Following methodological procedures and protocols is helpful. And empirical corroboration of the findings through applied research and community engagement might be useful.

Well, that's an interesting question. I am not sure about critical realism being a "metatheory" (whatever that may be) or about research needing metatheories, for that matter. However, the problem of when (with what evidence) a mechanism is validly identified is one that I keep encountering in my own work.

We need to start from the definition of a mechanism, in my case of a social mechanism. I use the following: A social mechanism is "a sequence of causally linked events that occur repeatedly in reality if certain conditions are given and which link specified initial conditions to a specific outcome". (Gläser and Laudel 2019: [4], based on Mayntz 2004: 241). From this follows that I must empirically identify specific sequences of events that occur repeatedly in my empirical material, which are triggered by conditions that are similar across cases and lead to outcomes that are similar across cases.

If I manage this, I can argue that I have a candidate for a mechanism. It depends on the amount and solidity of empirical evidence for me to claim that this is indeed a mechanism. The crucial point is the word "repeatedly" in the definition. If I find a sequence occuring only once, I have a candidate for a mechanism at best. If I find it occurring two or three times, I would still call it a candidate mechanims. Above that, I would probably be inclined to call the sequence a mechanism. However, this depends on the totality of the empirical evidence available. And I would always state that further research need to confirm that what I found is indeed a mechanism.

Andreas Vavvos Thank you for your response. I agree with you on the importance of adhering to methodological procedures. A question that continually arises for me is: how do I determine when I have identified a mechanism, and when can I consider my search to be comprehensive enough?

Jochen Glaser Thank you for your response. Firstly, I agree that a metatheory is not always necessary in research. However, I find the concept of metatheories, especially critical realism, particularly intriguing.

Your methodology for identifying social mechanisms—focusing on causally linked events that occur repeatedly—is fascinating. To fully comprehend these repeated causal links, do you employ any form of triangulation, or are these patterns discernible within the same dataset, such as through qualitative interviews? I'm curious about how different data sources or methodological approaches might enhance the reliability or depth of your findings, offering a more comprehensive view of the mechanisms at play.

I use qualitative interviews because I don't really see an alternative for getting the rich data you need for sequences of events and the conditions under which they occur. Documents sometimes play a role as additional sources of data. I don't see the point of using surveys because they follow a different logic.

There is a suggestion by philosophers that a complete explanation requires both a causal relationship (to be established with quantitative methods) and a causal mechanism (to be established with qualitative methods (the Russo-Williamson-Thesis). However, in my opinion this applies only if you want an explanation to apply to a particular population. Only in this case you need to know the mechanism and allinstances uin which it operates. Combining quantitative and qualitative methods that way also depends on both sets of methods operationalising the same concepts, which I think is extremely difficult to achieve.

Thus, while I agree that there is a general belief that more data and heterogeneous data provide a more comprehensive view, these data must describe the same phenomenon, which is rarely achieved.