7. What are the principles that underpin scientific knowledge?
“The order in which knowledge has been acquired by the human race will be a good order for its acquisition by the individual.” George Polya
Many of my recent blog posts have been concerned with existing examples of teaching practice and why they fall short of what I believe we should be aspiring towards as teachers. In my last post, I sought to outline a positive vision of what it means for a student to know something, along with how the teacher can support this development.
In this piece, I aim to lay out the principles upon which we are building the science curriculum at Bobby Moore Academy. The overarching point is that to build a curriculum one must understand the logical structures that underpin the creation of knowledge in a discipline.
I will argue that in the sciences, (especially physics), knowledge is created in a three step process. This borrows partly from Peter Godfrey-Smith’s excellent introduction to the philosophy of science, Theory and Reality, and partly from Richard Feynman’s characterisation of the scientific method.
- Explanatory inference (guessing what’s happening)
- Deductive reasoning (computing the consequences)
- Inductive reasoning (checking and generalising)
To teach students science is to do more than equip them with a list of facts. It is to give them a glimpse of the reasoning process by which these facts were reached. The history of science will therefore be important in my account.
Questions to consider
A typical aim of a science curriculum is to teach students to ‘think like scientists’. What does this mean? How do scientists think? Is there just one way?
As teachers, we must take a view on these questions. We must be clear on the norms that govern our disciplines, and our students must learn how to abide by these norms in their thought and speech. For this to happen, we must also be aware of the rules of reasoning they currently follow, and present them with situations that call those rules into question.
Our understanding of the norms of the discipline is codified in the curriculum; our understanding — and challenging of — the students’ everyday view of the world is manifested in pedagogy. I have discussed in previous posts how teaching breaks down if either of these constituents is missing.
I have also argued that an itemised list of facts and ideas does not make up a curriculum, no matter how thoroughly it covers the content of an examination. This is because knowledge is not something we have, but something we are.
Knowledge is an orientation to the world; it is only displayed when a person responds to a specific, concrete situation (see Gilbert Ryle — The Concept of Mind). To have knowledge is therefore not only to be equipped with information; it is also to show judgment (I have borrowed this distinction from Michael Oakeshott).
A person with a thorough knowledge of science is not simply someone who wins the nature round of her local pub quiz. Instead, it might be argued, she is able to spot the pattern that links a range of observations, whether in the laboratory or at home.
Furthermore, she is able to explain them, often with the aid of some sort of visual image or model. The ability to visualise is key to feeling at home in the scientific landscape (more on this later).
This is not to say that the facts and ideas contained within a specification document are not part of the scientist’s arsenal. She undoubtedly has lots of substantive knowledge and implicitly abides by the rules we may call disciplinary knowledge.
It must however be remembered that these are abstractions; the author of these documents has looked at a person with a sound command of her domain, pulled out the premises upon which she draws various conclusions, and typed them up into a document.
It is a picture whose subject matter is drawn from real life; not an imagined ideal of what we are aiming for.
In other words, the process leaves us with some useful rules of thumb about what good scientists tend to know and do, but it doesn’t leave us with a good scientist.
Rules must, by their very nature, be formulated in advance, yet much of what is required to be successful in a discipline is manifested in unpredictable, spontaneous action. This leaves the teacher in a difficult situation. How can we teach students to respond to situations we cannot possibly predict?
The psychologists’ solution
Cognitive psychology takes a simple (perhaps simplistic) view of this question. Its proponents would argue that because this tacit, being element of knowledge is unpredictable and therefore uncodifiable, we might as well ignore it.
We can all agree, they might say, that good scientists have a decent body of substantive and disciplinary knowledge. We will never be able to agree what else they have (or are), because it is inherently unpredictable and unmeasurable.
Therefore we should stick to what we know and allow all those fuzzy, hard-to-describe traits to develop naturally. As I once heard an advocate of cognitive psychology say, “if you want to improve something, you have to be able to measure it.”
The problem with adopting this approach, especially in an education system with high stakes exams, is that a student can be very successful at the subject without ever building an understanding of the discipline. This problem is particularly acute in science, which often takes on the appearance of a body of unchanging facts about the world.
One can memorise all the facts, learn all the methods, and even recite the history, without having any understanding of the scientific enterprise. I know, because it happened to me. I was an expert at passing exams, yet when I reached university — and exam questions were no longer reducible to recipes — I found myself out of my depth and resentful of the whole sorry business.
