For this blog, I read more research than I actually write about. I look at its relevance, its quality, and sometimes simply whether I understand it well enough. And sometimes I leave it aside because I am not yet sure what to do with it.

That is how I came across the work of Jukka Savolainen on sociology as a scientific discipline some time ago. It stayed with me. Not because it was entirely convincing, but because it touched on something you cannot easily dismiss. At the same time, I deliberately chose not to write about it until now. Precisely because the piece is clearly written as a provocation, and because its empirical basis is rather thin. That makes it tempting to use quickly in discussions, but also risky to adopt without caution. A more recent study now gives me the opportunity to discuss both with the necessary nuance.

For those unfamiliar with it, Savolainen argues that sociology as a discipline lags behind in methodological standards, such as open science and causal inference. His explanation is that the field is relatively ideologically homogeneous and that activist-oriented research plays a role in it. That is a strong claim. And although he supports it with comparisons across disciplines, it remains largely correlational work based on a very limited sample. He himself acknowledges that causality has not been established.

So the issue is not that there is nothing to it. The issue is that it is presented too quickly as an explanation.

And yet it kept nagging. Because what Savolainen points to is not entirely unfounded. He highlights a broader tension that also appears in other research.

A few days ago, for instance, a large-scale bibliometric study by Amy Shumin Chen was published. In this study, she maps the international landscape of the sociology of education, focusing mainly on the United States, the United Kingdom, and Taiwan, where she currently serves as president of the Association of the Sociology of Education.

What stands out in her work is not so much a problem as a pattern. In the United States, an empirical tradition dominates, strongly focused on measurement and policy. In the United Kingdom, the emphasis is much more on critical, theoretical, and discursive analyses. And in other contexts, such as Taiwan, there are attempts to combine the two, often without fully succeeding.

That in itself is not a weakness. On the contrary. It shows the field’s richness. But it also makes visible where the tension lies: between measuring and understanding, between evidence and critique, between policy and reflection.

And that is where both articles intersect.

Where Chen describes this tension as a historically developed diversity within the field, Savolainen interprets it as a problem of methodological weakness. That is a very different step. And it is precisely where we need to be cautious.

Because you can read the same observation in two ways. You can say that the field consists of multiple epistemological traditions that coexist and each contributes in its own way. Or you can argue that this diversity prevents the field from developing in accordance with stricter scientific standards.

The truth likely lies somewhere in between.

What is difficult to deny is that integration between these different approaches often remains limited. Empirically strong work and theoretical-critical analyses often coexist without truly engaging with one another. This is not a problem unique to sociology, as I will show shortly, but it is clearly visible here.

The question, then, is not which side is “right,” but how to make that tension productive.

For those, like me, working on evidence-informed education, this will sound familiar. There, too, you constantly see the tendency to choose: either hard evidence and measurability, or broader reflection and critical interpretation. Yet it is precisely the combination that becomes interesting when done well. It is something I try to do myself, though it often means navigating between paradigms.

I think this is not a unique story for the sociology of education. A similar analysis can be easily applied to education research and to the philosophy of education.

There, too, you see different traditions coexisting: research focused on effectiveness and measurement, work grounded in practice and context, and explicitly normative reflections on what education should be. These lines do not always align neatly. Sometimes they reinforce each other, sometimes they run in parallel.

That makes the field rich, but also complex. And it can make it difficult to arrive at shared standards or a common language.

So yes, if you wanted to, you could formulate a critique similar to Savolainen’s. You could argue that certain parts of the field place less emphasis on methodological rigour, or that normative convictions sometimes outweigh evidence. But again, if you move too quickly to that conclusion, you risk overlooking the nature of the field itself.

Education and philosophy of education are not only about what works, but also about what is desirable. And we should not reduce that latter question to causality or effect sizes.

Perhaps that brings us back to the same challenge. Not to let one approach prevail, but to strengthen the bridges between these different forms of knowledge. Because that is where things become truly interesting.