***** In this guest post, Simeon J. Newman and George Steinmetz reflect on some of the recent discussions about critical realism and present their own perspective on what they consider to be the essentials of a critical realist position******


The last several days have seen the topic of critical realism return to colloquial sociological discourse with the publication of Neil Gross’s review of two books by prominent critical realist scholars, Fabio Rojas’s amplification of it, and responses by Tim Rutzou and Jeff Guhin

But this discussion has been all but devoid of reference to the core features of critical realism. While Gross aims some choice words at the critical realist works he reviews, he does not ask, much less answer, the bigger question of whether the features of the works he reviews are critical realist in more than name, and if so, whether they are good or bad specimens of it. Since his comments are limited to a book review, Gross cannot really delve into his own views on the philosophy of social science, or examine alternative critical realist positions.

Rutzou responded to Gross (and to Rojas’s amplification) by focusing on its meta-theoretical nature and noting that critical realism is really very heterogeneous (as he has emphasized before; see his article What is Critical Realism, and his article on Bhaskar and Deleuze regarding the heterogeneity and open-mindedness of critical realism), but also not identifying it substantively. Guhin’s response focuses primarily on his pleasant engagement with critical realists and the Critical Realism Network’s events.

This entire recent debate is turning out to be a “dialogue des sourdes.” It is as if someone reviewed a work by a neoKantian dismissively and another defended other neoKantians, with both claiming to be speaking about Kant.

Curiously, if one were to take sides on the debate on critical realism as currently structured she would be doing so without considering the core features of critical realist thought. We take this as an invitation to outline what we think are the two essential features of critical realism and its main ancillary commitments. We think that one would have to engage these ideas to figure out where she stands (yea, nay, abstain) on the topic critical realism.

For us, at its core, critical realism consists of two philosophical commitments and a couple of other desiderata that together constitute a meta-theory [1].

  1. Ontological realism — this is the belief that there are things that exist and have effects in the world even if we are not currently aware of them. For example, imagine for a moment that we had no knowledge of patriarchy and no one ever called anything “patriarchical.” For a critical realist, patriarchy would still exist, for it is not a matter of names, concepts, or discourses alone. For an antirealist of the relevant kind, without this kind of discourse of patriarchy, patriarchy would not exist. This kind of ontological realism is a response to rampant skepticism and narrow-minded empiricism.

Critical realists conceptualize ontology using the ideas of “stratification” and “emergence.” The most profound level is the “real,” which consists of that which really exists even though we may be ignorant or have only indirect knowledge of it. The second level is the “actual,” or that about which we can, in principle, gain empirical knowledge of, but which, for whatever reason, we don’t yet know about. Third and finally is the “empirical,” that which we can and do have empirical knowledge of. The actual and empirical are “emergent” in the sense that they depend upon but are not reducible to the real. Sociologists at least since Durkheim have argued that the social depends upon but is not reducible to the biological or natural. Critical realists refer to a related but different, metaphysical concept, and maintain that the actual and empirical may even exert “downward causation” (i.e., have affects) on underlying real causal entities.

An additional ontological premise is the open system in which multiple causes combine to produce patterned regularities and historically-unique events. Critical realists maintain that all of social and even most of the natural world is an open system. 

Taken separately, none of these ontological ideas is distinctively critical realist, but as a package or cluster they do identify a distinctive position.

From this basis, critical realism posits a definition of science as the pursuit of truths about the “real,” and labels the objects of scientific inquiry “mechanisms.” According to critical realism, good sociologists try to identify the effects of, discover new, and refine the accounts of causal mechanisms. (Note that since critical realists have the ideas discussed in mind, they may use the term “mechanism” differently from, for example, analytic sociologists. And some critical realists prefer terms like “entity,” “structure,” or “causal power,” since the term “mechanism” can seem too mechanistic and harbor unwanted assumptions.)

