Into the unknown
The positivism strawman
I am often accused of being a ‘positivist’. Google’s version of the Oxford Dictionary defines positivism as, ‘a philosophical system recognizing only that which can be scientifically verified or which is capable of logical or mathematical proof, and therefore rejecting metaphysics and theism.’
This does not describe me.
I believe there are truths that cannot be scientifically verified or proved with logic or maths. In fact, most of what we deal with in education falls into this category. Let me give an example.
I tend to be in favour of school uniforms. To be clear, I don’t think school uniforms have much of an effect on anything and I think I could work quite effectively in a school with or without one. In a culture where school uniforms are uncommon, such as many American public schools, I don’t think it would be worth the fight to introduce one. However, I do think they have benefits, particularly for more marginalised students who cannot afford the latest fashions or those who are prone to being bullied for what they wear.
I am also of the view that uniform provides a ritual ground for rebellion. Teenagers are motivated to rebel against adult authority and so by rebelling over uniform, we offer a relatively harmless outlet for that. They can rebel, we can bring them back in line and nothing of importance is affected. Without uniform, their impulse to rebel may take more malign forms.
Can I scientifically prove this? No. But it is either right or wrong. So there is a truth out there—that this view of uniform is either right or wrong—that cannot be proved. Not only that, it matters and will affect the actions we take in schools.
Perhaps you disagree and believe this ‘ritual ground’ hypothesis could be proved one way or the other. If we tried using correlational data then we would have to deal with lots of confounds because schools with uniforms will differ from schools without uniforms in multiple and systematic ways. Perhaps we could run a randomised controlled trial and persuade schools to be randomly assigned to having a uniform or not. I cannot imagine this ever happening.
If you enjoy reading my posts and particularly if you find yourself sharing them, you should become a paid subscriber. This will give you access to my entire archive, as well as my weekly Curios posts that will keep you informed about everything worth knowing in the world of education.
Perhaps there is an even more significant point than the impracticality of the research. Imagine if we could run the experiment and we clearly demonstrated uniforms were associated with significantly better measures of behavioural outcomes, we would still not know whether this was because the ritual ground hypothesis is the reason or some other effect of uniforms. This latter point is true of all scientific theories. They exist in a provisional state, waiting to be proved wrong. I think a lot of people fail to appreciate this and paradoxically view science as certain.
So, we have something that is either true or false. It matters and it matters to me but I cannot scientifically prove it one way or the other.
Perhaps this example is a little technical so let’s try another one. Imagine we had scientifically validated evidence that giving students mild electric shocks enhanced learning. If I was a positivist, presumably, this evidence would be all that mattered to me. However, setting aside the small matter of this being illegal, I would not participate in any teaching system that used electric shocks on students because I believe that would be wrong. So I have now prioritised a moral conviction over scientific evidence. That does not seem like something a positivist would do.
I think educational progressivists do this often. They see an example of traditional teaching or explicit instruction and it offends their romantic views of what education should be. In essence, they find it immoral. This is why they talk of prisons or factories or robots and tend to characterise explicit teaching in moralising terms.
That’s fine. We can have that argument. We have processes in the world’s great democracies for adjudicating such disputes. I would also add that in societies with thousands of schools, there is the space for different models to co-exist and parents to have some degree of choice—the more the better, in my view.
However, the accusation of ‘positivism’ seems to be designed to avoid having this kind of debate. We don’t have to admit it is really about values.
This accusation also serves to imply there is something philosophically questionable about highlighting the available evidence for, say, explicit teaching or a knowledge rich curriculum and it does this by erecting a straw man, accusing opponents of having views they do not hold.
In reality, we need to reverse this perspective. I am not a positivist because positivism reflects a naive view of what science is and I value things other than empirical evidence. However, where evidence exists, I am strongly of the view that we should not ignore it and we should reckon with it somehow.
This is the trick that is being played. Accusations of positivism are a tactic for ducking the evidence we already have about the relative effectiveness of different teaching strategies. This is perhaps necessary for those whose beliefs come in conflict with this evidence.
