Why dual coding is difficult to get right in history (or English, or a whole host of other subjects)
In my last blog, I argued that dual coding works in science because pictures make explicit the logical connections between scientific concepts. Scientists develop models in order to visualise a phenomenon before making deductions that can be tested by experiment. Simple pictures, arranged in a way that captures the deductive process, can thus be highly effective in making these logical connections manifest to students.
Many have jumped on the idea that combining pictures with text must be a Good Thing in all subjects, even though the evidence in favour of dual coding typically refers to teaching science. Hence, we’ve ended up with teachers of history and other subjects scratching their heads and asking: “how does dual coding apply to me?”
Should history teachers draw illustrations of Henry VIII and his six wives? Should English teachers sketch a tableau of Macbeth losing his cool at the dinner table? How many loaves of bread do RS teachers need to draw when explaining the feeding of the five thousand?
Clearly, these questions are absurd. The reason for the confusion, I will argue, is that the logical connections between concepts in these subjects are of a completely different nature from those in science.
Dual coding must be thought of as a pedagogical tool. It forms an essential part of a science teacher’s toolkit but is not always the best tool for the job in other subjects.
Dual coding in history
In her blog ‘Examples of Dual Coding in the Classroom’, Kate Jones gives a number of examples of how teachers and students can build on the ideas of the Learning Scientists mentioned in my previous blog. These examples give some idea of why dual coding is less useful in history than in science. Many of the diagrams takes the form of ‘concept maps’, with pictures connected by arrows and captioned by text.
We face the same two problems that we saw with the concept map in physics. Firstly, the pictures don’t seem to offer much more than illustration. In no way do they help scaffold students’ disciplinary thinking; they do not shed much light on concepts and connections that students couldn’t visualise already.
Secondly, the connections between events and ideas are represented as arrows. While this gives a bit more of an inclination as to how things are connected, as we shall see in the next section, they only do so in a very limited way.
Historical reasoning — first and second order concepts
Whereas the Learning Scientists argued that “any visual image of the information” is likely to boost memory, I would argue that such diagrams are of limited utility as they fail to capture the peculiar logical structures of historical knowledge.
What are those structures? For the purposes of my argument, I will use the taxonomy of first and second order concepts, sometimes referred to as substantive and disciplinary knowledge.
First order concepts are made up of the ideas and vocabulary students need to be aware of (alliances, empire, democracy). Second order concepts govern the relationships between first order concepts; they include cause and consequence, continuity and change, similarity and difference, significance and interpretation.
In science, we saw how the content of a diagram represents the model being used, while the layout of the diagram captures the deductive process. Scientific reasoning ascribes causes — “if x is true and we do y, then we would expect z to happen” — and the 2x2 grid gives a good representation of this chain of causal reasoning.
Historical reasoning does not ascribe causes to things. An historical argument that attempted to explain a complex event in causal terms would be a bad historical argument. Instead, it employs the second order concepts mentioned above.
It attempts to describe change, and the extent to which change is gradual and continuous or sudden and discontinuous: “was historical event x the result of evolution or revolution?”. It ascribes significance: “this piece of evidence should be taken more seriously than that piece of evidence because …” And it involves interpretations: “this person said x, but, given y, what they really meant was z”.
Logical layout
Evidently, historical knowledge is made up of very different logical structures to the models and deductions characteristic of the sciences. The easiest of these to represent diagrammatically is continuity and change. Trivially, we could draw a timeline to show how one event proceeded after another, but this would be of relatively limited value: most students intuitively understand such sequential relationships anyway.
(The arrows in the example shared above function in this way. Yes — they represent sequential change, but it’s likely students would be able to visualise this even without the arrows present.)
A richer way of representing continuity and change in history might be by sketching graphs. This way we could show the pace of change and highlight any discontinuities (a revolution, a natural disaster, an economic crash). This would be a useful example of dual coding because it makes manifest the logical connections between events, in terms of a second order concept characteristic of historical reasoning.
When it comes to cause and consequence, we could potentially adopt the 2x2 grid model to show how one event led to another event, and how the outcome might have been different if the first event had not happened. There is a danger, however, that we lead students into the trap of ascribing singular, direct causes to historical events of the kind that are used in science. This would be to make a category error.
I’m not a history teacher, so I have only a limited idea of how we could incorporate the second order concepts of significance or similarity and difference into the layout of a diagram. (Larger images for more significant pieces of evidence? Side by side pictures to make comparisons?) And it’s even harder to imagine how we could capture the concept of interpretation in diagrammatic form. (History teachers — I’d be interested to hear your thoughts.)
Dual coding as a pedagogical tool
The point is that where the very nature of scientific knowledge makes it easily representable in diagrammatic form, the nature of historical knowledge means that dual coding is much less straightforward if you’re a history teacher.
The same could be said for English. How could the notion of texts talking to one another be captured within the layout of a diagram? Likewise art and design or music. Discerning between different styles and themes is much more subtle than the scientific pursuit of achieving absolute conceptual clarity.
As stated previously, dual coding is a pedagogical tool. It is useful in science because science deals in images of the world and proceeds from those images deductively. The layout of a good diagram makes a chain of causal reasoning manifest to students. The same could be said when using diagrams in certain topics in maths or geography.
It is less useful in the arts and humanities. In those branches of knowledge, the logical connections between concepts are very different. I will leave the detailed articulation of the nature of those connections to teachers with more expertise than me, but I’m pretty confident they are much less representable in diagrammatic form.
As we shall find out in my next blog, a more effective way to make these connections explicit in the humanities is through careful modelling of academic writing. If this is done effectively, teachers can scaffold historical or other forms of reasoning in much the same way science teachers do using pictures.