Importantly, critical realists conceive of mechanisms as operating in open systems in which they may interact and interfere with one another, leading them to be suppressed, subdued, or fused with one another.

Whether or not any such nemeses actually exist, critical realists often think of this kind of sociology as an alternative to both a “Humean” [2] view of sociology, which would say that sociologists principally identify correlations between observed variables and remain skeptical of and therefore aloof from any serious discussion of causes, and a strong social constructionism, which would say that we can’t actually use social inquiry to get at anything “real” (in the relevant sense) and therefore to try to do so is to kid ourselves. Critical realists think that we can (tentatively) identify the causes of correlations and outcomes, even if we must always be humble and realize that our conclusions are fallible.

  1. Epistemic relativism — this is the belief that our ways of acquiring knowledge should change as a function of what we acquire knowledge about, rather than the reverse. (This stems from the position that we only know things in the world under particular descriptions.) When we have new findings about the same exact object of inquiry, the discrepancy between the old and the new findings (barring measurement error) can only be attributable to the use of different epistemological approaches. Critical realism maintains that one’s epistemology should depend on the object of inquiry, or be accountable to the “real” mechanisms (and their interactions, fusions, etc.). To have it the other way around, say critical realists, is to commit the “epistemic fallacy,” or succumb to “the view that statements about being can be reduced to or analysed in terms of statements about knowledge” (Bhaskar, A Realist Theory of Science). (Of course, critical realism embraces fallibilism, maintaining that our conclusions–our descriptions about things in the world–may be wrong.)

One useful way of thinking about critical realism’s epistemic relativism is as an alternative to epistemological dogmatism. For example, Karl Popper put forth the idea that the only admissible scientific method consists in trying to falsify our hypotheses in order to “obtain the fittest theory within our reach by the elimination of those which are less fit” (Popper, Conjectures and Refutations). Such approaches have been rightly criticized for their inability to produce new findings. Thus epistemological relativism seems pretty much like a basic requirement for empirical research beyond mere confirmation studies. For critical realists, it is also important because mechanisms interact with one another in open systems, sometimes counteracting each other or subduing one another’s effects. This would lead a Popperian to incorrectly reject (on the basis of falsification) the theory being tested and deny the existence of the causes within that theory.

Beyond these two basic philosophical commitments, a minimalist critical realism entails two other positions. First is an endorsement of the idea of judgmental rationality–the belief that some explanations are better than others, and that we should be able to judge them using (relatively) rational criteria. Second is explanatory critique–the view that we can and should simultaneously explain and criticize what we study. We see these as ancillary to the core of critical realism, so we will not dwell on them here. But it is worth pointing out that here, too, as in the rest of critical realism (and indeed, as critical realism would suggest), current solutions to the problems of judgmental rationality and explanatory critique are themselves fallible. There is, for example, strong disagreement among critical realists on the very topic of critique. Some hew to a neo-Aristotelean view of virtue ethics, while others lean toward pragmatism or universalizing Kantian ethics. Others–ourselves included–believe that more is to be gained from further interaction between critical realism and the traditions of immanent critique in critical theory.

Clearly there is a lot at stake with accepting any or all of the philosophical positions sketched above. There are stakes in rejecting any or all of them as well. Anyone who is interested in critical realism (friendly, hostile, or otherwise) should at least be attuned to its claims about ontological realism (the real, open systems) and epistemological relativism (along with the idea of the epistemic fallacy), and perhaps also to its claims about the possibility of judgmental rationality and explanatory critique. Whether or not particular critical realists’ analyses are valid, this array of concepts form a minimalist critical realist metatheoretical position. If one desired to critique, dismiss, or defend the critical realist position, this is the place where the discussion must be had. These claims could be criticized or defended singly or as a group, but attacks on and defenses of other things would be to miss the point.


[1] We base ourselves on Roy Bhaskar’s first two books, A Realist Theory of Science and The Possibility of Naturalism.

[2] Named for David Hume’s account of causation as “constant conjunctions.”