8. Teaching Explanatory Inference in the Classroom
The previous post was heavy on theory. It sought to identify the logical means by which knowledge is created within the scientific disciplines. In this piece, I will focus more on classroom practice: how can we bring students to understand the process of explanatory inference (or, in Feynman’s terms, the guess)?
The first point to notice is that making students memorise a definition of explanatory inference, e.g. the generation of a theory that best explains previously observed events, is likely to be pointless. (I happen to think any memorisation of abstract definitions is largely pointless, as I described in an earlier piece).
In fact, using the language of philosophy of science at all is likely to turn off all but a handful of students, which is why Feynman’s description of this stage as the guess is so valuable to the teacher.
Guesses in the history of science
Science has historically proceeded by great thinkers making guesses that tie together a range of disparate observations. The best way I have found to convey this idea is to explicitly teach the history of science, and make it clear exactly what was going on in each episode.
The example most familiar to science teachers is likely to be the development of the atomic model, a key part of the GCSE specification.
Many science teachers I have encountered see this part of the curriculum as a pain to teach. “The students just have to memorise the names of a load of dead scientists!” they have said to me.
In a sense, it’s true. Dalton, Thomson, Geiger, Marsden, Rutherford, Bohr, Chadwick — it’s a long roll call that the students are expected to work their way through.
Teaching students about people is not something that comes naturally to scientists, and for many the temptation to sit back and leave it to the students to learn themselves is irresistible.
Thus I have seen a lot of students copying down accounts of the development of the atomic model that have themselves been copied and pasted from the spec.
Alternatively, I have seen this part of the course called a ‘project’, in which the students ‘do their own research’ and copy the information down from a textbook or the internet instead.
Either way, such approaches overlook the fundamental importance of these developments. They are important because they capture the essence of the three stage process: guess, compute the consequences, check.
In particular, the series of guesses about what the atom looked like — billiard ball, plum pudding, solar system — encapsulates the process of explanatory inference.
Bohr, for example, used the large-scale metaphor of the solar system to understand what was going on inside the tiniest of particles.
His guess was that the atom behaves as if it is a tiny solar system, with electrons orbiting the nucleus like the planets orbit the Sun (see Arthur I. Miller: Insights of Genius for more on this).
Their illustrative power is almost certainly why they have been included in the specification. How should we teach these stories then?
Further guesses in the history of science
My first suggestion would be to introduce them to some other episodes in the history of science earlier in the curriculum. As Christine Counsell has long argued, when faced with challenging topics, we must avoid the temptation to drill students on specifics that appear on the exam specification. We should broaden, rather than narrow, what we teach.
Put another way, as the Learning Scientists would advocate, the best way to help students understand a complex process is to show them as many concrete examples as possible. Through this, we encourage them to identify the deep structure linking the examples together through comparison and questioning.
Thus, by the time students learn about the development of the atomic model, we will have covered the following episodes:
Early particle theories of matter
Corpuscular theories moved science away from the Aristotelian conception of matter — everything is constituted of four elements which tend towards their natural states — towards the modern, energy-based view — all matter behaves as if it is made of tiny particles whose energy determines the properties and behaviour of the overall substance.
Particle and wave theories of light
Newton’s corpuscular theory could explain certain properties of light (motion in straight lines, reflection) but not others (e.g. diffraction). It took Huygens to suggest that light behaves as if it is a wave oscillating through some invisible medium to move theories of light forward.
While this position later proved problematic (it turned out that the luminiferous aether does not exist), it is an important scaffold in helping students to understand the nature of light (see my last post for the characterisation of hypotheses as scaffolds).
Plate tectonics and continental drift
In the nineteenth century, it was thought that mountains, earthquakes and volcanoes were caused by the Earth shrinking as it cooled. This theory, however, did not explain why Africa and South America looked as though they once fit snugly together.
Wegener suggested that the Earth behaves as if its surface is split into lots of plates which move around and cause the events outlined above. His theory was largely ignored in his lifetime, but subsequent discoveries supported the idea that plate movement is caused by convection currents in the mantle. Since then, his theory has been widely accepted.
Geocentrism and planetary movement
Or, in other words, the insights and discoveries of Copernicus, Kepler, Galileo and Newton. I’m currently trying to educate myself on this period of scientific history — their ‘guesses’ will be a key part of the Forces & Space scheme we are soon to write.
Looking beyond the specification
By looking at a number of examples of guesses in the history of science before we come to look at the development of the atomic model in Year 9, I hope that students will have a sense of why Dalton, Thomson, Rutherford et al made their suggestions about what the atom might look like.
The example of plate tectonics is particularly important. We recently taught rocks and plate tectonics to our Year 8s. I have spoken to a couple of teachers since; they have told me that they no longer teach rocks because it’s not on the chemistry specification at GCSE.
This is exactly what Christine Counsell warns against. Whether it’s on the final exam or not, Wegener’s insight offers a beautiful example of how scientific knowledge develops. By seeing it as an example before arriving at the development of the atomic model, students are much more likely to understand that topic when they get there.
(I would also argue that the rock cycle makes for an excellent precursor to the more abstract carbon, water and nitrogen cycles — another reason to keep rocks in the KS3 curriculum. And, of course, the main reason they should be there is because geology is an inherently interesting subject which students have a right to know about.)
Examples like this illustrate where curriculum thinking — which scientific episodes are worth addressing — overlap with pedagogical thinking — repackaging explanatory inference as something comprehensible to students in the form of as if sentences.
In the next post, I will adopt a similar approach to explore how we can teach students the art of deductive reasoning.
This is part of a series on Developing a Science Curriculum at Bobby Moore Academy.
Previous articles:
1. What’s going wrong with our Year 11s?
2. Isn’t there already a curriculum?
3. Does anyone learn because they love learning?
4. Do we need a curriculum at all?
5. Do we need to worry about pedagogy at all?
6. Navigating the landscape of the domains
7. What are the principles that underpin scientific knowledge?
