A critique of Downes’ Connectivism and defense of the PLE paradigm

Posted on Feb 3, 2010

While digging through my bookmark library for processing into Delicious and Netvibes, I came across the following from last year:


“Personal Learning Environments and Personal Learning Networks Call for Papers for an eBook by Athabasca University and the National Research Council of Canada”

There had been a conference last October, 2009 on the topic of PLE and other contextually relevant connectivity tools for use in education, though the website lists no (easy to find) resources for recorded video or audio of the proceedings. I will have to hunt it down elsewhere. The website had no feed, either (snicker!), and the discussions appeared to have petered out.

Among the conference speakers was Stephen Downes, who appears to be regarded as the expert, if not a flat-out advocate of the PLE paradigm. I examined some content on his site and download the following paper, “Learning Networks and Connective Knowledge”. The issue I downloaded is dated October 16, 2006 (note the emphasis of the now somewhat irrelevant MySpace, and nothing about Facebook – maybe it’s been updated?). Here are all of his publications.

My purpose in reading this piece was twofold: 1) I need someone to persuade me that this PLE thing is really worth all the trouble and complication it seems to cause; 2) Are there any communication theories that can impart value into the medium between sender/receiver within the PLE system that fundamentally shifts the quality of information towards greater resolution than the previous paradigm?

I should also mention that Mr. Downes does not present his work as didactic, or as a manifesto – he qualifies the writing as being still formative, though we do not know whether now, in 2010, he still feels the way he did in 2006 (unless I read everything else he’s done since then). So don’t take my somewhat summative statements here as though they represent his present stance.

Downes explains that prevailing Cognitive learning theory is errant in representing knowledge as some isomorphic “sentence” in the learner’s mind, as if propositions are real objects represented in a physical space in the brain. Instead, knowledge is expressed as a pattern of interconnected neurons, mapped over a variety of areas of the brain in such a way that imbues it with cross-experiential meaning, such as my definition of “Paris, France” meaning something different to someone else’s definition of “Paris, France”. The nuances in meaning are defined by the manner of connectivity – thus “knowledge is in the network”.

In Communication theory, there is a theoretical space between sender and receiver called the transactional distance, and communication is the act of transferring an object of knowledge from send to receiver. The channel through which this transfer takes place is the medium. If you believe, as Cognitivists would state, that knowledge is an object (or schema) of some sort, then it is plausible to transfer this object through the medium.

Downes rejects this model, referring, instead to a connection/emergence relationship between knowledge (information) and the mind.

Connection/emergence refers to the ability of the mind to functionally recognize emerging patterns that arise through the interaction in a network – a distributed representation, or meme (arguably defined as a unit or element of cultural ideas, symbols or practices). Knowledge, thus, does not exist in a place – it exists in the pattern of the network.

He states:

… More precisely, patterns of input phenomena – such as sensory perceptions – cause or create patterns of connections between neurons in the brain. These connections are associative – that is, connections between two neurons form when the two neurons are active at the same time, and weaken when they are inactive or active at different times.

Downes argues against the Cognitive schema object approach while at once advocating the neural scheme. Isn’t a Cognitive schema a pattern of interrelating principles/concepts/etc.? I guess the argument is more of a metaphysical one where Cognitivism functions in the network within a black box, whereas the Functional Connectionist would deny the existence of the boxes within the black box and only recognize the network. Downes argues against Cognitivism further (condensed below):

  1. Knowledge is subsymbolic. Mere possession of the words does not mean that there is knowledge; the possession of knowledge does not necessarily result in the possession of the words.
  2. Knowledge is distributed. There is no specific ‘mental entity’ that corresponds to the belief that ‘Paris is the capital of France’. What we call ‘knowledge’ is (an indistinguishable) pattern of connections between neurons.
  3. Knowledge is interconnected. The same neuron that is a part of ‘Paris is the capital of France’ might also be a part of ‘My dog is named Fred’.
  4. Knowledge is personal. Your belief that ‘Paris is the capital of France’ is quite literally different from my belief that ‘Paris is the capital of France’.
  5. What we call ‘knowledge’ (or ‘belief’, or ‘memory’) is an emergent phenomenon. Specifically, it is not ‘in’ the brain itself, or even ‘in’ the connections themselves, because there is no ‘canonical’ set of connections that corresponds with ‘Paris is the capital of France’. It is, rather (and carefully stated), a recognition of a pattern in a set of neural events (if we are introspecting) or behavioral events (if we are observing). We infer to mental contents the same way we watch Donald Duck on TV – we think we see something, but that something is not actually there – it’s just an organization of pixels.

The truth (his word) in Cognitivism or Functional Connectionism, he states, is proven in the empirical test of context. If meaning can only be stated in approximate terms (your meaning and my meaning overlapping in some form of majority alignment), then how might we be able to teach, and expect learners to learn, when what it is to know and what it is to teach is something that can neither be defined nor transferred? Thus, the reliance on the emerging patterns in the network.

Some thoughts:

We cannot assume to know why the brain behaves the way it does – only that it does, and that if it didn’t, it would fail to function in some way that has enabled us to survive as a species. To attempt to denote or define information by a system of relationships by connotation is risky, in my view: the result is not so much meaning but boundary formation – which reverts recursively back to human conceptions of “what is a pattern” or “how do I define the boundaries of a domain?”. Is there a point at which we can say, ontologically, the equivalent of “I think, therefore I am”, or more appropriately, “I am (in) the network, therefore I am”?

It is also philosophically difficult to say, on one hand, that we are imbued with the capacity to recognize patterns in networks, while also being the network. (One must be in the network, presumably, to be privileged to access to the network itself, right?). Wouldn’t this be grounds for a conflict of interest, to be at once in it and informed by it? I suppose this has been addressed in Bandura’s Social Learning theory as well – I’d have to look at it more.

Nor may we assume that the more connections on the network, or more neural nodes, then the “better” (or whatever way you want to define that – faster, deeper, efficient, related to other patterns, etc.). It may be, as the Frontline documentary “Digital Nation” asks, that the fewer connections the better, meaning that the brain has made itself so efficient that it need only absorb less neural resources (nodes) to emerge a particular pattern (such as automaticity). For all we know, neural mapping is an encumbrance on the brain, whose goal is to compress or simplify entropy, not expand it?

More later.

Main Menu