What I failed to understand then, and am only just beginning to glimpse now, is the distinctly human aspect of science, the ability that the Newtons, Maxwells and Einsteins had to put their finger on the key problems and reformulate them in a way that brought a seemingly unintelligible complex of ideas together (see Arthur I. Miller: Insights of Genius).
This brings us back to the question of what it means to think like a scientist. We may not be able to predict what a person with good scientific judgment might do in a particular situation, but we can describe how they would be likely to go about it.
The architecture of knowledge
My initial premise was that the aim of a science curriculum is to teach the students to think like scientists. So far we have seen the limitations of a curriculum that takes the form of a list, no matter how granular the codification of knowledge within it.
The question to be addressed next is how we teach the judgment aspect of knowledge; how can we ensure that students leave our classrooms after five years ready to respond to the world in a scientific manner?
It seems to me that our first task is to examine the nature of scientific knowledge. How is knowledge created within the disciplines of science? The importance of this question relies on a second claim, that the manner in which knowledge is created within the disciplines is analogous to the manner in which it is created in our minds.
(The concept of mind is obviously open to debate; for the purposes of my argument, we can call it the brain, the individual, or the self.)
Under this view, the web of interlinking concepts the disciplines have constructed over hundreds of years is comparable to the intellectual architecture within which we reason and draw conclusions.
As we have seen, to know science is to operate within a framework of scientific concepts and norms, one which shapes our responses to concrete situations (see Jan Derry’s work on inferentialism). The clues towards how this framework may be erected lie within the history of the disciplines themselves.
Classical assumptions
In constructing a science curriculum, the most illuminating period of science I have come across has been the late 19th century and early 20th century in physics.
- That era, because (as Miller and others argue), this was a period in which knowledge was created with the explicit goal of visualising the real world, a goal which I think is more appropriate to school science than the abstract mathematical formalism that followed from quantum mechanics.
- Physics, because it is typically regarded as the purest manifestation of knowledge creation according to the ‘scientific method’ (see below).
The philosophy of science pursued by the greats of this era of classical physics (Einstein, Maxwell, Bohr, Kelvin, etc), was based on visualisation (see Miller, Olson — Scottish Philosophy and British Physics).
Humans are faced with enormous amounts of incoherent and seemingly unconnected data; the role of science, this view suggests, is to identify relations between those data and simplify those relations to as few principles as possible.
To identify the relations between our observations, we assume that every effect must have a cause, because we cannot imagine something happening absolutely instantaneously. This is why questions of cosmology and the early universe hold such fascination for us; why we struggle to imagine what came before the Big Bang.
We seek to simplify and arrange these causal relations for two reasons.
- Firstly, because it is unbearable to have too many; our minds simply cannot comprehend the vastness of nature unless we organise it somehow.
- Secondly, because there is an aesthetic aspect to science; the theory which unites the widest possible array of observations and ideas has an irresistible elegance and appeal.
The scientific method
Under this picture of science, the most common route by which new knowledge is created is neatly summarised (as always) by Richard Feynman. In his words, first we guess, then we compute the consequences, then we check by experiment.
If any experimental observation contradicts the consequences predicted by our guess, then the guess is wrong; we must begin again.
The technical terms for this process are given by Peter Godfrey-Smith as follows:
Explanatory inference — the guess.
This is the suggestion of a model or process which would explain the observations we have made so far, e.g. that all matter is made up of small particles, or that a luminiferous ether pervades the entire universe. This may be called a hypothesis.
Deduction — computing the consequences.
This is the process in which we start from a number of premises, and draw conclusions about what should be true. The philosophical example given by Godfrey-Smith is:
- Socrates is a man.
- Socrates is mortal.
- Therefore Socrates is mortal.
In scientific deduction, we use the hypothesis as one of the premises, and draw conclusions by working out what would happen to one measurable quantity if we were to change another, e.g.
- All matter is made up of small particles.
- Temperature is a measure of the particles’ kinetic energy.
- If we increase the temperature, the particles should bounce off the walls faster.
- Therefore the pressure should increase.
Induction — checking by experiment.
This is the process by which we generate laws in science (particularly in physics), that give us predictions for what will happen when we make measurements in the future. It is by induction that we move from particular observations to general laws.
We design experiments to check whether the consequences computed by deduction always hold true. For example:
- When I increase the temperature of a gas, the pressure always increases.
- Therefore pressure and temperature are proportional.
It should be noted that by moving to induction we no longer rely on the models or analogies that led to us doing the experiments in the first place. These models or analogies may therefore be thought of as ‘scaffolding’ which helps to construct the final edifice (see Olson, p284).