No, educational progressivists are not superior beings with a more sophisticated philosophical perspective, they are just looking for a way to escape inconvenient truths.
I briefly kind of wondered if it was me who'd provoked this post by Greg, since my comment on last week's curios touched a little bit on epistemology. I'm currently taking courses on research methods in education/the social sciences, and we've seen the different epistemological paradigms that subtend different types of scientific research. The way it was explained to me, positivism is the paradigm according to which we can know exactly what reality consists of and understand relations of cause and effect between phenomena. It's a fairly naive position to take, so it was superseded by post-positivism, which stipulates that we can know an approximation of reality and understand relations of cause and effect in a probabilistic manner, through statistical analysis. This is the paradigm that subtends quantitative research. Qualitative research, on the other hand, can be based on a so-called constructivist epistemological framework, which stipulates that different versions of reality are constructed by the different participants and that research can consist of learning about these different constructed realities.
Now, this is all well and good. But it's true that when confronted with some inconvenient results produced by predictive quantitative research, some educationalists will retreat to the position that we cannot know for sure what works or doesn't work in education, it's naive to think otherwise, our students are not statistics but people in all their messiness, and we need to adopt a more nuanced view of what works in education and what research can tell us. Claims of "positivism", in other words. But I don't think this accusation is one that its targets would think accurately describes their views. Greg, for example, thinks research shows explicit teaching to be more effective for student learning than discovery learning. But he doesn't claim we know for sure that explicit teaching will be more effective for every student in every situation, or that we can predict for sure what a student will have learned after the process of explicit teaching. Greg adopts a post-positivist view of knowledge, not a positivist one.
What annoys me about some of these educationalists, and which I was complaining about in my comment on last week's curios, is that despite taking up the mantle of nuance and tarring claims about research that they disagree with with the label of positivism, they also aren't shy about themselves making claims about research results. I was reminded of this when reading this recent report that was brought to my attention by Professor Anna Stokke: https://sites.google.com/view/manitoba-math-myths. Briefly, the Canadian province of Manitoba (where I used to teach university-level mathematics for a few years) recently reduced the requirements to become a teacher, probably in response to a shortage of teachers. The thing about this that mainly attracted the attention and condemnation of Professor Stokke and her coauthors was that future elementary teachers in Manitoba will no longer be required to take any university-level mathematics course. However, the response by an education professor at the University of Manitoba was that said changes actually bring the process of teacher certification more in line with research in mathematics education. When asked for evidence for these claims, the professor responded with a series of cites, some of which having to do with Lee Shulman's classification of teacher knowledge, but none of which credibly support the professor's claims. Indeed, anyone who's read Shulman's seminal paper "Those who understand" will know that, first, it's an entirely theoretical paper, but also second, it actually insists on the importance of teacher content knowledge.
So in which way can the changes to the Manitoba teacher certification process be said to be in line with mathematical education research? Unless we define "mathematical education research" to be "some things mathematics education researchers talk about and believe, even though they have no evidence for it, and anyway belief in evidence is positivist". I honestly feel this sort of thing can fuel populism. We talk about how our decisions should be informed by science, but this sort of thing shows that "science" may not mean evidence, but simply some self-anointed researchers' opinions or (as suggested by Greg) moral intuitions that many people would disagree with. (I, for one, do not believe explicit teaching, or having students memorise times tables, to be immoral, for example.)
Thanks for writing this, Greg. I haven't been concerned about positivism being a problem in your work, but your piece helped me understand the problem in general. In many areas of life, not just education, I see people acting as if empirical research offers the only reliable source of information concerning a course of action. Such a position can be crippling because the research often isn't there and because, as you point out, scientific theories are provisional. Science provides valuable, but limited, guidance. We need to keep that in mind and use this knowledge appropriately. I wasn't aware that some educational progressivists are using charges of positivism to dismiss research, but I am not surprised.