In all of my efforts to participate in Connectivist MOOCS (#CritLit2010, #PLENK2010, #CCK11) I have run into a roadblock when discussion turned to “Learning Theories” and I have found myself unable to express (or perhaps even determine) what I want to say on this topic. My instinct is just to shout that the emperor has no clothes because none of the proposed “theories” make well-defined testable predictions, but I realize that this would be unduly dismissive of something that a lot of serious people take seriously.
Notwithstanding several helpful posts outlining the basic principles of the various “theories”, I can’t let go of those scare quotes because they don’t seem like true theories to me.
“Learning theories” as discussed by teachers of education are not scientific theories about how learning happens.
Yes, they do have value as (context dependent) foci of interest or beliefs about what matters in education, and as approaches or tools for analysing and engaging in its practice, but calling them “theories” is misleading.
Now, to be fair, not everyone does use the term “theories”. Some authors use “paradigm” instead for the more general frameworks (as here and here). But unfortunately, in the spirit of a discipline which introduces the term “manipulative” for something manipulable, some educational theorists seem to need to dress everything up with either pretentious (and poorly chosen) jargon or pseudo-scientific “just so stories” which “inform” their practice in a way that seems intended to create a sense of authority rather than just just honestly identifying that practice as focusing on whatever values they most care about. But aside from that propensity they are good people seriously interested in improving the lot of humanity, so perhaps now I can get past this mental block, and while interpreting their “theories” just as statements of value, move on to looking at their (and my own) practices in the light of those values.
Of course, the global sweep of the above paragraph is probably unfair, but I can’t help feeling that much of what I see of the field appears to involve dressing up the commonplace in unnecessarily formal attire.
In order to confirm or correct that impression, here is my take on each of several styles of paedagogical analysis with a view to helping me understand its status as a “theory”.
I should emphasize here that my knowledge of these things is very limited so what I say below may well be wrong. Please assist me by making corrective comments if you are able, and I will continue to edit this post in the light of what I learn.
Behaviourism is perhaps an anomaly here, as it is less about what should be valued as a goal of instruction than about how other “theories” should be judged – and even in that role it is largely misunderstood. Contrary to the apparent understanding of many of its opponents, the behaviourist point of view does not deny the “reality” or importance of higher level theoretical constructs but just asserts that in order to be meaningful they require an interpretation in principle in terms of observable behaviour (in much the same way that the theoretical concepts of physics are all really just abstractions of complex statements about relationships between length measurements – or more fundamentally “sense data”). There are of course learning contexts in which all that is required is induction of appropriate responses to specific stimuli, and for these it may be sufficient to impose an SRS reinforcement strategy of operant conditioning. Some educators may feel that this can also be applied to “higher level” learning outcomes by just using OC methods to induce the corresponding pattern of responses but this “radical behaviourist” approach ignores both the possibility that an SR characterization of “deep understanding” may be too complicated to describe (like doing thermodynamics by controlling the motion of individual molecules) and the fact that the student’s prior learning may provide much more efficient ways of influencing behaviour than conditioning it directly.
The problem with behaviourism is not that it’s false but that it is has, up to now, been trivially true. Pre-MRI there was no way to determine a person’s mental state other than observing his or her behavioural and physiological responses to stimuli (which could of course be quite complex – such as math exam problems or questions like “How are you feeling?”). And even with direct brain observation, what we are observing is still just another kind of physiological response so really nothing has yet changed in that regard. In future, perhaps direct observation of neural connections may allow us to “see” mental states without involving a stimulus of the organism, though of course as yet we are not able either to do that on sufficient scale nor to interpret the results if we could make the observations.
But in any case, behaviourism itself is not a theory of learning but rather a theory of observation and OC is usually identifed as little more than a technique – though it does have the potential of giving rise to a theory (which it would if it were elaborated to the extent of making predictions – as opposed to just observations – about rates of learning in various contexts).
I suppose also that the test of what I called the “radical behaviourist” claim (eg to see if using operant conditioning aimed at producing correct answers to calculus exam questions would succeed in producing students who could actually pass such an exam) would be a testable prediction and so might be called a scientific hypothesis, but a hypothesis is still not quite the same as a theory.
Cognitivism puts the focus on items of stored information, which from a behavioural perspective amount to S-R patterns in which certain questions produce specified answers. But since knowledge of a fact can be revealed in many ways there is not a one-to-one correspondence between the Q&A pairs and “pieces of knowledge”. There are doubtless some subjects which can be most efficiently summarised as a list of facts, but a “radical cognitivist” may go beyond that to assert that just learning facts is the best way to induce “higher level” understanding as well. As for behaviourism, cognitivism postulates a family of observables and its radical form makes a testable hypothesis, but again this is far less than is needed to qualify something that would be considered as a scientific theory in other contexts.
Conceptualism requires a higher level of learner “understanding” which often involves generalizing from a large collection of cognate facts to the abstraction of a higher level concept. A conceptualist attitude may suggest that explicitly drawing the learner’s attention to parallels is more efficient than just providing lots of examples, and a “radical conceptualist” may believe that encouraging the student to work with some abstractions before he or she is “ready” is an appropriate way to help develop the useful habit of looking for patterns rather than just waiting for them to become apparent.
Constructivism emphasizes the participation of the learner in actively putting together pieces of knowledge to generate or “construct” mental models which embody abstract concepts.
Social Constructivism values in particular the role of interpersonal interactions in the construction of mental concepts.
Connectivism appears to claim that what we should attend to, more than the state of the student as an individual, is that of his or her connections to a social network – perhaps with the idea that this will induce or reflect (or obviate the need for?) personal knowledge and understanding. Connectivists also often note analogies between the social network and the neural network (which, in a scientifically reductive theory of learning, presumably underlies all of our mental associations between “concepts”) and sometimes they even appear to be considering these quite distinct networks as directly related but Stephen Downes has recently emphasized their independence.