see networks as a metaphor for learning and knowledge. I see learning and knowledge as networks. In global, digital, distributed, and complex settings, a networked model of learning and knowledge is critical. Most disciplines in society have become too specialized to function in isolation. Global problems are too intractable to be tackled by any structure other than networks. Generalists have given way to connected specialization (as evidenced in the identification process of the corona virus (SARS)). Everything – form fixing my car, to my morning coffee, to my research, to my mobile phone, to healthcare – is a function of connected specialization. Novelty and innovation arises when we collide ideas or specialties that previously had not been brought in relation to one another.

Much of the world that we live in today can be explained through networks, including education, learning, research, and knowledge development. I have not read Deleuze, though it is on my agenda for my trip to Croatia this weekend, but I don’t see rhizomes as possessing a similar capacity (to networks) to generate insight into learning, innovation, and complexity. Terry Anderson describes a sense of alienation with the rhizomatic learning and states that: “If we really want to CHANGE systems, we have to insure that we don’t grow as rhizomes, reproducing clones of ourselves or establishing gardens in which only certain types of weeds can flourish.”

Rhizomes then, are effective for describing the structure and form of knowledge and learning – bumpy, lumpy, organic, and adaptive. But they fail to describe how learning occurs, how novelty happens, and how a rhizome becomes more than a replication of itself. Rhizomes can be a helpful way to think about curriculum, to think about how we develop educational content when we are connected (dang networks again) to one another and to information sources. However, beyond the value of describing the form of curriculum as decentralized, adaptive, and organic, I’m unsure what rhizomes contribute to knowledge and learning.

6 Responses to “Rhizomes and networks”

  1. Nate Angell says:

    Definitely read Deleuze/Guattari on rhizomes! It will be worth the effort…

  2. dave cormier says:

    Hi George. Lots here to respond to and i’ll bring it in to the broader discussion on my last summative post for the week… but i’d like to address one point that i think is central to where we differ.

    You speak of ‘entities’. As best i can tell from the way you describe and entity, you think of them as fully formed things that can be pointed to and the recombined with other things. You see them as having defined attributes that ‘activate’ on recombination. This position is, i think, Platonic. I read these as forms/ideals that are essential to their being an entity.

    I don’t think these ‘entities’ exist in this way. (no, i’m not saying that the world is full of mush) I think what you refer to are blackboxes (thanks Latour) that hid power structures and support existing ways of looking at things. It’s the tidiness that I see in your version of networks.

    The rhizomatic metaphor could easily be a metaphor for the SHAPE of actual knowledge. I’m not sure… that’s a big part of the thinking that i need to do coming out of this week. I’m definitely not saying it CAN’T be… rather… i’m not sure.

    Thanks for the thoughtful post. Need more time to think about it.

  3. Lawrie says:

    Hi George,
    I just want to disagree with the last paragraph. Rhizome (growth) does effectively, as you say, describe the formation of knowledge and possibly learning. But the tone with regards to curricula is almost dismissive. If we could develop a model of curriculum development based on Cormier’s ideas then it would have the potential to be truly adaptable to student needs, possibly genuine personalised learning experiences, where “just in time” resources and serendipitous network connections can be made depending on the context of the student. The point with rhizomes growth is that it is there below the surface, looking for the right time, the right conditions to flourish, just like learning. Curricula could do worse than to aspire to develop and adapt in the same way.

  4. Jaap says:

    I will read Deleuze, because I suspect some romantic (he is after all a French Philosopher with connections to romantic philosophers like Bergson and Nietzsche) views in his thoughts. I will read the books with this question in my mind.
    The romantic view could explain why the rhizomes / network discussion seems to be endless. (different world views in the background) and why ‘organic’ is used as a value and a bit of a negative value is set on ‘analytic’ .

  5. Asif says:

    Coming in from left field to offer a perspective on this.

    Perhaps rhizomes and networks represent two different types of learning.

    Rhizome learning results in replication of concepts whereas networked learning represents production of concepts.

    In an organization, the head honcho’s views/talking points will tend to be replicated by the underlings — i’ll say ‘rhizomatically’ — so you’ll see the same views/points articulated from different mouths at different levels down the hierarchy.

    At the same time, in the same organization, people at different levels will connect, communicate and learn from one another’s perspectives in the larger organizational network — and this will result in the production of new views in each one of the people connecting (a combination of their pre-existing views plus the new information they’ve received).

    So things like propaganda or peer pressure will function on rhizomatic learning, whereas things like open source communities or collective decisionmaking will function on networked learning.

    That’s not to say that the two are mutually exclusive or that it’s even possible to strip one away from the other.

    In terms of relationships between the two, I’ll use the following analogy: if the ecosystem is a network, then the potato plant is a rhizome within that network.

    So I’m proposing that networks can contain rhizomes. The question I’m asking myself now is: can rhizomes contain networks?

    Maybe it’s a question of scale.

    • Asif says:

      Me coming in to respond to my own comment (…sad, I know).

      Been thinking: if the Occupy movement is rhizomatic, in that it has popped up in many locations based on people learning about what’s happening in other locations (ie. it is self-replicating [at the moment]), then the connections being made by people within a single location (ie. Occupy Toronto) and the information shared between those people, according to the model I’m presenting above, represent the formation of networks and a process of networked learning.

      All that to say — to answer my question above — that yes, rhizomes can contain networks too.