The ASE’s Big Ideas of Science — hypotheses like all matter is made up of small particles, or objects can affect other objects at a distance — may likewise be thought of as ‘scaffolding’ for the construction of conceptual frameworks in the classroom, a point of discussion in the next section. (See Adam Boxer’s work for further discussion of Big Ideas.)
Ultimately, induction highlights the causal relations between observations of the world, and the process of explanatory inference organises these in the minimum number of categories possible. The process of deduction is essential in moving the process forward.
The logical structures that underpin each stage of Feynman’s model therefore give us an insight into the modes of reasoning that any good scientist employs. It is in this latter sense that they may become useful in the classroom.
Knowledge creation vs recontextualisation
We have discussed a process by which new knowledge may be thought to be created in the disciplines of the sciences (at least in the classical, pre-quantum era), but it is obvious that what takes place in the classroom of a school is an entirely different engagement.
As sociologists like Basil Bernstein and Michael Young have argued, the disciplines are institutions in which knowledge is created; schools and school subjects are where this knowledge is ‘recontextualised’ and passed on.
We must therefore ask ourselves what relevance abstract terms like explanatory inference, deduction and induction have in the very concrete circumstances of a Tuesday morning with Year 9.
The crux of my argument is that if knowledge is taken to be something displayed in our ability to respond to the world using judgment (rather than simply having a store of information and rules formulated in advance), and if that knowledge is created via certain modes of reasoning (i.e. the logical structures which underpin a discipline), then the curriculum must be based upon these modes of reasoning.
In science then, all of the facts, rules, concepts, methods, arguments, experiments and historical episodes students learn about must be tied together by the notions of explanatory inference, deduction and induction.
To learn that pressure is proportional to temperature is not to memorise that proposition as true, it is to learn that first this was observed by experiment; later it and a number of other observations were brought together by the particle model; then the likely consequences of this model were calculated mathematically, and the theory was said to have been proved to be true when these consequences were confirmed by further experiments.
In other words, students should not simply be taught the facts; they should be taught how to reason their way to the facts.
This is not to say facts should be neglected: retrieval practice and other memory boosting techniques are essential if a student is to become a scientist. They are necessary, but not sufficient parts of understanding a discipline.
To truly be at home in the landscape of a domain, a student must recognise her subject as a distinctly human practice. A student should thus not only be able to answer the question ‘what is a wave?’; they should also be able to explain why anyone would want to ask that question in the first place.
How can this be achieved in the classroom? If we want to teach the wide range of facts and ideas contained within an exam specification, it is not possible for students to study the history of science in enough depth for these principles to emerge.
Instead, we must frame every lesson within the modes of reasoning inherent to the discipline. To do so requires — to use Bernstein’s phrase again — some recontextualisation. This is the role of the curriculum.
Recontextualising science
The terminology of abstract logic is clearly not suitable for all but a minority of school classrooms, yet the thinking which terms like explanatory inference, deduction and induction aim to denote is not beyond a teenager’s grasp.
The role of the curriculum is to carefully and systematically provide opportunities for students to apply such modes of reasoning in order that they begin to do so effortlessly. It is only once they have reached this stage that they can be said to ‘think like scientists’.
The role of pedagogy is to ensure these opportunities are suitably designed and presented in order for students to take advantage of them. Language, which may be thought of as a manifestation of thought (see Ryle on thinking as talking to oneself), provides a basis on which students can begin to think according to the relevant modes of reasoning.
I have summarised the types of sentences that might be thought to structure scientific reasoning in the table below (thanks to Miller for ‘as if’). (Rosalind Walker has also done lots of work on sentences as thought structures.)
Conclusion
By making explicit the types of sentences that constitute reasoning within a discipline, and providing frequent opportunities for students to think in this way (i.e. by saying or writing such sentences), we may move beyond an abstracted view of knowledge towards something approaching judgment.
The true test of success in this endeavour is whether students can indeed reason for themselves in unpredictable situations, which leads us to the question of assessment. I will save that one for another day.
The point I wish to emphasise is that by equipping students with the powers of explanatory inference, deduction and induction, they may take any set of circumstances or observations and reason their way to a conclusion, something no itemised list of facts could ever hope to achieve.
Curriculum can never tell students what to do in an unforeseeable situation; its only concern is how they might go about making sense of it. Put another way, true knowledge is not something we have; it is who we are. It both shapes and is manifested in our responses to the world. This is what makes it so powerful.
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?
