Dave Cormier is a wonderfully sideways-thinking writer, such as in this recent discussion of the myth of learning styles. Dave’s post is not mainly about learning style theories, as such, but the nature and value of myth. As he puts it, myth is “a way we confront uncertainty” and the act of learning with others is, and must be, filled with uncertainty.
The fact that stuff doesn’t have to be true to be useful plays an important role in my latest book, too, and I have an explanation for that. The way I see it is that learning style theories are (not metaphorically but actually) technologies, that orchestrate observations about differences in ways people learn, to attempt to explain and predict differences in the effects of different methods of teaching. Most importantly, they are generative: they say how things should and shouldn’t be done. As such, they are components that we can assemble with other technologies that help people to learn. In fact, that is the only way they can be used: they make no sense without an instantiation. What matters is therefore not whether they make sense, but whether they can play a useful role in the whole assembly. Truth or falsehood doesn’t come into it, any more than, except metaphorically, it does for a computer or a car (is a computer true?). It is true that, if the phenomena that you are orchestrating happen to be the findings and predictions of science (or logic, for that matter) then how they are used often does matter. If you are building a bridge then your really want your calculations about stresses and loads to be pretty much correct. On the other hand, people built bridges long before such calculations were possible. Similarly, bows and arrows evolved to be highly optimized – as good as or better than modern engineering could produce – despite false causal reasoning. Learning styles are the same. You can use any number of objectively false or radically incomplete theories (and, given the many scores of such theories that have been developed, most of them are pretty much guaranteed to be one or both) but they can still result in better teaching.
For all that the whole is the only thing that really matters, sometimes the parts can be be positively harmful, to the point that they may render the whole harmful too. For instance, a pedagogy that involves physical violence or that uses threats/rewards of any kind (grades, say), will, at best, make it considerably harder to make the whole assembly work well. As Dave mentions, the same is true of telling people that they have a particular learning style. As long as you are just using the things to help to design or enact better learning experiences then they are quite harmless and might even be useful but, as soon as you tell learners they have a learning style then you have a whole lot of fixing to do.
If you are going to try to build a learning activity out of harmful parts then there must be other parts of the assembly that counter the harm. This is not unusual. The same is true of most if not all technologies. As Virilio put it, “when you invent the ship, you invent the shipwreck”. It’s the Faustian bargain that Postman spoke of: solving problems with a technology almost invariably creates new problems to be solved. This is part of the dynamic the leads to complexity in any technological system, from a jet engine to a bureaucracy. Technologies evolve to become more complex (partly) because we create counter-technologies to deal with the harm caused by them. You can take the bugs out of the machine, but the machine may, in assembly with others, itself be a bug, so the other parts must compensate for its limitations. It’s a dynamic process of reaching a metastable but never final state.
Unlike bows and arrows, there is no useful predictive science of teaching, though teaching can use scientific findings as parts of its assembly (at the very least because there are sciences of learning), just as there is no useful predictive science of art, though we can use scientific findings when making it. In both activities, we can also use stories, inventions, beliefs, values, and many other elements that have nothing to do with science or its findings. It can be done ‘badly’, in the sense of not conforming to whatever standards of perfection apply to any given technique that is part of the assembly, and it may still be a work of genius. What matters is whether the whole works out well.
At a more fundamental level, there can be no useful science of teaching (or of art) because the whole is non-ergodic. The number of possible states that could be visited vastly outnumber the number of states that can be visited by many, many orders of magnitude. Even if the universe were to continue for a trillion times the billions of years that it has already existed and it were a trillion times the size it seems to be now, they would almost certainly never repeat. What matters are the many, many acts of creation (including those of each individual learner) that constitute the whole. And the whole constantly evolves, each part building on, interacting with, incorporating, or replacing what came before, creating both path dependencies and new adjacent possible empty niches that deform the evolutionary landscape for everything in it. This is, in fact, one of the reasons that learning style theories are so hard to validate. There are innumerable other parts of the assembly that matter, most of which depend on the soft technique of those creating or enacting them that varies every time, just as you have probably never written your signature in precisely the same way twice. The implementation of different ways of teaching according to assumed learning styles can be done better or worse, too, so the chances of finding consistent effects are very limited. Even if any are found in a limited set of use cases (say, memorizing facts for a SAT), they cannot usefully predict future effects for any other use case. In fact, even if there were statistically significant effects across multiple contexts it would tell us little or nothing of value for this inherently novel context. However, like almost all attempts to research whether students, on average, learn better with or without [insert technology of interest here], on average there will most likely be no significant difference, because so many other technologies matter as much or more. There is no useful predictive science of teaching, because teaching is an assembly of technologies, and not only does the technique of an individual teacher matter, but also the soft technique of potentially thousands of other individuals who made contributions to the whole. It’s uncertain, and so we need myths to help make sense of our particular, never-to-be-repeated context. Truth doesn’t come into it.